Autonomy in Exchange for Accountability:An Initial Study of

Pennsylvania Charter Schools

Gary Miron and Christopher Nelson

The Evaluation CenterWestern Michigan University

October 2000

i

Autonomy in Exchange for Accountability:An Initial Study of Pennsylvania

Charter Schools

Executive SummaryGary Miron and Christopher Nelson

The Evaluation CenterWestern Michigan UniversityKalamazoo, MI 49008-5237

tel. 616 387-5895<www.wmich.edu/evalctr/>

October 2000

This is the final report of the 17-month initial study of Pennsylvania charter schools. The WesternMichigan University Evaluation Center has conducted this evaluation pursuant to a contract with thePennsylvania Department of Education (PDE). The initial study is a key research component of theCommonwealth�s overall accountability plan for charter schools. Section 1728-A of Pennsylvania�scharter school law (hereafter known as Act 22) requires an evaluation of the charter school programafter 5 years. While the 5-year report is to be largely summative in nature (providingrecommendations on the advisability of continuing or amending the program), the initial study isdesigned to be largely formative and to provide feedback to schools and policymakers regardingchanges that can be made to help these schools function more effectively and achieve theiranticipated goals. At the same time, this initial study is an important component of theCommonwealth�s overall accountability plan for charter schools and it provides a foundation forthe 5-year legislatively-mandated evaluation.

Charter schools are a new form of public schooling intended to provide alternative and diverseeducational programs, with the goal of improving academic achievement. Charter schools operateunder a contractual arrangement with a chartering entity, in the case of Pennsylvania this includesthe local school districts. The chartering contract frees schools from most of the rules andregulations that apply to traditional public school systems in exchange for increasedaccountability�ultimately, high student academic achievement. Charter schools can be formed bya variety of individuals or groups, including educators, parents, community members, for-profit andnonprofit organizations, and institutions of higher education. A charter is signed by its foundingmembers and a chartering agency and details what the school expects to accomplish with respect tostudent achievement and other outcomes. Unlike traditional public schools, charter schools may beclosed by their sponsoring entity if they fail to meet the standards set forth in the charter.

ii

The autonomy-accountability bargain creates for charter schools an �opportunity space� withinwhich they operate. Two general questions follow from this:

! How are charter schools using their autonomy?

! Are these uses of charter school autonomy leading to the student outcomes for which chartersare held accountable?

Indeed, the central evaluation question stated in the request for proposal (RFP) for this study is,�Does increased flexibility in exchange for increased accountability result in improved pupilresults?� Moreover, the RFP asked a number of more specific questions about uses of charter schoolautonomy and their impact on various educational outcomes.

! What effect does budget have on student results, nonacademic services, and school facilities?

! Are opportunities offered to charter school teachers, parents, and students to influenceclassroom and school policy significantly different from those offered at traditional publicschools?

! Are the opportunities (i.e., professional growth, salaries, benefits, employee rights) for teachersand other employees significantly different at a charter school than at a traditional publicschool?

! What is the impact of charter schools as related to district reform efforts?

! Is there evidence that, over the term of the charter, student learning has significantly improved?

! What are promising practices in charter schools that could be included in district systemicreform?

Methods and Limitations of the Study

The questions addressed in this study required multiple approaches for collecting and verifyinginformation. Thus, the study�s methodology represents a blend of quantitative and qualitativemethods. Our aim was to collect enough information to analyze all charter schools individually aswell as among and between groups of schools with similar characteristics. Since our mandate wasto evaluate the charter school reform rather than individual schools, this final report focuses ongeneralizations across the charter schools and does not make judgments about individual schools.Decisions regarding the nature and type of data to be collected were made with the PennsylvaniaDepartment of Education to ensure that they are worthwhile and of interest to decision makers andother interested parties. All decisions regarding the study�s design and methodology were guidedby The Program Evaluation Standards developed by the Joint Committee on Standards forEducational Evaluation.

The following data collection and analysis methods were used:! Surveys of staff, students, and parents (charter schools surveys developed by The Evaluation

Center and nationally-normed school climate surveys)

iii

! Reviews of (student) work samples (when available)! Interviews ! Diaries and logs (if available)! Document review! Portfolios (if available) ! Direct observation! Focus group meetings! Analysis of test scores and available demographic and financial data

Most of the analyses presented in this report are based on the 31 charter schools operating during the1998/99 academic year (1 of which closed at the end of the 1998/99 year). Where possible, the studyincludes secondary data on the additional 17 schools that opened during the 1999/2000 academicyear. Surveys were administered and interviews were conducted during both the 1998/99 and1999/2000 academic years. The evaluation team sampled all teachers and key administrators anda random selected 3 classes of students and between 25-35 families at each school. The evaluationteam worked diligently to ensure high response rates across all of the schools. Student response ratesfor each year were approximately 90 percent, while teacher response rates largely ranged from 73to 83 percent. Parent response rates were considerably lower (approximately 50 percent).

The study�s ability to provide evaluative conclusions was limited by the small number of charterschools and the relative newness of the charter school movement in Pennsylvania. Fortunately, bothproblems will be self-correcting as the movement continues to grow and mature.

Patterns of Growth in Pennsylvania Charter Schools

One of the key evaluation questions is whether the charter school law is providing students, parents,and teachers with new alternatives within the public school system. Thus, the report began bysummarizing growth trends in Pennsylvania charter schools.

! As of the 2000/01 academic year, there will be 66 charter schools in operation in Pennsylvania.A 67th school was closed after its first year of operation. Taken as a whole, these schools willenroll more than 20,000 students, or just over 1 percent of all public school students inPennsylvania.

! Throughout the life of the charter school law, there has been relatively steady growth both inthe number of charter schools and charter school students. Most of these schools are smallerin size (i.e., average enrollment is approximately 265 students) than the typical Pennsylvaniapublic schools but larger than the typical charter school nationwide. There is, moreover,evidence of a trend toward larger charter schools in Pennsylvania.

! Charter schools are concentrated in certain parts of the Commonwealth, particularlyPhiladelphia. Indeed, while Philadelphia enrolls approximately 11 percent of the public schoolstudents in the state, it has 51 percent of the charter schools and 69 percent of the charter schoolstudents.

iv

! Charter schools appear to target students of a reasonably wide variety of grade levels. Thereis, however, a tendency for charter schools to seek to serve students in the lower reaches of theelementary, middle, and high school grades. Similarly, charter schools� mission statementsindicate that they intend to serve a wide variety of educational interests and goals. However,there is evidence that a significant proportion seek to serve at-risk students.

Charter School Start-Up

The range of charter school alternatives depends, in the first instance, on founders� commitment toand skill in gaining charters and opening schools. For this reason, the report provides an analysisof the characteristics of the founding coalitions that seek charters and start up charter schools.

! Charter schools appear to be born of dissatisfaction with noncharter public schools, asevidenced by low PSSA scores. Lower performing districts, in turn, tend to have higherconcentrations of poor and nonwhite students.

! The founding coalitions behind charter schools often include public school teachers andadministrators, academics, and members of the business community. There is little evidencethus far that parents have played a significant role in founding charter schools (though they dobecome active in the operation of charter schools).

! Founding coalitions� goals have included providing a choice for poor children, creating a venuein which to operationalize ideas and practices hindered by district practices, promoting changein noncharter schools, and inculcating a particular ethnic or cultural perspective. Privateconversion charter schools, moreover, are often founded with an eye toward scaling up practicesalready employed in private schools and attracting a broader base of students.

! There is evidence that in order to be successful, founding coalitions often need to musterconsiderable political resources. In some cases this has involved tapping into support forcharter schools among influential community leaders and the more general dissatisfaction withpublic school bureaucracies. In some instances, however, it appears that host districts have seenadvantages in sponsoring charter schools, making such political tactics unnecessary.

! Some 78 percent of the charter schools in operation as of the 1998/99 academic year are closelyaffiliated with some sort of nongovernmental organization, including community developmentgroups, ethnic/cultural groups, and other nongovernmental organizations. Many schools arealso built upon the foundation of a preexisting school.

Charter School Finance

One of the most important inputs into charter schools is money. Therefore, the report examinedcharter school revenues and also provided an analysis of charter schools� expenditure patterns andsome simple indicators of financial viability.

v

! Analysis of charter schools� financial reports indicates that the median charter school receivedapproximately 81 percent of its total revenues from LEAs during the 1998/99 school year.There is, however, a large amount of variation among charter schools, with some charterschools receiving as little as 17 percent and some as much as 89 percent of their revenues fromother LEAs. Much of this variation, of course, is explained by the fact that some schools spendmuch more per pupil than others and that the size of the LEA varies with district expenditures.Beyond that, schools vary in their reliance on non-LEA revenue sources.

! Next to LEA transfers, the largest revenue source for charter schools is the federal government,mostly through Title I monies and special charter school grants. While there is considerableschool-by-school variation, the median charter school received approximately 7 percent of itstotal revenues from the federal government. The remainder of charter school revenues camefrom state grants (e.g., start-up grants) and �local� sources, including earnings on investments,charitable donations, and revenues from student activities (e.g., candy sales, car washes, and soon). In addition, a few schools relied on proceeds from extended term financing during the1998/99 school year.

! Analysis of charter schools� financial data indicates that the median charter school spentapproximately the same amount per pupil as its host district during 1998/99. Of that totalamount, charter schools typically spent a smaller percentage (59 percent) on instructional itemsthan their host districts (66 percent) and a larger percentage on support services (which includerenting and maintaining facilities) and on other noninstructional items.

! Analysis of financial reports indicates that charter schools appear to be relatively conservativein budgeting, taking in more than expected on the revenue side and spending less than expectedon the expenditure side of the ledger. Moreover, there is some evidence that schools benefitfrom experience, as second year schools had slightly lower expenditure variances than first yearschools (there was no discernible relationship on the revenue side).

! Analysis of end-of-year balances indicates that most schools ran surpluses for the 1998/99school year. Seven charter schools (23 percent), however, showed negative balances (deficits),the largest of which constituted 10.7 percent of total expenditures.

Student and Parent Characteristics

Another important set of inputs to charter schools is the people who attend, send their children to,and work/volunteer at the schools.

! On the whole, charter schools enroll roughly equal proportions of males and females.

! The proportion of low income students enrolled in charter schools is roughly comparable to hostdistricts; in both groups approximately two-thirds of students are eligible for free or reduced-price lunch.

! Charter schools, however, enroll a significantly higher proportion of nonwhite students (80percent) than their host districts (57 percent) and charter schools nationwide (52 percent).

vi

! Most charter school students (80 percent) previously attended a public school, while 17 percentattended a private school and 3 percent other schools. The proportion of students coming fromprivate schools roughly corresponds to the overall proportion of private school students in theCommonwealth (15 percent). Nevertheless, there is significant variation among charter schools.

! Evidence from surveys suggests that parents choose charter schools mainly because of theperceived quality of instruction, the school�s academic reputation, dissatisfaction with theirchild�s previous school, and because they agree with the school�s educational philosophy.Indeed, some 88 percent of parents sampled indicated that they are aware of their school�smission.

! The most common reason students cited for attending their charter school was that their parentsbelieve the school is good for them. Beyond that, students cite school safety, teacher quality,small classes, and dissatisfaction with their previous schools as the most important factorsdriving their choice. Approximately 60 percent of students report that they are aware of theirschool�s mission.

! Only one-quarter of surveyed parents indicated that they volunteer more than 3 hours per monthat their charter school. This is in spite of the fact that approximately half report that their schoolrequires such involvement.

Teacher Characteristics

Teachers and staff represent another important human capital input for charter schools.

! The majority of teachers and staff are female (71 percent), while the gender split foradministrators and directors is close to equal. This generally matches the gender distributionin noncharter public schools.

! Charter school teachers are generally younger than their counterparts in noncharter publicschools, with approximately 50 percent under the age of 30 compared with 11 percent in otherschools.

! Charter school teachers are also more likely to be nonwhite than their counterparts in otherpublic schools. Approximately 40 percent of teachers are nonwhite, with African Americanscomprising the largest group, followed by Hispanics, Asian/Pacific Islanders, and NativeAmericans.

! The proportion of charter school staff who report that they are classroom teachers (52 percent)is similar to the proportion in noncharter public schools in Pennsylvania (49 percent) and publicschools nationwide (52 percent).

! Some 76 percent of teachers surveyed in 1999/2000 reported that they are currently certified toteach in Pennsylvania. This represents a decrease from 82 percent in the previous year. On theother hand, the percentage of teachers certified to teach in other states rose from 2 percent to4 percent over the same period, as did the percentage of teachers working to obtain certification(12 to 17 percent).

vii

! The vast majority (75 percent) of teachers with university degrees had attained a BA as theirhighest level of education. However, a third of the teachers stated they are working towardanother degree.

! The average experience level of the teachers surveyed was just under five years.

! The most commonly cited reason for joining the faculty of a charter school was an interest inbeing involved in a school reform effort, followed by the opportunity to work with like-mindededucators. Other popular responses included small class sizes, safety, and the school�sacademic reputation.

! Some 97 percent of teachers indicated that they are familiar with their school�s mission. Ofthese teachers, 72 percent believe that their school�s mission is being followed �well� or �verywell.�

Professional Opportunities for TeachersOne of the goals of Act 22 is to provide enhanced autonomy and professional developmentopportunities for teachers. ! Many teachers have come to charter schools seeking autonomy in creating and implementing

curriculum. ! There is evidence at a number of charter schools that there was a conscious effort to involve

teachers in developing curricula. ! Teachers indicated that they had autonomy in curriculum decisions and freedom to utilize

creative approaches. Indeed, many teachers report that they have considerable flexibility andopportunities for creativity in their day-to-day activities.

! There is clear evidence in slightly more than half of the charter schools that the teachers in theschools work collaboratively.

! Evidence from teacher surveys indicates that approximately one-quarter to one-third of teachersare satisfied with their school�s physical facilities, while one-third to one-half were satisfiedwith resources available for instruction and other educational functions.

! The average charter school paid teachers an annual salary of approximately $30,000, comparedwith the state average of approximately $48,000 and an average of $40,000 for schools withsimilar levels of teacher education and experience and similar per pupil expenditures. In spiteof this, some 30 percent of teachers report that they are satisfied or very satisfied with theirsalaries.

! There was a measurable difference between initial teacher expectations and current experienceon many topics, the largest of which were the effectiveness of leadership and administration,communication between parents/guardians and the school, availability of support services tostudents, parents� ability to influence the direction and activity of the school, and the extent towhich students receive sufficient individual attention. There was also a large gap betweenexpectations and experience in teacher empowerment and the degree to which they are able toinfluence the steering and direction of the school.

viii

! Other areas of concern for teachers included class size, emphasis on academics, and parentalinvolvement.

! Despite reported problems, many teachers reported that they are satisfied with their teachingenvironment, and about 75 percent planned to returned to the school the following year.

! Charter schools devote considerably more time to teacher professional development activitiesthan noncharter schools, with the average charter school allotting 7 professional days andnoncharter schools 5 days. A strong emphasis on graduate study was frequently reported, withsome programs requiring it.

Innovations in Governance, Curriculum, and Instruction

One of the stated goals of Act 22 and other charter school statutes is to encourage the developmentand diffusion of innovative practices. In this sense, charter schools are to be public education�s�R&D.� The report examines innovations in leadership, organization, discipline, curriculum, andinstruction.

! Teachers and parents generally expressed approval of their school�s leadership, with 67 percentof parents and 53 percent of teachers indicating that the statement �This school has goodadministrative leadership� was true.

! Approximately 70 percent of teachers agreed that their school�s leaders set high standards,communicate them effectively to others, and set a good example by working hard themselves.

! Parents, teachers, and students, moreover, often have enhanced opportunities to participate inschool decision making. These include (a) involvement in the strategic planning process, (b)attendance at planning meetings, (c) attendance at school board meetings and other avenues ofaccess to board members, (d) input via school surveys and interviews, and (e) formal appealsand grievance processes.

! Approximately half of teachers (54 percent) said that they are involved in decision making attheir school. A similar proportion of students report that administrators listen to their ideasabout the school. Approximately 83 percent of parents, moreover, said that it was true or partlytrue that �I am able to influence the direction and activities in the school.�

! Many teachers indicated that their workloads did not leave them enough time to remaininvolved in school decision making. Parents cited work and family obligations as barriers totheir participation.

! Notable organizational and disciplinary practices include extended hours and Saturday sessions,inclusion of both middle and high school grades in order to address common problems inmaking the transition from middle to high school, and preventive disciplinary measures thatfocus on behavior modification techniques.

! Curricular innovations in charter schools often come in the form of a focus on a particularethnic or cultural perspective. Other schools focus on social and vocational skills, with some

ix

offering individualized instructional plans for all students. In some cases, charter schoolsemploy packaged programs such as Success for All reading, Everyday Mathematics, orDiscovery Works.

! Evidence of charter schools� use of technology was scant and allowed for no firm conclusions.However, the weighted average number of students to computers in charter schools is 4.5,compared with 7.5 for their host districts. At least one school maintains student portfolios on-line, and at least one school offers on-line courses that students can take from home.

Special Education

Special education presents a challenge for all public schools, especially charter schools. Thus, thereport provides a preliminary assessment of special education in charter schools.

! There is an important conflict between the spirit of charter schools laws, which seek toderegulate charter schools, and the preemption of special education by federal law. To itscredit, since enactment of Act 22, the Commonwealth has made considerable headway inclarifying charter school obligations and providing guidance to charter schools in proposing newregulations in this area, and also in providing support and technical assistance to help charterschools provide special education services that are in compliance with IDEA.

! The average proportion of students with disabilities in the charter schools was 10.5 percent in1999-00, which is slightly lower than the statewide average of 12.5 percent. Among the schoolsthat opened during the first two years of the reform, the proportion of students with disabilitieswas higher (12.1 percent) than for the 17 schools that opened for the 1999/2000 school year (7.8percent). However, there was great variation among the schools, with some schools reportingas many as 42 percent special education students and some reporting no such students.

Student Achievement in Charter Schools

The charter school �bargain��autonomy in exchange for accountability�requires that charter schoolsdemonstrate improved student outcomes. Thus, the report provides a preliminary assessment ofstudent achievement in Pennsylvania�s charter schools. A number of caveats are discussed in thereport regarding our analyses of student achievement. Nonetheless, the evaluation presents somesuggestive findings that might be useful in program improvement.

! The data on student achievement in charter schools precludes conclusive statements aboutcharter schools� impacts on student learning due to a number of limitations. Among thelimitations the following are most noteworthy: (i) the charter school initiative is still quiteyoung, leaving charter schools with little time to demonstrate their ability to improve studentachievement; (ii) data was not available for many of the charter schools; (iii) no data exists forsuch important characteristics as students� precharter school achievement rates, which seriouslydiminishes the evaluation�s ability to provide valid assessments of charter school impact. Forthese reasons generalizations to the broader movement are tenuous.

x

! Charter schools as a group produced Pennsylvania System of School Assessment (PSSA) scoresthat were considerably lower than all noncharter public schools in the Commonwealth. Suchcomparisons, however, probably measure differences in the types of students who choose toattend charter schools more than any impact the charters have on their students.

! Four schools for which time series data are available posted gains of more than 100 points (thescale ranges from 1000 to 1600). In doing so, these schools as a group outgained their hostdistricts as a group by some 86 points over the same period. Single cohort, pre/post commercialtest data (e.g., ITBS, SAT-9) from a nonrepresentative sample of charter schools also show thata substantial number of charter schools posted gains in student achievement.

! A cross-sectional comparison of each charter school with its host district(s) found that charterschools as a group were outperformed by approximately 50 points on the PSSA. However,there is evidence that host districts provide flawed comparisons to charter schools.

! Likewise, a cross-sectional comparison of each charter school with demographically similarnoncharter public schools found that charter schools as a group were outperformed by theirapproximately 50 points on the PSSA. However, data limitations restricted the analysis tocomparisons on the basis of family income (as indicated by the proportion of students qualifyingfor free or reduced lunches) only.

Other Indicators of School Quality

There is legitimate debate about precisely what types of student outcomes charter schools should beheld accountable for. While most stakeholders seem to agree that student achievement is animportant (if not the only) goal of charter schools, others argue that schools should also be judgedon their ability to satisfy their customers. ! Survey evidence suggests that as a group charter school students report that their level of

academic performance has improved since they moved to a charter school. Teachers areperhaps a little less sanguine, with most initially expecting that student achievement would haveimproved more than it has. Still, more than half the teachers believe that student achievementis on the rise at their school.

! Approximately two-thirds of parents and teachers reported that charter schools are serving needsnot well served by other schools.

! A nonrandom sample of charter schools indicates that the average school has a waiting list of125 students, or 45 percent of total current enrollment.

! A nonrandom sample of student rosters indicates that the average charter school lost 38 percentof its students from Spring 1999 to Spring 2000. Moreover, data from the Philadelphia schooldistrict indicates that some 1,800 students left Philadelphia charter schools to return to districtschools.

! In spite of the significant enrollment instability in some charter schools, more than 90 percentof parents surveyed report that the quality of instruction in their charter school is high and thattheir child receives adequate attention.

xi

! Approximately half of the students surveyed said they would recommend their charter schoolto a friend.

! Charter schools had an estimated attendance rate of 90 percent for both 1998/99 and 1999/2000.! Nearly three-fourths of parents and teachers reported that their charter schools had high

expectations for students. ! Nearly three-fourths of students said that their teachers encourage them to think about their

future. ! Approximately one-third of students said that other students at their charter school were more

interested in learning than students at their previous school. ! Fewer than one-half of respondents report being fully satisfied with school facilities.

Issues for Further Consideration

This formative evaluation seeks to identify program strengths and weaknesses in order to facilitateimprovements. It also provides a foundation for a 5-year summative evaluation to be released in2002. Thus, we conclude by identifying a number of important policy issues and research questionsraised by this report.

Issues to be considered in future evaluation and research activities include those listed below:! How effective is the charter approval process as a quality control mechanism? ! Do some district approval and oversight processes lead to stronger charter schools than others?! Do charter schools approved on appeal face a hostile environment? If so, does this affect school

quality and student outcomes?! What explains variations in charter schools� expenditure patterns?! Do charter schools offer more educational value for the money than noncharter public schools?! Will observed trends in charter school affiliation with nongovernmental organizations continue?! Will the role of private management companies continue to increase over time? And what

effects might this have on the autonomy and effectiveness of the schools?! What explains variations in student achievement across charter schools?! Why do students and parents choose to leave charter schools?

Policy issues that should be considered include these:! What are the implications of charter schools� reliance on nongovernmental organizations for

school quality and the future growth of the movement?! How should charter schools balance the provision of services for special needs students with

the movement�s goal of providing head-to-head competition with noncharter public schools?! How should charter schools, local authorizing agencies, and other stakeholders balance market

versus other forms of accountability in decisions to grant and renew charters?

xii

Table of Contents

page

Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiiList of Tables, Figures, and Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xviList of Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xixAcknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xx

Chapter 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1 What Are Charter Schools? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.2 Objectives of the Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.3 Structure of the Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41.4 Challenges Confronting the Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

Chapter 2 Methodology of the Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.1 Approach and Strategies for Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . 82.2 Specific Methods for Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112.3 Data Analyses and Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.4 Limitations to the Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Chapter 3 A Snapshot of Pennsylvania�s Charter Schools . . . . . . . . . . . . 18

3.1 Number of Charter Schools and Students . . . . . . . . . . . . . . . . . . . . . . . . . . 183.2 Spatial Patterns in Pennsylvania Charter Schools . . . . . . . . . . . . . . . . . . . . 213.3 Mission and Target Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283.4 Education Management Organizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313.5 Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

xiii

Chapter 4 Charter School Start-Up: Resources and Challenges . . . . . . . 33

4.1 Legal and Administrative Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334.2 The Charter Appeals Board . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364.3 Founders� Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 394.4 Founders� Organizational Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404.5 Types of Individuals Involved in Charter Founding Coalitions . . . . . . . . . . 424.6 Political Constraints and Opportunities . . . . . . . . . . . . . . . . . . . . . . . . . . . . 434.7 Characteristics of Chartering Districts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444.8 A Digression on the �Philadelphia Question� . . . . . . . . . . . . . . . . . . . . . . . 474.9 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Chapter 5 Charter School Finance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

5.1 Revenue Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 505.2 Expenditure Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535.3 Fiscal Viability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 575.4 Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Chapter 6 Student and Family Characteristics . . . . . . . . . . . . . . . . . . . . . . . 63

6.1 Sampling of Students and Parents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 636.2 Description of Charter School Students and Families . . . . . . . . . . . . . . . . . 656.3 Reasons for Choosing Charter Schools, Responses from Sampled

Students and Parents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 736.4 Awareness of School Mission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 776.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

Chapter 7 Teacher and Staff Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . 79

7.1 Description of Charter School Teachers and Staff . . . . . . . . . . . . . . . . . . . . 797.2 Educational Background and Years of Experience of Pennsylvania

Charter School Teachers and Staff . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 837.3 Reasons to Seek Employment at a Charter School . . . . . . . . . . . . . . . . . . . . 887.4 School Mission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 897.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

xiv

Chapter 8 Working Conditions, Professional Development and Levels of Satisfaction for Charter School Teachers . . . . . . . . 91

8.1 Working Conditions for Teachers and Staff and Levels of Satisfaction . . . 928.2 Initial Expectations and Current Experiences of Teachers and Staff . . . . . . 968.3 Professional Development in Pennsylvania Charter Schools . . . . . . . . . . . 988.4 Teacher Induction Plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1018.5 Turnover of Teachers and Staff in Pennsylvania Charter Schools . . . . . . . 1028.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

Chapter 9 Innovations in Pennsylvania Charter Schools: Governance, Curriculum, Instruction, and Assessment . . . . 104

9.1 A Framework for Thinking About Innovation . . . . . . . . . . . . . . . . . . . . . . 1059.2 Organization and Governance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1089.3 Curriculum and Instruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1139.4 Conceptual Issues Related to Innovation . . . . . . . . . . . . . . . . . . . . . . . . . . 1189.5 District Impact and Transfer of Innovations . . . . . . . . . . . . . . . . . . . . . . . 1209.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126

Chapter 10 Special Education and Pennsylvania Charter Schools . . . . 128

10.1 Federal and State Legislative Requirements . . . . . . . . . . . . . . . . . . . . . . . 12910.2 Pennsylvania Charter School Legislation and Special Education . . . . . . . 13010.3 Pennsylvania Charter Schools and Special Education . . . . . . . . . . . . . . . 13410.4 Related Survey Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14010.5 Summary and Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143

xv

Chapter 11 Performance Accountability: An Analysis of Student Achievement in Charter Schools . . . . . . . . . . . . . . . . . . . . . . . 145

11.1 Data and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14611.2 Limitations to the Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14911.3 Findings from the Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15311.4 Recommendations for Improving the Evaluability of

Pennsylvania Charter Schools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16611.5 Summary and Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168

Chapter 12 Other Indicators of School Quality . . . . . . . . . . . . . . . . . . . . . 171

12.1 Data and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17112.2 Subjective Ratings of Student Performance . . . . . . . . . . . . . . . . . . . . . . . 17212.3 Market Accountability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17312.4 Attendance and Discipline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17512.5 Other Attributes of School Environment . . . . . . . . . . . . . . . . . . . . . . . . . 17712.6 Comparisons with National Norms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17912.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

Chapter 13 Summary of Findings for Further Consideration . . . . . . . . 184

13.1 Summary of Main Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18413.2 Research and Evaluation Issues for Further Consideration . . . . . . . . . . . 19313.3 Policy Issues for Further Consideration . . . . . . . . . . . . . . . . . . . . . . . . . . 195

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198

Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201

xvi

List of Tables, Figures, and AppendicesT a b l e s

Table 2:1 Data Collection Strategies and Information Sources . . . . . . . . . . . . . . . . . . . . . . . . . 9Table 2:2 Matrix of Objectives of the Study and Sources of Information . . . . . . . . . . . . . . . . 10Table 2:3 Sample Size and Response Rates on Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13Table 4:1 Regional Charter Schools, Summer 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35Table 4:2 Organization Bases of Pennsylvania Charter Schools . . . . . . . . . . . . . . . . . . . . . . . 41Table 4:3 Chartering and Nonchartering Districts Compared . . . . . . . . . . . . . . . . . . . . . . . . . 45Table 5:1 Charter School Per Pupil Expenditures Compared with Host Districts and All

Pennsylvania Public Schools, 1998/1999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54Table 5:2 Percentage of Total Expenditures Devoted to Instructional Items, 1998-1999 . . . . 55Table 5:3 Instructional Expenditures as a Percentage of Total for Charter Schools

Minus the Same Percentage for Host Districts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56Table 5:4 Expenditure Patterns in Charter School and Other Public Schools . . . . . . . . . . . . . 57Table 5:5 Variances in Revenues and Expenditures as a Percentage of Budgeted Revenues

and Expenditures, 1998-1999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58Table 6:1 Proportion of Low Income Students in Charter Schools, Sending

Districts, and Nonsending Districts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69Table 6:2 Parents� Reasons for Choosing Their Charter School, Rank Ordered by

Mean Scores, 1999-00 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74Table 6:3 Students� Reasons for Choosing Their Charter School, Rank Ordered

by Mean Scores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76Table 7:1 Age Distribution of Charter School Teachers Compared with

National Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81Table 7:2 Percentage of Classroom Teachers in Pennsylvania Charter Schools Compared

with Regular Public Schools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82Table 7:3 Role and Amount of Formal Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84Table 7:4 Role and Highest Academic Degree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84Table 7:5 Highest Academic Degree of Pennsylvania Charter School Teachers as Compared

to Noncharter School Classroom Teachers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84Table 7:6 Years of Experience by Role and in Various Types of School, 1999-00 . . . . . . . . . 85Table 7:7 Charter School Classroom Teachers by Gender, Employment Status, Certification

Type, Race/Ethnicity, Level of Education, and Totals for 1999-2000 . . . . . . . . . . 86Table 7:8 Reasons for Seeking Employment at This School (Rank Ordered According

To Means) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88Table 7:9 Level of Satisfaction with the Mission of the School (N=507) . . . . . . . . . . . . . . . . 89Table 8:1 Teacher Expectations and Current Experience With Regard to Innovative

Practices and Autonomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

xvii

Table 8:2 Levels of Teachers and Staff Satisfaction with Working Conditions . . . . . . . . . . . 96Table 9:1 Prevalence of Commonly Reported Instructional Innovations in 31

Charter Schools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117Table 9:2 Availability of Computers and Technology in Pennsylvania Charter Schools . . . . 118Table 10:1 Proportion of Students with Disabilities Enrolled During the 1999-2000

School Year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134Table 11:1 Number of Charter Schools Participating in the PSSA by Subject/Grade . . . . . . . 152Table 11:2 Aggregate Comparison Between Charter Schools and All

Pennsylvania Schools, 1998/99 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154Table 11:3 Summary of Host District Comparisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155Table 11:4 Two-Year Change Scores on the PSSA, 1997/98 to 1998/99 . . . . . . . . . . . . . . . . 156Table 11:5 Non-PSSA Achievement Test Data Provided By Charter Schools . . . . . . . . . . . . 159Table 11:6 Non-PSSA Achievement Test Data: Grade Equivalents . . . . . . . . . . . . . . . . . . . . 159Table 11:7 Non-PSSA Achievement Test Data: Normal Curve Equivalents . . . . . . . . . . . . . . 160Table 11:8 Non-PSSA Achievement Test Data: Gain in Percent Achieveing At Grade Level 160Table 11:9 Non-PSSA Achievement Test Data: Percent Students Whose Scores Improved

From Pre- to Posttest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160Table 11:10 Income Adjusted PSSA Scores for Charter Schools (Regression Residuals) . . . . 162Table 11:11 Summary of PSSA Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169Table 12:1 Charter School Waiting Lists, Summer 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174Table 12:2 Attendance Rates in Charter and Noncharter Schools Compared,

1998/99-1999/00 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176Table 12:3 Descriptions of Subsales on the School Climate Survey . . . . . . . . . . . . . . . . . . . . 180

F i g u r e s

Figure 3:1 Growth in Number of Charter Schools, 1997/98 - 2000/01 . . . . . . . . . . . . . . . . . . . 18Figure 3:2 Number of New Schools by Year, 1997/98 - 2000/01 . . . . . . . . . . . . . . . . . . . . . . . 19Figure 3:3 Number of Students Enrolled in Charter Schools, 1997/98 - 2000/01 . . . . . . . . . . . 19Figure 3:4 Distribution of Students Across Schools, 2000/01 Projections . . . . . . . . . . . . . . . . 20Figure 3:5 School Size by Age of School . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Figure 3:6 Cumulative Number of Charter Schools: Philadelphia Compared with All PA . . . 22Figure 3:7 Pennsylvania Charter Schools by Year, 1997-98 . . . . . . . . . . . . . . . . . . . . . . . . . . . 23Figure 3:8 Pennsylvania Charter Schools, 1998-99 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24Figure 3:9 Pennsylvania Charter Schools, 1999-2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25Figure 3:10 Pennsylvania Charter Schools, 2000-2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26Figure 3:11 Pennsylvania Charter Schools, 1997-2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27Figure 3:12 Grade Level Distribution of All PA Students, 1998-99 . . . . . . . . . . . . . . . . . . . . . . 28Figure 3:13 Charter School Enrollment by Grade, 1999 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29Figure 5:1 Revenue and Expenditure Variance Compared, 1998/99 . . . . . . . . . . . . . . . . . . . . . 59Figure 6:1 Distribution of Sampled Students by Grade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66Figure 6:2 Distribution of All Pennsylvania Charter School Students by Grade . . . . . . . . . . . . 66

xviii

Figure 6:3 Distribution of Charter School Students by Race/Ethnicity, 1999-00 . . . . . . . . . . . 67Figure 6:4 Average Household Income of Charter School Families . . . . . . . . . . . . . . . . . . . . . 69Figure 6:5 Type of School Attended Before Charter School: Responses from Sampled

Students and Parents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72Figure 7:1 Distribution of Charter School Teachers and Staff by Ethnicity . . . . . . . . . . . . . . . 80Figure 7.2 Distribution of Teachers and Other Staff by Grade, 1999-2000 . . . . . . . . . . . . . . . 82Figure 9:1 Charter/Host District Venn Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106Figure 9.2 Models for the Diffusion of Innovations Between Charter Schools and

Non Charter Schools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121Figure 10:1 The Proportion of Students with Disabilities Enrolled in PA Charter Schools . . . 135Figure 10:2 Individual Attention for Students: Teacher/Staff Initial Expectation vs.

Current Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141Figure 10:3 Individual Attention for Children: Parents� Initial Expectation vs.

Current Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141Figure 11:1 Charter and Host District Change by Percent Low Income Student . . . . . . . . . . . 157Figure 11:2 Predicted and Observed Math 5 Scores, 1998/99 . . . . . . . . . . . . . . . . . . . . . . . . . . 161Figure 11:3 Predicted and Observed Reading 5 Scores, 1998/99 . . . . . . . . . . . . . . . . . . . . . . . 163Figure 11:4 Predicted and Observed Math 8 Scores, 1998/99 . . . . . . . . . . . . . . . . . . . . . . . . . . 163Figure 11:5 Predicted and Observed Reading 8 Scores, 1998/99 . . . . . . . . . . . . . . . . . . . . . . . 164Figure 11:6 Predicted and Observed Math 11 Scores, 1998/99 . . . . . . . . . . . . . . . . . . . . . . . . . 164Figure 11:7 Predicted and Observed Reading 11 Scores, 1998/99 . . . . . . . . . . . . . . . . . . . . . . 165Figure 11:8 Predicted and Observed Writing 6 Scores, 1998/99 . . . . . . . . . . . . . . . . . . . . . . . . 165Figure 11:9 Predicted and Observed Writing 9 Scores, 1998/99 . . . . . . . . . . . . . . . . . . . . . . . . 166Figure 12:1 Performance at Previous and Charter Schools . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172Figure 12.2 Number of Students Returning from Charter to Philadelphia Public

Schools, 1999/2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175Figure 12:3 Would You Recommend This School to a Friend? . . . . . . . . . . . . . . . . . . . . . . . . 175Figure 12:4 Discrepancy Between Students�, Parents�, and Teachers�

Perception of School Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177Figure 12:5 Agreement with �Students are More Interested in Learning� by School of Origin . . 178Figure 12:6 School Climate Survey: Student Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179Figure 12:7 School Climate Survey: Teachers and Staff Results . . . . . . . . . . . . . . . . . . . . . . . 181Figure 12:8 Teachers� and Staff Results on the Nationally-Normed School Climate Survey . . 182

A p p e n d i c e s

Appendix A Evaluations Questions Appendix B Survey Results from Teachers/Staff, Students, and Parents/Guardians Appendix C Financial Tables Appendix D Results on Student Achievement

xix

List of Acronyms

ADM Average Daily Member

AFR Annual Financial Report

CAB Charter Appeals Board

CAOs Chief administrative officers

CISC Central Instructional Support Center

COE Current operating expenditures

CSS Charter School Survey

CSPD Pennsylvania State ComprehensiveSystem of Personnel Development

CTBS Comprehensive Test of Basic Skills

EMOs Education managementorganizations

FTEs Full-Time Equivalents

FAPE Free appropriate public education

IDEA Individuals with DisabilitiesEducation Act

IEPs Individual Education Plans

ISSP Instructional Support System ofPennsylvania

ITBS Iowa Tests of Basic Skills

ITED Iowa Test of EducationalDevelopment

IU Intermediate Unit

KEC Keystone Education Center CharterSchool

LEAs Local Education Agencies

MAT Metropolitan Achievement Test

NCEs Normal Curve Equivalents

NCES The National Center for EducationStatistics

NGO Nongovernment Organizations

OEI Office of Educational Initiatives

PDE Pennsylvania Department ofEducation

PEPS Pennsylvania Educational PolicyStudies

PSSA Pennsylvania System of SchoolAssessment

RFP Request for Proposal

SASS School and Staffing Survey

SAT Stanford Achievement Test

SCS School Climate Survey

WMU Western Michigan University

WRAT Wide Range Achievement Test

xx

AcknowledgmentsThis was a complex study, and it could not have been completed without the cooperation andcommitment of a number of individuals. First and foremost, we want to thank the charter school staffwho made time for us during our site visits, even in times of extreme pressure, and the many parents,students, and teachers who took the time to return surveys and participate in interviews and discussions.We also thank the representatives of charter authorizers and others who provided us with informationand suggestions. Second, we wish to thank the many people at the Pennsylvania Department ofEducation who assisted us during the course of the evaluation. Most of all, we thank Drs. Dale Bakerand Timothy Daniels, each of whom provided support throughout the duration of the study. GreggSpadafore and Betsy Becker greatly assisted us with requests for data and documentation, as well as inhelping with matters regarding our contract. Others at PDE who provided invaluable assistance wereWalter Howard, Lenny Locke, Barbara Nelson, and Dennis Peachey. We also thank the Education,Policy, and Issues Center in Pittsburgh for allowing us to use the extensive historical data files they hadcompiled for another project.

We were fortunate to have a talented evaluation team at our side. Deborah Lehmann researched andwrote most of chapter 10, helped keep us sane by coordinating the site visits and the qualitative analysis,and ensured that relevant data were reported back to the schools. Chia-Lin Hsieh helped with thequantitative data analysis, while Catherine Awsumb Nelson interviewed district personnel. The teamgreatly benefited from a number of student research assistants. In particular, we thank Jamie Helsen,Carla Howe, and Kim Reynolds, each of whom was instrumental in helping us organize and analyze thequalitative data. Deborah Bullock, Dianne deVries, Valaida Fullwood, Brian Landry, Gary Smith, andFlossie Stepeny helped us conduct the field research for the project. We also thank Jacqui Kelleher andSandra Vegari who reviewed and provided comments on some of the chapters, Sally Veeder for readingand editing what must have seemed like endless chapter drafts, William Post for preparing the maps, andChristine Hummel for preparing the report cover.

Finally, we express our appreciation to Daniel Stufflebeam, who founded The Evaluation Center andhas developed it into a national and international presence in research and evaluation. Dr. Stufflebeamprovided guidance and support during the preparation and conduct of the study and provided commentsand advice on the final report.

From this long list of names and contributions, one can clearly see that this was a team effort. We feelconfident that the evaluation has benefited from the expertise and diverse perspectives of the variouscontributors and evaluation team members. While we recognize and express our appreciation for thecontributions made by these many persons, we are mindful that we are responsible for the content of thereport including errata.

Gary Miron and Christopher NelsonOctober 2000

1

Chapter OneIntroduction

This is the final report of the 17-month initial study of Pennsylvania charter schools. The EvaluationCenter has conducted this evaluation pursuant to a contract with the Pennsylvania Department ofEducation (PDE). The initial study is a key research component of the Commonwealth�s overallaccountability plan for charter schools. Section 1728-A of Pennsylvania�s charter school law(hereafter known as Act 22) requires an evaluation of the charter school program after 5 years.While the 5-year report is to be largely summative in nature (providing recommendations on theadvisability of continuing or amending the program), the initial study is designed to be largelyformative and to provide feedback to schools and policymakers regarding changes that can be madeto help these schools function more effectively and achieve their anticipated goals. At the same time,this initial study is an important component of the Commonwealth�s overall accountability plan forcharter schools and it provides a foundation for the 5-year legislatively-mandated evaluation.

Charter schools are intended to provide alternative and diverse educational programs; with the goalof improving academic achievement. The intent behind this new form of public schooling is thatby providing further autonomy to schools, they can pursue innovative teaching practices and createa diversity of school options from which parents can choose.

Charter schools operate under a contractual arrangement with a chartering entity, in the case ofPennsylvania this includes the local school districts. The chartering contract frees schools from mostof the rules and regulations that apply to traditional public school system in exchange for increasedaccountability�ultimately, high student academic achievement. Charter schools can be formed bya variety of individuals or groups, including educators, parents, community members, for-profit andnonprofit organizations, and institutions of higher education. A charter is signed by its foundingmembers and a chartering agency and details what the school expects to accomplish with respect tostudent achievement and other outcomes. Unlike traditional public schools, charter schools may beclosed by their chartering entity if they fail to meet the standards set forth in the charter.

Some states have pushed to open a large number of charter schools within a short period of timewhile providing little oversight or support, but this has not been the case in Pennsylvania. Whilethere is a clear desire to expand the charter school reform in the Commonwealth, the Office ofEducational Initiatives at the Pennsylvania Department of Education has maintained highexpectations for the schools in terms of accountability and at the same time has provided technicalassistance to help the new schools operate effectively and in compliance with applicable regulations.

1 The closure of this school provides an indication that schools are being held accountable to their charterand those that do not operate in compliance with applicable regulations will be closed.

2 We use the term �charter school concept� to denote the generic set of ideas that characterize most orall charter school laws. We distinguish the charter concept from its operationalization in particular state charterlaws, such as Act 22

2

This report also expands upon the first annual report of the initial study, released in Spring 2000.The first year report was largely descriptive in nature and relied on preliminary site visits and theresults of student, parent, and teacher surveys administered in the charter schools during Spring1999. This report builds upon that analysis, but includes additional data from student achievementtests, secondary analysis of relevant data sets provided by PDE, additional site visits, anddocumentary evidence gathered by members of the evaluation team.

This report focuses primarily on the 31 charter schools open during the 1998/99 academic year. Oneschool that opened during the fall of that year closed just before the end of the academic year.1 Inaddition, we have included data on the schools that began operations during the 1999/2000 schoolyear. Due to the limitations in the scope of the study, most data on the 17 charter schools thatopened for the 1999-00 school year came from secondary sources.

1.1 What Are Charter Schools?

Charter schools are public schools that operate under a contractual arrangement with a charteringentity such as a state, local board of education, or an independent chartering authority. Thechartering contract frees schools from most traditional public school system rules and regulationsin exchange for increased accountability�ultimately, high student academic achievement. Charterschools can be formed by a variety of individuals or groups, including educators, parents, communitymembers, for-profit and nonprofit organizations, and institutions of higher education. A charter issigned by its founding members and a chartering agency and details what the school expects toaccomplish with respect to student achievement and other outcomes. Unlike traditional publicschools, a charter school may be closed by its chartering entity if they fail to meet the standards setforth in the charter.

Nationally, the charter school movement began in 1991 with the passage of Minnesota�s charterschool law. California followed suit in 1992 with the passage of its own law. Since then, each yearhas seen the addition of several new charter school laws. At last count, 36 states and the District ofColumbia had enacted charter school laws. Pennsylvania�s Act 22 was passed in 1997, the yearwhen similar legislation was passed in Mississippi, Nevada, and Ohio (RPP, 2000). Six charterschools were approved to operate in 1997, and the total number grew to 31 during the 1998/99school year, 47 during the 1999/2000 school year, and 66 during the 2000/01. Chapter 3 providesa more detailed overview of the growth of the charter school movement in Pennsylvania.

Critics and skeptics have charged that the charter school concept is an �empty vessel.� Whether trueor not, the charge stems from two features of the charter school concept.2 First, the concept has

3

attracted followers from both the political left and the political right�from the ranks of teacher unionsto ardent advocates of privatization. Wells et al. (1999), for instance, interviewed key policy makersin six states and found that while some charter proponents simply seek to reform the public schoolsystem without turning to vouchers, others see charter schools as a stepping stone on the way to afull-blown voucher system. Hence, the charter concept has proved to be quite flexible politically.Second, when compared with other education reform packages, the charter concept is quite agnosticon many core issues, including curriculum, instruction, assessment, and others. Indeed, by design,the concept counsels policy makers to delegate most such decisions to individual schools and theirstakeholders and thus stands in the tradition of �site-based decision making.� The charter concept,in short, gives enhanced autonomy to schools in the belief that doing so will unleash previouslyunrealized potential that already exists in schools, ultimately leading to improvements in studentachievement. Hence, like current reforms from across the public policy spectrum, the charterconcept seeks to replace �one-size-fits-all� solutions with an �opportunity space� that charter schoolscan fill with their own experience and innovation (RPP, 1998).

Charter school autonomy, however, does not come as a blank check. Instead, charter schoolspurchase their autonomy in exchange for greater accountability. Charter proponents, however, havein mind a particular kind of accountability, one they believe is more compatible with schoolautonomy than older versions. According to proponents of the charter concept, traditional schoolpolicies hold schools accountable for inputs and processes (e.g., �number of hours� requirements)in the belief that if schools adhere to these rules they will, as a matter of course, produce desiredstudent outcomes. Such accountability designs assume that central policymakers have enoughknowledge about educational processes to prescribe the right inputs and processes for a given set ofoutcomes. The newer �performance accountability� design, by contrast, turns this relationship onits head. Advocates are typically skeptical of central policy makers� knowledge of and wisdom abouteducation. Thus, instead of prescribing means in the belief that doing so would generate the rightends, performance accountability designs prescribe policy goals in the belief that teachers,administrators and other officials �on the ground� are best able to design effective and efficientmeans toward those goals. In short, performance accountability refocuses accountability andmonitoring from inputs and processes to outputs and outcomes. Accordingly, school officials (andother public administrators) receive more autonomy in deciding how to pursue policy goals andperhaps less autonomy over what those goals ought to be. Thus, where critics see an empty vesselin the charter concept, charter proponents see flexibility, innovation and, ultimately, improvedstudent outcomes. In many respects, the purpose of this evaluation is to determine whether thisautonomy-accountability trade-off leads, as promised, to improved student outcomes.

1.2 Objectives of the Evaluation

If charter school autonomy creates an opportunity space in which the schools operate, then there aretwo key questions:

! How are charter schools using their autonomy?

4

! Are these uses of charter school autonomy leading to the positive student outcomes for whichcharters are held accountable?

The central evaluation question stated in the request for proposal (RFP) for this study is, �Doesincreased flexibility in exchange for increased accountability result in improved pupil results?Moreover, the RFP asked a number of more specific questions about uses of charter schoolautonomy and their impact on various educational outcomes.

! What effect does budget have on student results, nonacademic services, and school facilities?

! Are opportunities offered to charter school teachers, parents, and students to influence classroomand school policy significantly different from those offered at traditional public schools?

! Are the opportunities (i.e., professional growth, salaries, benefits, employee rights) for teachersand other employees significantly different at a charter school than at a traditional public school?

! What is the impact of charter schools as related to district reform efforts?

! Is there evidence that, over the term of the charter, student learning has significantly improved?

! What are promising practices in charter schools that could be included in district systemicreform?

A complete list of evaluation questions is included in Appendix A.

1.3 Structure of the Report

This report is divided into four major parts. The architecture of the report builds on the familiarinput-process-outcome model of policy systems. According to this model, policy outcomes are afunction of inputs (both financial and human) that are transformed into outcomes through processes.Each of the 4 parts of the report addresses an important component of the autonomy-accountabilitytrade-off.

The first part provides important background information on Pennsylvania�s charter school law andon the evaluation itself. Chapter 2 summarizes the methods used to gather and analyze data. Thischapter is quite general, leaving detailed discussions of methods to later chapters. Chapter 3completes the background part by providing a brief snapshot of Pennsylvania charter schools,including patterns of growth in the number of schools and students, trends in school size, patternsof spatial distribution, and the primary target populations served by charter schools.

The second part of the report considers some of the most important inputs to charter schools. First,the process by which charter applications are proposed and ultimately granted or denied is the firstimportant hurdle would-be school operators must face. The charter application and approvalprocess, therefore, determines the range of charter school opportunities available to students, parents,and teachers. It is also the first important step in the accountability process, as local school districtsand other actors to seek to identify important student needs and ensure that charter schools willeffectively produce desirable student outcomes. Thus, chapter 4 includes a discussion of the legal

5

and administrative contexts of charter school start-up (including the role of the Charter AppealsBoard), the goals and resources of the �founding coalitions� that have received charters, and the keydemographic and political characteristics of districts that have chosen to grant charters.

The second part of the report continues with a preliminary analysis of charter school finance.Autonomy requires both discretion (the freedom to develop and implement educational policies atthe school level) and resources. Hence, any discussion of charter school autonomy must include ananalysis of charter schools� revenue sources, general spending priorities, and capacity to budget andplan effectively. The part of the report dealing with inputs concludes with a pair of chaptersexamining the key demographic characteristics of charter school students, parents, and teachers. Ageneration of scholarly research on educational productivity beginning with Coleman (1966)suggests that schools� ability to produce student outcomes is conditioned by students� family andcommunity backgrounds. Thus, these chapters lay the foundation for the chapter (in the fourth part)on student test scores. The chapters on students, parents, and teachers also examine some of theattitudinal characteristics of these actors. Indeed, charter school theorists often argue thatparticipants� commitment to a school�s mission and school leaders� ability to form coherent �teams�are crucial to charter school success.

The third part of the report examines some of the educational processes Pennsylvania charter schoolshave employed to date. Chapter 8 examines teacher professional development and other features ofteacher working conditions in charter schools. Chapter 9 examines innovations in Pennsylvania�scharter schools and explores charter schools� governance practices as well as curriculum,instructional techniques, and assessment methods. Chapter 10 provides a brief look at specialeducation in Pennsylvania charter schools. Each of these chapters seeks to identify innovativepractices that might be adopted by other public schools.

The fourth and final part of the report examines two sets of student outcomes. Chapter 11 providesan in-depth analysis of scores on the Pennsylvania System of School Assessment (PSSA).Unfortunately, the analysis is limited by the small number of charter schools and our inability to get1999/2000 results in time for inclusion in the report. Nonetheless, the chapter provides a usefulearly look at patterns of student achievement in Pennsylvania charter schools. There is, however,legitimate debate about the precise student outcomes for which charter schools should be heldaccountable. The debate is particularly poignant for the significant number of charter schools thatserve at-risk populations. Moreover, some have argued that a more appropriate measure of charterschool success than test scores is �market accountability��the extent to which parents and studentshave �voted with their feet� for charter schools. Thus, chapter 12 supplements the picture of schoolsuccess provided in chapter 11 with an examination of a number of alternative indicators. Theseinclude student and teacher perceptions of school quality, transfers into and out of charter schools,attendance, and various aspects of the schools� educational climates and cultures.

Chapter 13 provides a summary of major findings and then discusses relevant policy issues andhighlights areas that deserve/require further research or evaluation.

3 Readers should note that Section 2.4 in the next chapter addresses a number of limitations, bothmethodological and analytical in nature.

6

1.4 Challenges Confronting the Evaluation3

One of the challenges confronting any evaluation concerns the overall frame of the evaluation. Insome cases, evaluators take the existence of the program for granted and seek to find ways toimprove it. This is most closely associated with formative evaluation. In other cases, evaluatorsseek to assess whether the program should continue at all. Usually, such evaluations assess theextent to which the program realizes some preordained social or policy goal. This is most closelyassociated with summative evaluation. This evaluation of Pennsylvania charter schools combineselements of both formative and summative evaluation. The first three parts of the report addressissues that are more formative in nature, seeking to identify strengths and weaknesses in theimplementation of the charter school law. The last part�on student outcomes�seeks to assess theextent to which the program is achieving a variety of student outcomes. Thus, the latter chaptersprovide a preliminary assessment of the program�s overall desirability, as measured by its ability toachieve its stated goals.

However, there are a number of limitations to our ability to provide summative feedback at thisstage. Ultimately, the decision to revoke, continue, or revise Pennsylvania�s experiment with charterschools must include the following considerations. First, how much gain in student outcomes(achievement and otherwise) is enough to justify the program�s existence? One way to address thisquestion is to estimate the cost of a unit improvement in various outcomes relative to the valuepolicymakers place on those improved outcomes. Unfortunately, we were unable to generate reliablecost effectiveness estimates in this study. This is because there are still relatively few charter schoolsin Pennsylvania�and even fewer on which we could make comparisons for any given studentachievement outcomes (e.g., 8th grade math). Thus, we could not provide a satisfactory answer tothe evaluation question, �What effect does budget have on student results, nonacademic services,and school facilities?� Clearly, good methods exist for estimating cost effectiveness and benefit-costratios for educational programs (see, e.g., Grissmer et al., 2000). At this point, however, the dataare insufficient to support reliable estimates. Fortunately, the appearance of more charter schoolsin Pennsylvania should enable future evaluators to provide better cost estimates. Second,policymakers should consider the opportunity costs of the charter school program. In other words,might the resources expended on charter schools be better spent on other programs designed topursue the same goals? Estimates of opportunity costs are even trickier than estimates of fiscal costs,since they inevitably involve tough choices about program and value tradeoffs.

Another challenge the evaluation team encountered came in assessing the extent to which variouscharter school practices are innovative. The challenges were part philosophical and part practicalin nature. Philosophically, the concept of innovation is highly contested, with little agreement byscholars and others on its definition. We entertain two competing definitions of innovation at thebeginning of chapter 9. Nonetheless, our judgments of innovation are clearly sensitive to choice ofdefinition. More practically, given the scope of the project, we found it difficult to provide

7

systematic assessments of the extent to which any given charter school practice was unique relativeto its host district schools. Thus, we relied mostly on less rigorous comparisons between charterschool practices and those that are �typical� nationwide.

Still another challenge derives from the controversial nature of charter school policy and schoolchoice policies more generally. Indeed, the apparent bipartisan consensus on charter schools masksdeeper disagreements about how charter policies should be designed. At the heart of this controversylie legitimate differences of opinion on important value questions, such as the ultimate goals ofeducation and school policy. As we note in the report, charter school stakeholders hold variouspositions on the relative importance of equity, efficiency, and choice. Thus, it is often difficult todisentangle factual disagreements about the impact of charter schools from value disagreementsabout the ultimate goals of charter schools. Where possible, we have tried to identify how ourfindings might affect the pursuit of these various goals (see Chapters 11 and 12).

A final challenge came in interpreting the empirical findings from the evaluation. As is often thecase with new programs, the findings in this report are mixed. Some aspects of Pennsylvania�scharter school policy appear to be going well, others not so well. Readers should bear in mind,however, that most of the findings in the text of the report represent aggregate generalizations.Hence, to say that some aspect of the charter school experiment is going poorly does not imply thatall charter schools are doing poorly on that dimension. Similarly, to say that some aspect of theprogram is going well does not imply that all charter schools are doing well on that dimension. Toaccount for such variations, we have sought, where possible, to include descriptions of school-to-school variation in the text. In addition, we have included detailed school-level tables on a numberof variables. Generally, where the data in question are considered public, we have provided theschool-level tables. Where the data are considered nonpublic, or particularly unreliable, we have notprovided the school-level tables. We encourage readers to pay close attention to the tables andappendices to gain a full appreciation of the range of charter school experiences in Pennsylvania.

8

Chapter TwoMethodology of the Study

The array of questions addressed in this study required multiple approaches for collecting andverifying information and for capturing the various perceptions that exist. Our aim was to collectenough information to analyze all charter schools individually as well as among and between groupsof schools with similar characteristics. The data collected allowed us to provide feedbackinformation to individual schools to help them make improvements as well as to address theevaluation questions regarding the charter school reform as a whole. Since our mandate was toevaluate the charter school reform, rather than individual schools, this final report focuses ongeneralizations across the charter schools and does not make judgments about individual schools.Decisions regarding the nature and type of data to be collected were made with the PennsylvaniaDepartment of Education to ensure that they are worthwhile and of interest to decision makers andother interested parties.

2.1 Approach and Strategies for Data Collection

We used the following methods for collecting information: ! Surveys of staff, students, and parents (charter schools surveys developed by The Evaluation

Center and nationally-normed school climate surveys)! Reviews of (student) work samples (when available)! Interviews ! Diaries and logs (if available)! Document review! Portfolios (if available) ! Direct observation! Focus group meetings! Analysis of test scores and available demographic and financial data

The general strategies for collecting information are summarized in Table 2:1. Table 2:2 shows amatrix of the evaluation questions and sources of data/information for each question.

We are aware of the fact that charter schools are of considerable public interest and that they arebombarded with requests for information and to serve as subjects for a variety of studies. Inaddition, we know that they have been targeted by the public media for stories related to an array of

Table 2:1 Data Collection Strategies and Information SourcesStrategy/Source

Question-aires

Interview DocumentReview

FocusGroup

TestData

Direct Observation

Work Sample

Portfolio(if available)

Diaries or Logs(if available)

Students X X X X X X

Teachers X X X X X X

School CAOs/Directors

X X X

Parents X X X

Local SchoolDistricts thatCharter

X X X

CommunityMembers

X

PDEPersonnel

X X X

School Records X

Schools X X X X X

Table 2:2 Matrix of Objectives of the Study and Sources of Information

General Objectivesof the Study

Charter schoolsurveys (forteachers/staff,students, and

parents/guardians)

School climatesurvey (for

teachers/staff,students, and

parents/guardians)

Inter-view

Docu-ment

review

Focusgroup

Testdata

(PSSA)

PSSA school

indicators

Directobserva-

ation

Worksample

Portfolio(if avail-

able)

Diaries orlogs (if

available)

Does increased flexibility inexchange for increasedaccountability result in improvedpupil results?

X X X X X X X X

What effect does budget have onstudent results, non-academicservices, and school facilities?

X X X X X X X X X X

Are opportunities offered to charterschool teachers, parents, andstudents to influence classroom andschool policy significantly differentfrom those offered at traditionalpublic schools?

X X X X X

Are opportunities (i.e., profession-algrowth, salaries, benefits,employee rights) for teachers andother employees significantlydifferent at a charter school thanfrom a traditional public school?

X X X X X X X X

What is the impact of charterschools as related to districtreform efforts?

X X X X X X X

Is there evidence that, over theterm of the charter, studentlearning has significantlyimproved?

X X X X X X X X

What are promising practices incharter schools that could beincluded in district systemicreform?

X X X X X X X

11

topics. While this attention may be complimentary and initially well received, it becomes aconsiderable drain on the resources of the relatively small staffs of charter schools. Usually, thosein administrative roles at charter schools are unaccustomed to these requests and do not have thesupport services to maintain a responsive position. These conditions posed potential complicationsfor this study; therefore, we made efforts to use existing data that may be required for other reports.We also focused on only those issues that are important and necessary for this study and selectedrespondents who were considered to be knowledgeable about the issue(s) being addressed and whocould contribute to the quality of the information/data that we collected. The planning andcoordination of the data collection in the schools was much improved over time, and we hope theprocess of obtaining information was viewed as time well spent by the informants and useful bystakeholders.

As noted in the matrices, some of the data we collected are quantitative in nature and some arequalitative. Sources of information for answering the key evaluation questions often included acombination of qualitative and quantitative data/information. Likewise, a variety of sources wereoften used to provide a basis for a response to a question. We often use the term �triangulation� inour data/information-gathering efforts. While this implies three indicators or sources, we oftenemployed more than three sources to provide a more detailed and dependable explanation.

For example, we examined the level of satisfaction with the schools from the vantage point ofstudents, teachers, administrators, and parents. We considered evidence of academic achievementfrom test scores, as well as self-rated performance by students, and parents� opinions about theirchildren. In addition, we asked teachers about other types of achievement that reflected the missionand goals of the school. Input about the role and effectiveness of local school districts that grantedtheir charter was supplied by charter school representatives, appropriate Pennsylvania Departmentof Education personnel, and others identified by the authorizers themselves.

2.2 Specific Methods for Data Collection

This section contains brief descriptions of the data collection methods. Further details on thesemethods are included in the sections that contain the respective results.

Surveys

Four different surveys were used in the course of the study. Charter school surveys developed byThe Evaluation Center were administered to teachers/staff, students, and parents/guardians. Aschool climate survey from the National Association of Secondary School Principals was also used.While the questions in the charter school surveys were targeted to each group (i.e., parents, students,and charter school staff), the same school climate survey was administered to all three informantgroups in the charter schools.

All schools were visited at least two times for the purpose of administering surveys. The first roundof these visits occurred in May and early June 1999. The second round occurred during March and

12

April 2000. The fieldwork went rather smoothly, and most schools were prepared and eager to workwith us. At a few schools, however, the level of cooperation from school representatives wasminimal. During these visits, questionnaires were administered to students, teachers/staff, andparents/guardians. Interviews were also conducted; and documentation, where available, wascollected about the school.

Below a brief description of the questionnaires and targeted informant groups is included as well asinformation about the timing of the administration of the questionnaires and the actual datacollection process. Appendix B contains copies of the surveys used in this study.

Teachers/staff charter school survey. All teachers and school personnel who work more than 5hours per week and who are involved with instruction, including administrative and professionalsupport personnel, were asked to complete this questionnaire. The respondents were asked tocomplete the questionnaire, enclose it in an envelope, and then return it to a designated person at theschool. Teachers were instructed not to place their names on the questionnaire, although they wereasked to check their name off a list so that we could trace and follow up with missing respondents.Since the completed forms were to be collected, sealed, and mailed to the external evaluator by adesignated person at each school, ample assurance was given that the responses would beanonymous. A cover letter explained the purpose of the survey, and each teacher received anenvelope in which to enclose the survey.

Student charter school survey. This questionnaire was used only with students in grades 5-12.This meant that a few schools that catered only to lower elementary grades were not included. Threeclasses of students were selected at each school. These questionnaires were administered by amember of the evaluation team, and all of the students in these classes were asked to complete aquestionnaire. The purpose of the survey and the manner in which the results would be used wereexplained to the students before they began completing the forms. Students in grades 7-12 couldtypically complete the questionnaires on their own, after initial instructions. More instructions forindividual items was provided to students in Grades 5 and 6.

Parent/guardian charter school survey. Depending on the size of the school, between 25 and35 families were selected at each school to complete the survey. Families were randomly selectedfrom a roster of all students by a member of the evaluation team. Additional details regarding thesampling can be found on the evaluation web site (http://www.wmich.edu/evalctr/) in the documententitled �Instructions for Administering the Parent/Guardian Survey.� A cover letter explained thepurpose of the survey, and each parent received a self-addressed, stamped return envelope in whichto enclose the survey. School participation in this component of the study was optional during thefirst round of data collection in May 1999. This was because of the short space of time availablebefore the end of the school year to administer the survey and conduct a thorough follow-up. Duringthe second round of data collection, two dollars were enclosed in each envelope going home toselected parents. This served as a means of expressing our gratitude for the time parents took incompleting and returning the survey.

13

School Climate Survey for teachers/staff, students and parents/guardians. This is a commercialinstrument developed by the National Association of Secondary School Principals. Theadministration of this instrument was coordinated by the external evaluators or by a travelingobserver who worked as part of the evaluation team. One advantage of the School Climate Surveyis that national norms are available so that charter schools can compare how they rate compared withother public schools across the nation.

The summarized results from each survey were returned to each school for its own planningpurposes. Additionally, a short report containing the responses to the open-ended questions werereturned to the schools. When returning the results of the second round of surveys, we also providedthe schools with a primer to help them understand and interpret the results for their school.

Response rates on surveys. The purpose of our sampling was to build an accurate compositepicture of the target population of staff, students, and parents across all charter schools in the state.We pieced together this picture by sampling representative groups of informants at each school. Ourstrategy in sampling teachers/staff was to receive a high response rate from all teachers/staff in thecharter schools. For students, the strategy was to select three representative classes at each school.In many cases this involved sampling 100 percent of all the students at grade 5 or above. In all othercases, the three classes represented a large portion of all enrolled students.

Table 2:3 Sample Size and Response Rates on Surveys

1998/99 1999/2000

TargetPop.

AchievedSample

ResponseRate

TargetPop.

AchievedSample

ResponseRate

Teacher/Staff Charter School Survey 609 447 73.4% 649 537 82.7%

Student Charter School Survey 1021 923 90.4% 1221 1105 90.5%

Parent/Guardian Charter School Survey 577 292 50.6% 777 364 46.8%

Teacher/Staff School Climate Survey � � � 366 316 86.3%

Student School Climate Survey � � � 842 755 89.7%

Parent/Guardian School Climate Survey � � � 25 17 68.0%

Since one of the key purposes of the charter school reform is parental choice, parents are clearly oneof the most important informant groups. Unfortunately, parents are also the most difficult groupfrom which to collect information. Many other studies invest time and effort into sampling allparents but then invest little effort into follow-up. In order to achieve a representative sample, ourstrategy was to sample a smaller group of parents at each school and then work hard to obtain a highresponse rate from this randomly selected group. Either of the two approaches would likely haveyielded a similar number of returned surveys, but from our experience we find that the parents who

14

initially respond are either extremely critical or extremely positive about the school. In other words,a small, well-drawn sample is better than a large, poorly drawn sample, since the former is morelikely to be representative of the target population. Table 2:3 illustrates the overall sample andresponse rate by informant group and year.

Interviews and Site Visits

During the site visits when we administered surveys, as well during other site visits, we conductedinterviews with the CAOs and with other staff members. In some schools we also had the chanceto meet with parents and community members. As on other occasions, the purpose of the visits wasto collect information about innovative or unique aspects of the schools, as well as to inquire aboutevidence of success according to the school mission statement.

Document Review

The annual reports the charter schools prepared and submitted to the Pennsylvania Department ofEducation in August each year were the primary source of documentation regarding the operationand performance of the charter schools. For 6 schools we obtained 3 annual reports and for 24schools we received 2 annual reports. Additionally, we received annual reports from most of theschools that began operation during the 1999-00 school year.

Wishing to be as unobtrusive as possible, we requested documentation already produced by theschools that would likely contain the information we wished to collect regarding each individualcharter school. During site visits, we also asked for descriptive information/evidence about theirschool�s success and its ability to fulfill its mission as well as any innovative or unique aspects ofthe school in terms of curriculum, instructional methods, or governance/administrative/operationalaspects.

Analysis of Data Available from the Pennsylvania Department of Education

From the Pennsylvania Department of Education web site we were able to download data pertainingto charter schools and their host districts. By host district, we are referring to the public schooldistrict in which the charter school resides. We downloaded databases pertaining to head counts,finance, and PSSA test results. We were able to include data for the 1997-98 and 1998-99 academicyears. In some cases, we were also able to include data from the 1999-2000 school year. Since mostof our comparisons were made with host districts, we extracted the records for the charter schoolsand their matching host districts. Next we merged the charter school and host district data into thesame records for each year. Finally, we merged records for each school and year into the samedatabase. The structure of these databases allowed us to conduct longitudinal analyses of the charterschool data relative to the host district.

15

2.3 Data Analyses and Reporting

Quantitative and qualitative data were collected and analyzed according to professionally acceptablestandards of practice. The survey results were scanned by machine in order to enter the quantitativeresponses to closed-item questions. After processing and scanning the surveys, the data weredisaggregated and sorted by school. Descriptive statistics were used to analyze the data (i.e., largelyfrequencies, means, standard deviations). Templates were developed for reporting the results backto each school. After compiling profiles from the surveys, the results were formatted and printed.All the results were shared with the schools and with PDE .

As the surveys were collected and returned to The Evaluation Center, all of the open-endedresponses were typed up and recorded in a separate database with responses linked to school ID, roleof informant, and question number. The written comments from teachers/staff, parents, and studentswere returned to each school. All comments were stripped of identifying information in order toassure the anonymity of the respondents.

As we collected the data, it was organized and integrated into a relational database. Archiving thedata in the database facilitated the simultaneous analysis of district-level, school-level, andindividual-level data. At the heart of the database are three sets of tables. Most central to theevaluation are tables on the 31 charters covered under the evaluation contract. These tables arelinked to school-level summaries generated from the student, staff, and parent surveys, and to datafrom the PDE�s Pennsylvania School Profiles.

In addition to data on the 31 �core� charter schools, the relational database included data from allPennsylvania schools from 1996 to 1999 on dozens of dimensions. Thus, in cases where data areavailable, we were able to compare charter schools to their host districts, to demographically similarschools, and to all Pennsylvania schools.

Data analyses are summarized in tables with appropriate explanatory narratives. Preliminary copiesof formal reports were submitted to the PDE contact for review. The purpose of the optional reviewof the reports (formally or informally) by PDE-designated persons and any other personnel is tocorrect errors and omissions and to ensure readability by stakeholders.

In all cases, the Program Evaluation Standards were followed in the conduct and operation of thisstudy.

The collected data yielded information to help us make judgments about individual charter schools,groups of charter schools, and the charter school initiative as a whole. For example, analysis isconducted with the following comparisons in mind:

! Compare charter schools over time (of course, this will be limited to 2 years for 6 of the schools,although additional comparisons can be made with conversion schools).

16

! Compare each charter school with its host school district or a school with similar backgroundcharacteristics including demographics, size, location, and educational level. Comparisonswould consider test scores, financial data regarding revenues and expenditures, etc.

! Results from the charter school surveys and School Climate Survey have been merged withschool-based data regarding the demographics, size, location, etc. This will allow comparisonof degree of satisfaction and quality of school climate with school characteristics.

In recognition of the various stakeholder groups, decision makers, and interested parties, specialefforts were made to communicate the procedures, findings, conclusions, and recommendations inunderstandable formats. In order to provide evaluative information to the various stakeholders, aweb site (http://www.wmich.edu/evalctr/) was established that contained information about theevaluation as well as copies of the data collection instruments and other fieldwork-relateddocumentation.

All schools received the disaggregated results from the surveys administered at their own school aswell as averages for all charter schools. In addition to this, all comments and responses to the open-ended items on the surveys were typed, organized, and returned to the schools. Care was taken toremove references or comments that would reveal the source of the comments.

2.4 Limitations to the Evaluation

There are a number of limitations to this study that need to be weighed and considered. These aredescribed in the following paragraphs.

Polarization. The most important limitation is perhaps the very polarized nature of the reformand the strong vested interest on the part of many of the informants. Due to this, there is a tendencyfor information to be painted �black� or �white.� For an evaluator, this makes such a study a challenge.

Sampling. Compared to other similar studies, the response rate was extremely good. Appendices A-C contain information on the samples. Nevertheless, the response rates from parentsand guardians were not as high as for the other informant groups. Likewise, response rates on theschool climate survey were not as good as those for the charter school survey. We did extensivefollow-up to raise the response rates but, since the school climate survey was administered in May2000, there was less time before the close of the year to follow up on persons not responding.Schools with a response rate lower than 45 percent were removed from the data set

We opted to have the charter schools assist us in distributing and collecting questionnaires. Whilethis helped us to obtain a higher response rate, it also limited the data, since a number of informantscomplained that their responses would not be confidential. We gave strict instructions not to openany of the sealed envelopes at the school, although on a few occasions this happened by accident.

17

Because of the strong vested interests, there is obviously the possibility of misleading informationbeing provided by those we interviewed. Wherever possible, we tried to double-check information,or when references were made to financial issues or testing results, we attempted to confirm suchinformation using the databases we obtained from PDE.

Timing. The fact that the evaluation was so short in duration is also a limitation, since theimpact of the charter schools can hardly be measured over so short a period of time. We have usedPSSA test scores as one indicator of success, but mostly view this as an indicator of the type ofstudents enrolling in the charter schools. Furthermore, due to the time that is required to prepareand recheck school statistics before they are released, we have often been dependent on interpretingolder data on school characteristics.

Start-up phase of schools. A considerable portion of the schools we visited were in their firstyear of operation. We are well aware of the growing pains of opening a new school and the heavydemands on the personnel who run these schools. It is likely that the new schools require a few yearsin order to implement their plans. A few years of operation is also often required in order to secureor renovate a permanent facility. We recognize that the schools have been in a start up phase andthat any fair summative evaluation would need to wait a few more years. For these reasons theevaluation is largely formative in nature and when we describe outcomes, we qualify them andremind the reader of the specific limitations that apply.

1 This total accounts for the closure of one school, Creative Education Concepts.

18

66

30

47

60

20

40

60

80

1997/98 1998/99 1999/2000 2000/01Year

Num

ber o

f Sch

ools

in O

pera

tion

Figure 3:1 Growth in Number of Charter Schools, 1997/98-2000/01

Chapter ThreeA Snapshot of Pennsylvania�s Charter Schools

One of Act 22�s stated goals is to � . . .provide parents and pupils with expanded choices in the typesof educational opportunities that are available within the public school system� (1702-A). In thischapter we provide a snapshot of current Pennsylvania charter schools with an eye toward helpingthe reader assess the range of public school opportunities available. This will also help set the stagefor analyses presented in subsequent chapters. In particular, we address the following questions:

! How many charter schools are there and where are they?! To what extent are private management companies a prevalent actor in Pennsylvania charter

schools?! What kinds of students do charter schools seek to serve and according to what educational

missions?

3.1 Number of Charter Schools and Students

The simplest indicator ofthe growth of charterschools is the number ofschools. As Figure 3:1shows, there were 66charter schools in operationin Pennsylvania as of Fall2000.1 This is up from 6during the first full year ofthe charter school law.This total accounts for theclosure of 1 school,C r e a t i v e E d u c a t i o nConcepts, which bothopened and closed duringthe 1998/99 academic year.

19

6

19

17

25

0

5

10

15

20

25

30

1997/98 1998/99 1999/2000 2000/01Year

Num

ber o

f New

Sch

ools

Figure 3:2 Number of New Schools by Year, 1997/98 - 2000/01

0

5,000

10,000

15,000

20,000

25,000

1997/98 1998/99 1999/2000 2000/01(projected)Year

Num

ber o

f stu

dent

s

0.0%

0.2%

0.4%

0.6%

0.8%

1.0%

1.2%

Percent of PA enrollment

# Students

% PA enrollment

Figure 3:3 Number of Students Enrolled in Charter Schools, 1997/98-2000-01

While the overallnumber of charterschools has increasedfrom year to year, therate at which newschools appeared hasbeen less constant.Figure 3:2 shows thatthe rate of increase wasgreater between thefirst and second yearsof the charter law,slowest between thesecond and third years,and middling betweenthe third and fourthyears. This growth pattern is quite common across the country, with a large number of charterschools approved in the earlier years followed by a slowing in the creation of new schools. Ifanything, the initial �takeoff� in Pennsylvania is more pronounced than in most other states (RPP,1998).

Not surprisingly, the number of students in charter schools has grown considerably since theenactment of Act 22. PDE records show that 1,179 students were enrolled in Pennsylvania charterschools during 1997/98. That number is projected to rise to 20,533 during the 2000/2001 academicyear. The number of charter school students as a percentage of all Pennsylvania public schoolstudents has risen in similar fashion from only 0.06 percent in 1997/98 to a projected level of 1.1percent during the2000/01 academicyear. Figure 3:3illustrates growth inthe number of charterschoo l s t u d e n t s .Unfortunately, nationalfigures on charterschool enrollment ared a t e d , m a k i n gcomparisons difficult.As of 1998 /99 ,Pennsylvania�s 0.3percent charter schoolenrollment placed itjust below the nationalmedian for that year.

2 These projections correspond exactly with estimates based on the latest year for which we haveenrollment data, 1999/2000. During this year, the median school size was also 256.

3 This estimate is based on enrollments for 1998/99, the most recent year for which school-levelenrollment data are available.

20

0

2

4

6

8

10

12

14

16

18

010

020

030

040

050

060

070

080

090

010

00>10

00

School Size

Num

ber o

f Sch

ools

Figure 3:4 Distribution of Students Across School, Projections for 2000/01

However, comparisonswith other states shouldtake into account thefact that Pennsylvania�scharter law is relativelyrecent. Compared withthe cohort of stateswith laws enacted in1997 (Pennsylvania,Missouri, Nevada, andOhio), Pennsylvaniahad the highest charterschool enrollment as apercentage of all publicschool students as of1998/1999.

Focusing again onPennsylvania, based on projected enrollments for the 2000/01 school year, the median charter schoolwill have 265 students.2 There is wide variation around this value, however. Most charter schoolsare small, with enrollments ranging from 100 to 400 students. Some, however, enroll as many as1,000 students. Figure 3:4 summarizes patterns in school size. In spite of this, the median charterschool is still much smaller than the median public school in Pennsylvania (577).3 Nonetheless,Pennsylvania charter schools are larger than other charter schools in the nation. Indeed, during the1998/99 school year the median charter school in the U.S. enrolled only 137 students (RPP, 2000).

Finally, there appears to be a trend toward larger schools, even though the number of small schoolscontinues to grow apace. Indeed, Figure 3:5 provides box plots showing the range of school sizesas of Spring 2000. The lines in the middle of the boxes show the median school size, while thelower and upper boxes show the first and third quartiles respectively. The �whiskers� extendingabove and below the boxes show schools with more unusual sizes. Generally, the more elongatedthe box and the longer the whiskers the more dispersed the distribution. We have also included ahorizontal line at the median size for all schools in the United States. The three boxes represent(from left to right) schools that opened in 1997/98, 1998/99 and 1999/2000, respectively. The figureshows that even though the median size of older schools and newer schools is approximately thesame (shown by the middle line in the boxes), a greater proportion of the newer schools are larger,

21

Enrollment by First Year of Operation

0

200

400

600

800

1998 1999 2000

Figure 3:5 School Size by Age of SchoolNote: The enrollment figures are from April 27, 2000

as evidenced by the longer whisker on top of the right most box. Indeed, several of the schoolsopening in the 1999/2000 school year are larger than the national median for all (non charter) publicschools (505).

3.2 Spatial Patterns in Pennsylvania Charter Schools

Markets work efficiently when, among other things, there is a wide range of suppliers to satisfyconsumer preferences. This not only ensures that most consumers get what they want, but thatcompetition among the many suppliers keeps quality high and prices low. Act 22 does not, to besure, create a pure and unalloyed market for educational services. Indeed, �prices� are set by districtfunding rates (see chapter 5) with the funds deriving, ultimately, from tax revenues and funneledthrough school districts. Nonetheless, it is clear that in enacting the charter school law the GeneralAssembly sought to �provide parents and pupils with expanded choices . . . within the public schoolsystem� (Act 22, sec. 1702-A). For such choices to become a reality, there must be a sufficientnumber of charter schools offering a wide range of �products.� These schools should also bedistributed geographically such that they are within reasonable distance of many or most families.

In order to assess the spatial distribution of Pennsylvania charter schools, we plotted the locales ofthe schools by their first year of operation. Figure 3:7, for instance, shows the locations of the 6charter schools opened during the first full year of the program. The first year established a patternthat has continued to this day, with most charter schools in Philadelphia and fewer in other regionsof the Commonwealth. Indeed, with the exception of Keystone Education Center in Mercer County,

4 This figure excludes the one charter closed during the period under investigation.

22

413

2532

2

1722

34

0

5

10

15

20

25

30

35

40

1997/98 1998/99 1999/2000 2000/01

Cum

ulat

ive

Num

ber o

f Sch

ools # P hilad. Schools

#Non-P hilad. S chools

Figure 3:6 Cumulative Number of Charter Schools: Philadelphia Compared with All Pennsylvania

each of the first crop of charter schools was either in Philadelphia or the surrounding communities(Chester County Family Academy is located in the West Chester school district, just to the west ofPhiladelphia). The 1999/2000 school year once again evidenced more bias toward the PhiladelphiaSchool District, with 12 new schools in Philadelphia and only 5 elsewhere. Figure 3:6 shows thatthe distribution of charter schools evened out somewhat in the following year, with 9 new charterschools in Philadelphia and 16 in other parts of the state. The 1998/99 cohort of schools included,for the first time, schools in Pittsburgh, the state�s other major urban area. There were also schoolsin Erie, Harrisburg, State College, Lancaster, and other smaller urban areas.

After the approval of charter schools to open during the 1999/2000 school year, a number ofstakeholders (including program officers of Western Pennsylvania foundations) were concerned thatwhile charter schools were flourishing in Philadelphia, the movement was languishing in the rest ofthe state. Indeed, while Philadelphia has approximately 11 percent of the public school students inthe state, it has 51 percent of the charter schools and 69 percent of the charter school students. AsFigure 3:7 shows, the rate of growth in the number of charters outside of Philadelphia did slowduring this period.4 However, the rate of growth in the remainder of the state appears to be catchingup with the approval and opening of the 2000/01 round of charter schools. Seven schools are slatedto open in Philadelphia during 2000/01 and 12 elsewhere. We explore the reasons for the largenumber of Philadelphia charter schools in the next chapter. Figures 3:8 to 3:10 show new schoolsfor 1998/99, 1999/2000, and 2000/01. Figure 3:11 summarizes the spatial growth patterns for all4 years in a single map.

Figure 3:7

Figure 3:8

Figure 3:9

Figure 3:10

Figure 3:11

5 These estimates come from a database maintained by PDE’s Office of Educational Initiatives

6 The only source of enrollments by grade level we could find is the Pennsylvania School Profilesdatabase. Unfortunately, data for the 1999/2000 school year will not be available until after completion of thisreport.

28

0%10%20%30%40%50%60%70%

K 1 2 3 4 5 6 7 8 9 10 11 12Grade Level

% S

choo

ls

% schools 1998 %schools 1999 %schools 2000

Figure 3:12 Planned Grade Level Coverage by First Year of Operation

3.3 Mission and Target Population

In this section we briefly assess the range of “products” offered by Pennsylvania’s charter schools.Chapter 9 provides much more detail on charter schools’ curricula, instruction, pedagogy, andoperations. For the time, however, we focus on charter schools’ target populations, or the studentgroups they seek to serve. For the purposes of this analysis, we operationalize “target population”in terms of (a) grade levels served and (b) school missions. We begin by examining the grade levels targeted by charter schools. We emphasize that this analysisexamines the grades schools intend to serve.5 Figure 3:12 illustrates the proportion of charterschools that provide instruction or intend to provide instruction at the various grade levels. One canclearly see that the charter schools opened in the first year of the reform were much more likely tocater to students at the lower or upper secondary levels than the schools that opened in the followingtwo years. It is important to recall, however, that the trends in this figure illustrate the proportionof schools at each grade level but not the proportion of students. We defer description of charterschools’ actual student populations until Chapter 6.

Figure 3:13 presents charter school enrollments by grade for the most recent year for which data areavailable (1998/1999).6 Figure 3:13 also provides comparable enrollment data for all Pennsylvania

29

0%

3%

6%

9%

12%

15%

K 1 2 3 4 5 6 7 8 9 10 11 12Grade Level

Perc

ent S

tude

nts

% PA Students% Charter Students

Figure 3:13 Grade Level Distribution of All Pennsylvania Students Compared to Charter Schools, 1998-99

schools for the same year. Inspection of the bar graphs reveals that the distribution of enrollmentis quite even across grade levels for all Pennsylvania students. Indeed, the percentage of studentsenrolled ranges only from 7.1 percent (kindergarten) to 8.3 percent (10th grade). The charter schooldistribution, by contrast, is quite lumpy, with percentages ranging from 2.6 percent (11th grade) to13.1 percent (9th grade).

Given that it is generally more expensive to educate older students than younger students, one mightexpect that the distribution of charter school enrollment would be weighted toward lower gradelevels. Thus, it is particularly noteworthy that the highest percentages come in the upper grades.Moreover, the specific pattern of lumpiness in the distribution is notable. The peaks come inkindergarten, 6th grade and 9th grade–the youngest students in the elementary, middle, and highschool grades. This suggests that charter schools might be planning to “grow from the bottom”; thatis, to start with a group of students in lower grades and add new cohorts of students in these lowergrades while retaining the original group.

Most charter schools, moreover, include clusters of grade levels that differ from the traditionalelementary, middle, high school categories. If we define elementary grades as K-6, middle schoolgrades as 7-9, and high school grades as 10-12, then we find that 29 of the 30 schools we examinedinclude nontraditional groupings. Of those 29 schools, 27 include either both elementary and middleor middle and high school grades. The remaining 2 schools cover elementary, middle, and highschool grades.

In short, Pennsylvania charter schools apparently intend to serve a reasonably representative sampleof students by grade level. There is, indeed, some unevenness in the distribution. However, onemight expect that this will even out as schools grow in the coming years. In any case, this issue bearsfurther attention by policymakers as the charter schools mature.

30

A second way to assess charter schools’ target populations is by examining their missions. Someschools’ missions focus on educational philosophies or pedagogical techniques. Others, however,focus on certain populations, such as at-risk students. If a large proportion of charter schools focuson at-risk students, policymakers might conclude that charter schools are not truly competing withother public schools. Indeed, the latter might be happy to allow charter schools to educate thesestudents.

In order to assess charter school missions, we examined formal mission statements as reported inannual reports each school must submit to the Pennsylvania Department of Education’s Office ofEducational Initiatives. Given the scope of the study, we restricted our attention to the 30 charterschools open as of Spring 1999. Readers should bear in mind that missions are simply concisestatements of a school’s aims and should not be interpreted as comprehensive lists of offerings.Indeed, many charter schools offer programs that are not directly reflected in their missionstatements.

Charter schools’ mission statements vary quite widely. In order to simplify exposition, we haveidentified nine core themes and the number of mission statements that make explicit reference tothem. Most mission statements include more than one theme. We have excluded studentachievement from our list, since all mission statements refer to it in one way or another.

� standards driven (4)� college preparatory (1)� science and technology (7)� bilingual and bicultural education (5)� parent and community involvement (4)� early intervention (1)� at-risk students (9)� preparation for work (4)� character education, community service, and citizenship (10) As one would expect, these themes tend to appear in clusters. That is, mission statements thatinclude one theme are more likely to also include a related or similar theme. With some exceptions,schools targeting at-risk students tend to reside in low income communities. Moreover, schoolstargeting at-risk students are more likely than others to also emphasize character, community service,and other values, and to emphasize preparation for work. Finally, schools targeting lower gradelevels are slightly more likely to focus on character issues, while schools focusing on higher gradelevels are slightly more likely to focus on science, technology, and bilingual education.

This brief examination of charter school missions, then, suggests that charter schools provide areasonably wide variety of educational choices. There is, however, evidence that a fairly largeproportion of charter schools target at-risk students, whose resident districts might be happy to seeleave. This, then, raises questions about the extent to which some Pennsylvania charter schools truly

31

compete for the same students. We shall provide a more thorough examination of charter schools’student populations in Chapter 6.

3.4 Education Management Organizations

One of the more controversial emerging issues in charters schools–and public education generally–isthe use of private education management organizations (EMOs). Proponents of contracting out toEMOs argue that competition, the profit motive, and freedom from what they view as cumbersomegovernmental bureaucracies, allows private management companies to provide more value for themoney. Critics argue, by contrast, that such companies threaten to diminish the local and democraticcharacter of charter schools. Recent years have seen the growth of a number of such companies,ranging from national, full service organizations such as Edison, Mosaica, Nobel, and Advantage,to any number of small “mom and pop” outfits. In Michigan, for instance, nearly 72 percent of allcharter schools contract out all or part of their services to private for-profit management companies(Horn & Miron, 2000).

In contrast to Michigan and other states, private management companies have played a relativelysmall role in Pennsylvania charter schools thus far. Of the 30 schools we examined in detail, onlytwo had substantial relationships with private management companies. Mosaica Academy CharterSchool in Bensalem (just north of Philadelphia) is, as the name suggests, run by the MosaicaEducation Inc. Chester Community Charter School in the Chester Upland School District (formerlyknown as Archway Charter School) has a contract with a private company that operates that school.

There is evidence that the role of private management companies is on the increase, however. Asof the 1999/2000 school year 5 charter schools were managed by 4 different management companiesand enrolled a total of approximately 1,800 students. Projections for the 2000/01 school year suggestthat a total of 8 schools will be run by 5 management companies. Aside from the small companyoperating the Chester Community Charter School, the remainder will be operated by large nationalEMOs, including 4 schools run by Mosaica, 2 schools run by Edison Schools Inc. (one of these beingthe state’s first public school conversion), 1 school operated by Nobel Learning Communities Inc.,and one school run by Advantage Schools Inc. Therefore, EMOs will be operated around 12 percentof the charter schools in Pennsylvania. Enrollment projections for the 2000/01 academic yearsuggest that all told, schools operated by EMOs will enroll over 4,000 students. Using theenrollment projections in Section 3.1, this would amount to over 19 percent of all Pennsylvaniacharter school students.

3.5 Summary

This chapter had two goals. First, we sought to provide an overview of the state of the charter schoolmovement in Pennsylvania. Second, the chapter sought to set the stage for issues and analyses thatcome up in subsequent chapters.

32

As of the 2000/01 academic year, there will be 66 charter schools in operation in Pennsylvania. A67th school was closed after its first year of operation. Taken as a whole, these schools will enrollover 20,000 students, or just over 1 percent of all public school students in Pennsylvania.Throughout the life of the charter school law, there has been relatively steady growth both in thenumber of charter schools and charter school students. Most of these schools enroll far fewerstudents than the typical Pennsylvania public school but more than the typical charter schoolnationwide. There is, moreover, evidence of a trend toward larger schools.

Given the size of Pennsylvania and its student population, it would be unreasonable to expect thatcharter schools would provide meaningful choices to all students after just a few years of operation.Nonetheless, it is clear that charter schools are focused in certain parts of the Commonwealth,particular Philadelphia. Indeed, while Philadelphia enrolls approximately 11 percent of the publicschool students in the state, it has 51 percent of the charter schools and 69 percent of the charterschool students.

Charter schools appear to target students of a reasonably wide variety of grade levels. There is,however, a tendency for charter schools to seek to serve students in the lower reaches of theelementary, middle, and high school grades. Many of these schools are planning to “grow from thebottom,” which might well even out the distribution of grade levels covered. Similarly, charterschools’ mission statements indicate that they intend to serve a wide variety of educational interestsand goals. However, there is evidence that a significant proportion seek to serve at-risk students.From one perspective, this is a good thing, since these are the students who have been least servedby other public schools. However, if one of the goals of the charter school law is to fostercompetition, this focus may mean that many charters are, in fact, targeting a particular strata ofstudents and thus not providing genuine competition for other public schools.

Finally, we note that private management companies have played a relatively small role inPennsylvania charter schools up to this point. There is evidence, however, that this is changing.Based on projections, over 19 percent of all Pennsylvania charter school students will attend schoolsrun by management companies in the 1999/2000 academic year.

In the next chapter, we seek to provide an account of the process by which groups and individualsobtain charters and open schools. Along the way, we will attempt to provide explanations for somefor the patterns mentioned in this chapter, including the concentration of charter schools inPhiladelphia and, more generally, why some districts appear to provide more fertile ground forcharters than others. In subsequent chapters we provide more detailed descriptions of charter schoolstudents, parents, and teachers, as well as charter school curricula, instruction, assessment, andgovernance.

1 In the language of causal inference, data on successes can at best identify a set of sufficient conditions ofsuccess (i.e., factors that can bring about an application�s success). Such data cannot, however, help us to assesswhether these factors are also necessary conditions of success (i.e., applications that do not have these factors fail).Demonstrating causality requires evaluators to show that a given factor (or set of factors) is both necessary andsufficient to the production of the outcome in question. Thus, we cannot make causal inferences based on these data.

33

Chapter FourCharter School Start-Up: Resources and Challenges

One of the assumptions of charter school laws is that there will be individuals and groups with theresources, will, and expertise to set up and successfully run charter schools. Without theseindividuals, charter school choice would be an empty abstraction. This chapter provides descriptionsof the types of individuals and groups (which we shall call �founding coalitions�) that havesuccessfully started charter schools in Pennsylvania, including their goals and their organizational,fiscal, and political resources. The first section describes the legal and administrative context ofcharter school start-up. In the second section we examine the Charter Appeals Board, to whichcharter applicants may appeal charter denials. The third section discusses the characteristics offounding coalitions. The chapter�s fourth section considers the characteristics of charter schools�host districts in order to determine whether certain types of districts provide more fertile ground forcharters than others. The fifth section takes a brief digression to consider some of the reasonsbehind the concentration of charter schools in Philadelphia.

The scope of the project prevented us from gathering data on all charter applicants. Thus, most ofour discussion draws upon the characteristics of successful founding coalitions. Unfortunately, suchdata cannot help us determine whether these characteristics were important causes or driving factorsin whether a given charter proposal was successful or unsuccessful.1 This issue deserves furtherexamination, since application success determines the range of choices available to parents andstudents and because the application and approval process is the first accountability hurdle charterschools must face. Thus, it is the first point at which public authorities can seek to ensure thatcharter schools use their autonomy in ways consistent with the public interest.

4.1 Legal and Administrative Context

In order to assess the legal requirements and restrictions on charter school start-up, we must addressthree questions: (1) Who may apply for charters? (2) What resources are provided for founding

2 The median estimate is based on 55 grants for which we know the amounts. Data were gathered fromlists of planning grant recipients provided by the Office of Educational Initiatives.

3 Personal correspondence with Gregg Spadafore, Pennsylvania Department of Education, August 29, 2000.

34

coalitions? and (3) Who may grant charters and according to what criteria? Each question isaddressed in the paragraphs that follow.

Act 22 is relatively permissive on the question of who may apply for charters and on founders�ability to build upon preexisting schools. First, the law allows virtually any individual or group toapply for charters, except sectarian and for-profit organizations. Among the types of organizationsand individuals specifically mentioned in the statute are teachers, parents and guardians, nonsectariancolleges and universities, not-for-profit corporations, associations, or any combination of theaforementioned categories. Second, Act 22 places few restrictions on founders� ability to build uponpreexisting schools. In addition to new start-ups, the law allows founders to convert both public andprivate schools to charter schools. This distinguishes it from charter laws like Georgia�s, which onlyallows for public conversions. Applications for public conversions, however, must present a petitionwith the signatures of 50 percent of the parents and 50 percent of teaching staff. Finally, unlikemany state charter school laws, Act 22 places no caps on the number of charter schools.

Starting a new school is a very resource intensive activity. Before opening a charter school, itsfounders must spend a considerable amount of time planning and drafting the charter application.This requires the time of individuals with skill and experience in education, finance, andorganizational design. Founders must also begin to find physical facilities. Below we include a listof the topics and issues that charter applications must address:

! Identification of charter applicant! Name of school! Grade or age served! Proposed governance structure! Mission and education goals, including

curriculum and assessment methods! Admission policy ! Criteria for student evaluation! Suspension and expulsion policy! Involvement of community groups! Financial and audit plans

! Complaint procedures! Description and address of physical facility! Proposed school calendar! Proposed faculty and professional

development plan! Plans for student participation in school

district extracurricular activities! Report of criminal history records for

employees! Official child abuse clearance statements! Plan for liability and insurance coverage

Relative to many other states, the Pennsylvania Department of Education has been very successfulin obtaining federal and state start-up funds and distributing these funds more quickly to schools.This includes a pool of funds for planning grants to facilitate the development of charterapplications. According to official records, the Pennsylvania Department of Education�s Office ofEducational Initiatives has awarded 249 planning grants over the past 5 years. The median grantamount is $23,150.2 Some 66 percent of those applying have received planning grants;3 28 percent

4 Office of Educational Initiative lists of planning grants (see note #3). 5 Personal correspondence with Ron Cowell, former chair of the Pennsylvania House of Representatives

Committee on Education, May 11, 2000. 6 In re Collegium Charter School, CAB 1999-9; in re Hills Academy Charter School, CAB 1999-12; in re

Phoenix, CAB 1999-10.

35

of those receiving grants eventually received charters.4 In addition, all approved charter schoolsautomatically receive a start-up grant of $825 per student, which includes both federal and statemonies. We shall have more to say about charter schools� fiscal resources in the next chapter.

Compared with other charter school laws, Act 22 is fairly restrictive in its requirements forchartering agencies. According to Act 22, only districts (LEAs) may sponsor charter schools. Thisdistinguishes Act 22 from �stronger� or �more permissive� charter laws�such as those in Minnesota,Michigan, North Carolina, and elsewhere�that allow universities, state agencies, and other publicbodies to grant charters and oversee the schools. Charter applicants, however, may apply for�regional charters.� Regional charters are granted and overseen by more than one LEA. The ideabehind regional charters was to encourage institutions that serve more than one schooldistrict�museums, universities, etc.� to propose and operate charter schools.5 The Charter AppealsBoard (CAB), however, has pointed out that Act 22 permits charter schools to draw students frommore than one district, whether they hold regional or single district charters.6 In so ruling, the CABmay have taken away much of the incentive for founding coalitions to apply for regional charters.Indeed, of the 70 charters approved as of Summer 2000, only 5 (7 percent) have been regionalcharters. Of these, the number of sponsoring districts has ranged from 2 to 13. In most of thesecases, it appears that the sponsoring districts were involved in the early planning stages of the charterapplication and expressed early support. In at least one case, a sponsoring district was among thegroup sponsoring the charter application. Table 4:1 lists the regional charter schools and the numberof sponsoring districts.

Table 4:1 Regional Charter Schools, Summer 2000

Name of School Number of Districts Year Opened

Northeast Charter School 13 1998-99Lehigh Valley Charter School 4 1999-00Centre Learning Community 3 1998-99

SUSQ-CYBER Charter School 3 1998-99Keystone Education Center 2 1997-98

Once they receive charter applications, Act 22 directs LEAs to judge them in light of four criteria.First, applications must demonstrate sustainable support for the charter school plan. The issue ofhow precisely to gauge community support has come up in a number of cases heard by the CharterAppeals Board. In these cases, the Board has made it clear that applications must demonstrate

7 In re Shenango Valley Regional Charter School, CAB 1999-11.8 In re Souderton Charter School Collaborative, CAB 1999-2; in re Ronald H. Brown Charter School,

CAB 1999-1; Shenango; in re Hills Academy Charter School, CAB 1999-12; in re Phoenix Academy CharterSchool, CAB 1999-10; in re William Bradford Academy Charter School, CAB 1999-8.

9 In re William Bradford Academy Charter School, CAB 1999-810 In re Vitalistic Therapeutic Center Charter School, CAB 1999-6.11 Personal correspondence with Ron Cowell, former chair of the Pennsylvania House of Representatives

Committee on Education. May 11, 2000.

36

support for the particular school in question, not just the charter concept.7 The remainder of thecases, however, make it clear that the Board prefers a fairly liberal interpretation of the term�support.� Indeed, the Board has held that applications must only show that there is support for thecharter school and that evidence of opposition to the school is irrelevant. Moreover, the CAB hasheld that lack of support from any particular stakeholder group (e.g., district teachers) is notnecessarily fatal to the application.8 The Board has also held that the requirement that applicationsprovide evidence of community support does not imply that charters must show why stakeholdergroups support the charter. Moreover, it has held that support for a district in no way diminishes acharter application.9

The second requirement is that applications must demonstrate that the charter is capable of providinga �comprehensive learning experience� to its students. Third, charter applications must provideinformation on a number of issues listed in the statute, including governance structure, admissionspolicies, discipline policies, and many others (sec. 1719-A). Finally, charter applications mustdemonstrate that the proposed charter school has the potential to serve as a model for other schools.However, the Charter Appeals Board recently ruled that the economic feasibility of a charterinnovation for a district should have no bearing on whether the charter is approved.10

4.2 The Charter Appeals Board

The Charter Appeals Board (CAB) plays a critical role in the charter application process. It isbeyond the scope of this report to provide an extensive evaluation of the CAB and its processes. Wecan, however, provide an overview of its processes and some of the policy issues raised by its rolein the start-up process.

The CAB was in many ways the result of a political compromise between those who wished to giveLEAs sole authority to approve, oversee, and renew charters and those who wished to spread suchauthority among a number of actors.11 As part of the compromise, the CAB was not formallyconstituted until July 1, 1999�some two years after the effective date of Act 22. The CAB iscomposed of seven members, including the Secretary of Education and six others appointed by the

12 Descriptive information on the CAB and its procedures is taken from PDE�s web site:http://www.pde.psu.edu/charter.html.

13 As of Summer 2000, only one charter had been revoked, that of Creative Education Concepts. Thecharter school appealed the revocation to the CAB, which upheld the district�s position in a unanimous vote

37

governor and confirmed by the Senate.12 The six include one individual from each of the followingcategories:

! parent of a school-aged child! local school board member! public school teacher! faculty member or administrator in higher education! businessperson! member of the State Board of Education

The CAB�s primary function is to hear the appeals of charter schools and their founding coalitionswhose applications for charters have been denied by LEAs. These can include denials of originalapplications, denials of applications resubmitted after an initial denial, and revocation of a charter.13

The CAB employs a quasi-judicial process in considering appeals, complete with hearings, records,and counsel for both sides. In order to certify an appeal, applicants must gather signatures from 2percent of the adults in the community or 1,000 adults, whichever is less. The petition for appealand the signatures are then presented to the local Court of Common Pleas for a hearing on the�sufficiency� of the petition. Provided the petition is in order, the Court of Common Pleas forwardsit to the CAB. Having received a certified appeal, the CAB provides written notice of acceptance,assigns a docket number, and requests that the district provide a certified record of the charter denialproceedings. The official record of the appeal includes the following:

! charter application! supplemental materials submitted by the charter applicants! transcripts of testimony taken by the district! exhibits offered in conjunction with testimony before the district! any other documents the district relied upon in making its decision! the district�s written decision to deny the charter

With the record in hand, the CAB must assign a hearing officer to the case and meet within 30 daysof receiving the materials. This officer holds prehearing conferences with counsel for both thedistrict and the charter school. After a formal hearing the CAB renders a written decision. If itupholds the denial or revocation of a charter, it simply notifies both parties of its decision. If theCAB overturns district denial or revocation, the LEA must grant (or reinstate) the charter within 10days of receiving the notice. If the LEA fails to grant the charter within this period, the charter issigned by the Chair of the CAB.

14 Unfortunately, we could not determine how many denials were not appealed or whether those that wereappealed are distinctive in some way. For instance, it is possible that only the strongest applications are appealed.Hence, we cannot say with any confidence that any given appeal has a 44 percent chance of success. Indeed, thecharter school success rate might be lower if more cases were appealed.

15 Duquesne Charter Schools Project, �Pennsylvania�s Charter Appeal Board: A First Look at its Impacton Charter Approval.� April 2000.

16 �The Charter Schools Appeals Board.� A panel discussion on the Charter Appeals Board at thePennsylvania State Charter School Conference. State College, PA. April 2000. The discussion included bycriticisms and defenses of the CAB. The panelists were Amos Goodall, Esq., Scott Etter, Esq., and Connie H. Davis.

38

As of July 31, 2000, the CAB had issued decisions on 23 appeals. Of those decisions, 10 overturnedthe district�s denial of a charter. Thus, charter schools have prevailed in 43 percent of the casesappealed to the CAB. In all but one of those cases, the LEA failed to grant the charter within theprescribed 10 days, requiring the chair of the CAB to sign the charter.14

Act 22 provides judicial review of CAB decisions by the Commonwealth Court. As of April 2000,five CAB decisions had been so appealed. Four of those appeals came from a district whose initialdenial was overturned by the CAB.15 The remaining appeal came from a charter school appealingthe CAB�s rejection of an earlier appeal.

Along with providing a �second chance� for founding coalitions frustrated by LEAs� denials orinaction, through its written decisions the CAB also provides interpretations of Act 22. We havereferred to these precedents when expounding relevant sections of the Act.

As the CAB concludes its first year of operation, some stakeholders have raised concerns about theCAB�s processes and its role in charter approval. While it is well beyond the scope of this reportto provide a complete evaluation of the CAB, we note three such concerns. It is important to notethat this evaluation makes no claims about the veracity of these claims. They are merely offered asissues for further consideration.

First, some stakeholders have questioned whether CAB members have the time and resources tofully consider the appeals, especially in light of the fact that most or all members have other full-timejobs, and given the extensive written records associated with some of the appeals. One appealrecord, for instance, was approximately 12 inches thick. Defenders of the process counter that boardmembers have a staff that can summarize documents and otherwise ease the burden.16 Second, somestakeholders worry that the burdens of the appeals process favors districts, which generally havemore legal and financial resources to draw upon than charter schools. As we have seen, charterschools have won just about as often as they have lost CAB decisions (Duquesne Charter SchoolProject, 2000). With no way of observing the denials that were not appealed, however, it is possiblethat a great number of denied founding coalitions balked at the costs of appeal and decided ex antenot to pursue an appeal. Thus, looking at charter schools� �batting average� before the CAB withoutunderstanding the factors that determine how and whether they �come to the plate� in the first placemight provide a wrong estimate of the Board�s proclivities. Moreover, some stakeholders note thefact that all but one of the appeals of CAB decisions have come from districts as further evidence

39

that charter schools are less able to bear the procedural burdens. Finally, some stakeholders pointout that because many decisions are rendered close to the beginning of the school year this createsan intolerable amount of uncertainty for charter school personnel (Duquesne Charter School Project,2000). It is important to note, however, that the timing of CAB decisions depends in part on factorsoutside the CAB�s control, including when LEAs deny charters and when founding coalitionspetition the CAB.

A final issue, and one that reaches beyond CAB processes, concerns how charter schools approvedon appeal will fare in what is likely to be a hostile LEA environment. Because much of this reportfocuses on the charter schools open as of the 1998/99 school year, we cannot provide an empiricallygrounded answer to this important question. We hope that subsequent evaluations and scholarlyresearch will explore this matter. Nevertheless, it is difficult to imagine that charters approved onappeal will have smooth relationships with their LEAs.

4.3 Founders� Goals

The previous section described the legal and administrative context of charter development andapproval. Statutory provisions and administrative structures, however, tell us only what is legallyand institutionally permissible. In order to ascertain what is practically possible, we must lookfurther to the resources, constraints, and opportunities that would-be founding coalitions face inseeking to found charter schools. We begin by examining some of the personal characteristics ofthose who have successfully founded charter schools in Pennsylvania. Given restrictions in thescope of the study, this examination is limited to schools in operation during the 1998/99 schoolyear.

Charter school founders are examples of what is known in the policy literature as �policyentrepreneurs� (see, e.g., Scheider and Teske, 1992, Mintrom, 2000). Policy entrepreneurs are thosewho look for and seize upon opportunities to bring about new policies, institutional structures, andorganizational forms. Whether a given individual or organization is a policy entrepreneur, therefore,depends both on personal characteristics (vision, ambition, charisma) and on the types ofopportunities afforded by their environments. While it is difficult to observe vision and ambitionin large groups of individuals, we can offer observations on the identities and goals of successfulcharter founders.

Interviews and document analysis suggest that charter school founders in Pennsylvania had fairlyclear goals when they set out to development their applications. Among these were

! Provide choice for poor children.

! Provide a venue for the realization of good ideas in a district that were reportedly hindered bybureaucratic encumbrances.

! Promote change in the host district and surrounding districts.

! Inculcate a particular cultural or ethnic perspective.

40

! In the case of conversion schools, continue and extend previous services. In some cases, thisinvolves extending preschool services into lower elementary levels. In other cases, it involvesproviding services at the same grade levels but to a broader population of students. In a numberof cases, the desire to procure a more reliable revenue stream was an important motive forconverting a private school into a charter school.

4.4 Founders� Organizational Resources

Actualizing these goals, however, can often be a very costly and time-consuming undertaking. First,and perhaps most obviously, developing a charter and shepherding it through the approval processtakes money. Fiscal resources might be necessary to buy materials or to compensate people who taketime off from jobs and other remunerative activities in order to develop the charter. Second,founders are more likely to succeed if they can draw upon individuals with certain technical skillsrelated to education. Other things being equal, charter applications are more likely to persuadeauthorizing bodies if the founders appear to know what they are talking about. Many foundingcoalitions have included current and former teachers and administrators who bring such expertiseto the table. In addition, applicants must assure authorizing bodies that they know how to managepersonnel, plan, budget, and account for expenditures. Third, charter applicants must have a stockof what social scientists call �social capital.� Social capital is what allows individuals, each withtheir own preferences and plans, to come together to work toward shared goals. Social capital isoften developed through longstanding relationships among individuals. The importance of suchresources is underscored by the fact that, under the terms of Act 22, charter applicants must showthat they have the capacity to provide a comprehensive educational experience for their students andthat they could provide a model for emulation by other schools. Preexisting organizations can provide all three types of resources. Not surprisingly, therefore, thefirst and most notable common characteristic of successful charter school founding coalitions is thatmost included personnel from either preexisting schools, community development organizations,ethnic and racial groups, or other nonprofit organizations. Such organizational resources haveincluded:

! Preexisting schools converted to charter schools, which often provide personnel andinstitutional wisdom from years of running schools

! Preexisting social service organizations provide both administrative expertise and access tofunding sources (both within the organization and through longstanding relationships withexternal funders).

! Relationships with universities often bring the technical expertise of education school faculty.

! Ties to business leaders often bring access to funders and other influential members of thecommunity.

41

Thus, while from a legal point view many of Pennsylvania�s charter schools are new start-ups, eventhese schools often draw upon rather impressive organizational and social networks.

Nonprofit community-based organizations are extensively involved in the founding and start-up ofcharter schools in Pennsylvania. This is a rather unique feature of the Pennsylvania initiative thatmakes it different from other states. Partnerships with nonprofit community organizations have longbeen considered a potential source of support for public schools, particularly in urban areas. Thereare likely a number of reasons to explain why community organizations are involved in only alimited way in our public schools. Charter schools in Pennsylvania, however, have been effectivein working with these organizations. Likewise, community organizations have come to see charterschools as a means of promoting their own goals by supplementing the services they already provide.In the Chapter 10 on school organization and governance, we will examine in greater detail the roleand contribution of nonprofit and for-profit organizations in the charter school initiative.

Other charter schools maintain relationships with for-profit companies. During the 1998-99 schoolyear, 2 of the 31 school in operations were involved with for-profit educational managementorganizations (EMOs). In the 1999-00 school year, 5 of the 47 schools were operated a total of 4different EMOs, which altogether enrolled approximately 1,800 students. In most of these instances,the EMO not only manages the charter school but also helped write the charter application andproposal. For the 2000-01 school year, there will be 5 EMOs operating 8 schools with more than4,000 students enrolled in them. This represents a big jump, but compared with other states, theinvolvement of EMOs is still quite limited. The low profile of EMOs in Pennsylvania�s charterschool reform sets it apart from a number of other states. In Michigan, for instance, some 72 percentof charter schools are operated by EMOs (Horn & Miron, 2000).

We close this section by offering a simple typology that summarizes the role of organizations in thecharter school start-up process (see Table 4:2). On the horizontal access we distinguish schools thathave clear connections to community and other nongovernmental organizations from those that donot. As we can see from the table, most Pennsylvania charter schools (78 percent) have a clearrelationship with a nongovernmental organization. The vertical axis distinguishes schools that weredeveloped from preexisting schools from those that were started from scratch. Here there is a moreeven split, with 46 percent of charter schools built on the foundations of preexisting schools and 53percent with no such foundation.

Table 4:2 Organizational Bases of Pennsylvania Charter Schools

Nongovernmental Organization Base No Organizational Base

Based on preexisting school 39% 7%

Start-up school 39% 14%Source: Analysis by The Evaluation Center

42

4.5 Types of Individuals Involved in Charter Founding Coalitions

Some individuals found among the ranks of charter school founders are former public schoolteachers and administrators. In most cases these individuals are in some way dissatisfied with publicschools. Interviews with charter school chief administrative officers (CAOs) suggest that manymembers of founding coalitions are quite pessimistic about the current system of public education.For at least one founding coalition, the public school system was so bad that just about anyalternative, including but not limited to charter schools, �had to be better.� Other founders, bycontrast, were less strident in their criticisms of public education. Indeed, a founder of anothercharter school remarked that his school�s founding coalition believed that public schools and theirstaffs are full of good ideas, but that talented individuals have too few opportunities to innovate inmost public schools. Indeed, many charter school founders report that they maintain friendships andprofessional relationships with officials in their former schools. A few (though not many) reportexchanging ideas with district personnel.

A second group of individuals found among charter school founding coalitions is universityacademics. Such individuals often view charter schools as a venue in which to try out new andinnovative ideas about education. The level of such individuals� involvement in charter schoolsranges from central to merely advisory. One charter school, for instance, was founded by auniversity education professor and graduate students. In other cases, committees formed to draftcharter documents have included university faculty. In still other cases, academicians have advisedfounding coalitions without formally participating in the drafting of charter documents.

Members of the business community make up a third group represented in founding coalitions. Inat least a few cases, leaders from local businesses have either helped create the initial vision for acharter school or served on committees created to draft charter applications. Often, businessinvolvement comes through participation on boards that oversee nonprofit organizations.

In a small number of cases, charter proposals were tendered by groups formed specifically for charterdevelopment. In one case, the group that developed and submitted the charter application had cometogether initially to sue the local district for providing an inferior education to many students.

A group conspicuously absent from many founding coalitions is parents. There is evidence thatparents become involved in the later stages of charter school development and operation (see chapter9). However, our interviews produced little mention of parents as direct and driving forces in charterdevelopment. This does not rule out the possibility that parents indirectly influence charterdevelopment. Indeed, one might argue that the need to attract students (and their parents) in orderto obtain funding means that founding coalitions must anticipate parents� preferences when decidingwhether to write a charter application and in designing the charter document. Indeed, some CAOsindicated as much during interviews. Nonetheless, we found little evidence that parents are takingthe lead in instigating the charter development process in Pennsylvania.

43

4.6 Political Constraints and Opportunities

The charter approval process is far from a narrowly technical one. The authority to approve or denycharters is given to school districts through local school boards. As democratically elected bodies,school boards respond at least in part to their constituencies. Consequently, the decision to approveor deny a charter is at least partially a political process. Indeed, one of the criteria on which LEAsmust evaluate charter applications is the level of sustainable community support. Thus, we must addlocal political climate to the list of factors that influence the founding of charter schools. As policyentrepreneurs, successful founding coalitions must not only provide resources, goals, and vision, theymust also be able to read and exploit opportunities created by the constellation of local politicalforces.

Generally, school boards have considerable incentives to deny charters. First, because students wholeave district schools to attend charters take with them most of their per pupil financial allotment,districts stand to lose considerable sums of money. In the extreme, this could force districts to layoff teachers and staff, cut back on programs, or even close schools. Second, and less tangibly,approving a charter might be seen as an admission by the district that it has failed to provide a highquality education for all its students. With this might come a diminution of public prestige and adecline in district staff morale.

In spite of these considerable incentives, Pennsylvania school districts had, as of Summer 2000,approved 70 charters. While the scope of the study prevented us from conducting a completepolitical analysis of charter approval, we asked charter school CAOs to identify the reasons theythought districts supported their charter schools. In many instances charter school CAOs reportedthat denying the charter would have subjected the local school board to a political backlash. In thewords of one CAO, the district approved the charter �because they felt they had to.� A fewinformants indicated that the local boards were under the impression that, under the law, they didn�thave a choice to deny charter applications. When the appeals board started reviewing cases,however, districts obviously became aware that they had a choice of denying the charter application.

Political support for charter applications often comes from two sources. First, charter schools oftentap into general skepticism about �big government� and educational �bureaucracies.� Indeed, manycharter applicants are no doubt helped by the public perception that they are dynamic andentrepreneurial Davids fighting entrenched and ossified district Goliaths. Second, foundingcoalitions can often draw upon their ties with powerful community leaders. Ties with communityorganizations often align charter applicants with highly visible community figures, includingbusiness leaders, foundation program officers, prominent academics and, in some cases, mayors andother elected officials.

Not all charter applicants assume the role of David fighting Goliath, however. One CAO, forinstance, said that his founding coalition was successful in part because it was �upbeat about publiceducation.� By emphasizing that public schools already have a great deal of talent and potential forinnovation, this coalition sought to frame its efforts as a form of cooperation with the district.

17 The analysis in this section reports bivariate relationships�relationships between any one of a numberof district attributes and whether the district sponsors a charter school. Such analyses, however, can often bedeceiving. Apparent correlations between two variables can be �spurious.� That is, they can falsely assign to onevariable the influence of another. For instance, it is well known that the relationship between expenditure levelsand test scores often �disappears� once we �control for� family income. In order to account for this possibility,we confirmed the bivariate analyses with multivariate analysis (Poisson regression), which allows for such complexcorrelations. With the exception of the per-pupil expenditures variable, all of the reported bivariate relationshipsstood up in the multivariate analysis. The effect sizes, however, did change in the multivariate analysis. Readersare invited to contact the authors for details on this analysis.

44

Indeed, not all charter approvals come about as the result of political muscle and tactics. Somedistricts apparently think they have something to gain from approving charters. First, manyPennsylvania charter schools target at-risk populations and other students who are likely to createheavy burdens for district schools. Indeed, it is not unusual for districts to actively support and evencosponsor such charter schools. The CAO of one such charter school speculates, however, that hisfounding coalition would have faced rough sledding had its charter concept targeted mainstream orgifted students. Second, several charter school CAOs speculated that charter schools provide aconvenient whipping boy for districts. For instance, several CAOs reported that district personnelhad cited funding losses due to charter schools to justify poor performance on standardized tests andcutbacks in programs. Finally, in at least one instance, a charter school and its host district jointlyprocured a multi-million-dollar foundation grant. The CAO of the charter school speculates thatfunders might not have given the grant to the district without the presence of the charter school. Tosome, therefore, approving a charter appears not as an admission of failure but as an indication thatthe district is working hard to improve its programs.

We must emphasize, however, that our inability to examine the attributes of unsuccessful charterschool applications and founding coalitions leaves us unable to assess whether the aforementionedresources and conditions have any causal importance. Indeed, it may be nothing more thancoincidence that most successful founding coalitions, for instance, have strong ties to preexistingcommunity organizations and schools. We hope that others will examine this issue in greater detail.

4.7 Characteristics of Chartering Districts

Having examined some of the attributes of the founding coalitions that develop and present charterapplications, we now examine the characteristics of chartering districts. Indeed, founding coalitions,like all policy entrepreneurs, must draw upon the fiscal, human, and political resources they find intheir communities. The analysis, therefore, begins to shed light on whether districts with certaintypes of characteristics are more fertile ground for charter schools than others. Once, again,however, we are limited by the fact that we did not have good data on the incidence of charterproposals. Indeed, the analysis below simply compares districts with approved charters with districtswithout approved charters.17 We have no systematic way of knowing how many of the districtswithout approved charter schools rejected a charter proposal. Unlike the previous sections, theanalysis in this section includes all 70 charter schools approved as of Summer 2000.

45

We began our analysis by investigating differences in the number of potential charter school studentsin a district. In many ways, this is very difficult to estimate, since it is a function of not only thenumber of students in a district but also the number of others in surrounding districts who might beinterested in an alternative to their districts� offerings. Moreover, the population must include bothpublic and private school students. Given the limits of the study, we simply compared the publicschool enrollment for chartering and nonchartering districts. Data come from PDE databases for the1998/99 school year, the most recent year for which such data are available. Table 4:3 presents themedian total district enrollment for both chartering and nonchartering districts. Chartering districtstend to have significantly higher enrollments than nonchartering districts. Indeed, where the mediannonchartering district has 2,076 students enrolled, the median chartering district enrolls 3,480students. In order to test whether charter schools are more likely to be proposed and approved indistricts with concentrations of students at certain grade levels, we performed the same analysis forelementary, middle school, and high school grades. We found, however, that the same patternpersists across all levels.

Table 4:3 Chartering and Nonchartering Districts ComparedVariable Nonchartering Chartering

Median total district enrollment (1998/99) 2,076 3,480

Median per-pupil expenditure (1997/98) 6,470 6,654

Mean PSSA score (1998/99 ) 1310 1293

Median % low income students (1998/99) 23.8 63.5

Median % nonwhite students (1990) 7.4 53.0

Median % adults with BA or more (1990) 16.7 16.2Source: PDE databases and U.S. census filesNote: Mann-Whitney rank sum tests show that all group differences are statistically discernible at the .07level or lower. PSSA scores are averages across all grades and for both the math and reading portions.Averages are weighted by test participation rates. Writing scores are omitted from the averages because notall schools administered the exam. The income, race, and educational variables are all weighted by totaldistrict enrollment.

Under Act 22, charter school students come with per-pupil subsidies. Thus, we might hypothesizethat districts with higher per-pupil expenditures will be more likely to attract charter applications.As Table 4:3 shows, the median per-pupil expenditures of chartering districts is $184 higher thanthe same figure for nonchartering districts. Hence, it appears that districts with higher expendituresare more likely to receive and approve charter applications. However, the relationship reverses (butremains small) once we control for student achievement, race, and other demographic factors (seefootnote #21). Another limitation to this analysis derives from the fact that we lacked reliable data

18 Readers should bear in mind that free/reduced-price lunch status is a function of both student need andschools� diligence in registering students for lunch programs. Thus, the indicator likely includes a considerableamount of �noise.�

19 While we recognize the obvious limitations of these data, we judged them sufficiently useful to includein the analysis. Data were extracted from the Pennsylvania Education Policy Studies database, compiled byProfessor Emeritus William Cooley of the University of Pittsburgh.

46

on failed charter applications. Hence, it is possible that many districts with low subsidies attract andthen reject charter applications.

As discussed above, charter applications can be assisted by public support for charter schools. Sinceit is widely believed that charter schools are in part a response to public dissatisfaction with publicschools, we might hypothesize that lower performing districts would be more likely to attract andapprove charter schools than high performing districts. In order to test this, we calculated for eachdistrict a mean PSSA (Pennsylvania System of School Assessment) score for all grades. Table 4:3shows that, as expected, chartering districts tend to have lower overall PSSA scores thannonchartering districts. Indeed, the average chartering district reported a combined PSSA score of1293 for 1998/99 (the most recent year for which such data are available). The average noncharteringdistrict, by contrast, reported a combined PSSA score of 1310.

Because low achievement is correlated with demographic factors, we might also expect that charterschools are more likely to be located in districts with high concentrations of low income andnonwhite students. To test the income hypothesis we examined data gathered by PDE on thepercentage of students eligible for free or reduced-price lunch.18 Unfortunately, we were unable tolocate race variables in PDE files, forcing us to rely on 1990 U.S. census data.19 Table 4:3 providesevidence for both hypotheses. In the average chartering district, 64 percent of the students qualifyfor free or reduced-price lunch, compared with only 23.8 percent in nonchartering districts. Inaddition, approximately 53 percent of students in chartering districts were nonwhite, compared with7.4 percent in nonchartering districts.

Finally, as discussed above, submitting and working for the approval of a charter application takesa tremendous amount of fiscal and human resources. While we had no way to observe the supplyof fiscal resources of would-be founding coalitions across districts, we can use adult education levelas a proxy for social capital. Once again, data limitations forced us to rely on 1990 U.S. census data.Table 4:3 shows that, as expected, chartering districts have slightly higher levels of adult educationthan nonchartering districts.

The foregoing analysis is subject to important limitations. First, it is difficult to measure many ofthe factors that might lead to charter applications and approvals across districts. Moreover, there aremany other factors, such as district political environment, that remain unaccounted for. Nonetheless,these data provide a useful general profile of chartering districts. Demographically, the averagechartering district is larger, poorer, and less white. And while chartering districts generally havelower performing schools than others, they also tend to have adults with a little more education.

20 Personal correspondence with Ron Cowell, former chair of the Pennsylvania House of RepresentativesCommittee on Education. May 11, 2000.

47

4.8 A Digression on the �Philadelphia Question�

Many observers have noted the high concentration of charter schools in Philadelphia. Indeed, of the70 charter schools approved as of Summer 2000, 36 (51 percent) were in Philadelphia. Moreover,as of 1999, 69 percent of charter school students attended Philadelphia charter schools. Bycomparison, Philadelphia enrolls only 11 percent of all public school students in Pennsylvania. Inthis section we attempt to offer some preliminary explanations for this concentration.

One possible explanation is that there is something inherent in Act 22 that favors Philadelphiacharter schools. We find no evidence of this, a conclusion confirmed by a member of the legislativecommittee that worked on Act 22.20

Another possible explanation is that Philadelphia possesses many of the resources and factors that,based on the literature and observation, would seem conducive to the development of charterschools. For example, the Philadelphia school district has a large �supply� of potential charter schoolstudents. Indeed, whereas the median Pennsylvania school district enrolls 2,045 students, as of 1999the Philadelphia public schools enrolled 192,284. Philadelphia also possesses in abundance manyof the �risk factors� discussed above. First, 82.6 percent of Philadelphia students are eligible for freeor reduced-price lunch, compared with a median of 25.1 percent statewide. Second, 78.1 percentof Philadelphia students are nonwhite, compared with a median of 2.5 percent statewide. Notsurprisingly, Philadelphia students also tend to do relatively poorly on the PSSA. Indeed, thecombined district average PSSA score for the 1998/99 school year was 1135, compared with amedian of 1310 statewide. In addition to demographic factors, there is evidence that Philadelphiaalso has many of the social capital resources discussed above in its relatively dense network ofnonprofit community, church, and ethnic associations. Finally, the Philadelphia Board of SchoolDirectors appears to be quite supportive of charter schools, even if the district office has a reputationfor opposing them.

4.9 Summary

This chapter explored the process by which founding coalitions develop charter applications and bywhich they are considered and ultimately approved or denied by districts. We began with anoverview of the legal and administrative context of charter start-up, including a discussion of whomay apply for and grant charters. Under Act 22 virtually any individual or group may apply for acharter, with the exception of for-profit and sectarian groups. Act 22 is somewhat more restrictive,however, in defining who may grant charters. Act 22 grants primary authority to consider andapprove (or deny) charters to local districts (LEAs). Since July 1999, however, the Charter AppealsBoard (CAB) has heard appeals from denied applicants, those who believe their charter has beenunjustly revoked, and those who believe that an LEA has wrongly chosen not to renew a charter.

48

Because of its short history, it is difficult to assess the appeals process and its role in the charter start-up process. While the scope of this report prevents a full evaluation of the CAB, the process doesraise a number of concerns that policymakers might wish to address. The appeals process can bequite complex, time consuming, and expensive. Thus, some observers doubt that it is a feasibleavenue for some charter applicants. If this is true, then the appeals process might not provide thecounterbalance to LEA sovereignty that it was designed to. The resource requirements of appealsmight also have tilted the process in favor of districts, which tend to have more legal expertise andother resources than charter schools. In spite of this, charter schools have won nearly half theirappeals. We do not know, however, how many denied applicants were dissuaded from pursuing anappeal by the costs. While it is not our place to judge whether the process is too onerous,policymakers should consider whether its complexities and costs to charter schools are consistentwith the balance of power Act 22 sought to create between LEAs and other outside actors in thecharter approval process.

Another issue raised by the appeals process concerns the fate of schools whose charters are approvedwhen the CAB overrules the LEA. The first such schools just recently began operation and,therefore, remain beyond the scope of this study. However, it seems likely that these schools willhave unusually strained relations with their host districts. We urge policymakers and otherstakeholders to monitor this process closely.

Finally, we considered the resources required to start a charter school and the types of districtconditions that appear to provide fertile ground for charters. Not surprisingly, charter schools appearto be born of dissatisfaction with district public schools, evidenced by low PSSA scores. Thesedistricts, in turn, tend to have higher concentrations of poor and nonwhite students. The chapter�smost striking finding, however, is the extent to which successful founding coalitions appear to haverelied on ties with preexisting schools, community development groups, ethnic/cultural groups, andother organizations. These organizational ties are a key strength of Pennsylvania�s charter schools.Founding any new institution, not least a school, takes a tremendous amount of resources�fiscal,human, and political. Pennsylvania charter schools are no doubt stronger and more viable becauseof these resources. Moreover, some might argue that charter schools should not be approved unlesstheir prospects for long-term survival are enhanced by these resources. From another point of view,such dependence on external groups casts some doubt on the scalability of the charter school reformin Pennsylvania. Indeed, one might imagine that there are only so many organizations willing andable to undertake the burden of supporting charter schools. If Pennsylvania�s charter school law isdesigned to provide viable educational choices for a large proportion of its students, suchorganizational dependence might in the long run limit many students� access to these choices.Nevertheless, the evidence suggests that Pennsylvania�s charter movement continues to grow apace.However, policymakers should closely monitor the start-up process in the coming years.

In the end, the charter school start-up process is a part of Act 22�s accountability design. Approval,after all, is the first point at which public authorities and their constituents can seek to ensure thatcharter schools are likely to use their autonomy in ways consistent with the public interest. Aspolicymakers consider the start-up process, they may decide that it is better to approve fewer schools

49

in order to minimize the risks associated with letting some bad schools come into existence. Or, theymay decide to liberalize the start-up process in order to minimize the opportunity costs associatedwith failing to approve some potentially strong and innovative schools.

In the next chapter we return to the issue of charter school resources by examining the operatingfinances of the Commonwealth�s first 30 charter schools.

1 There is considerable academic debate over the effectiveness of increasing spending on school�production.� However, even those who contend that increased expenditures are not a sufficient condition ofstudent achievement often agree that it is a necessary condition. Moreover, recent studies have begun to find thatvariations in student achievement are, indeed, associated with variations in expenditures (see RAND, 2000;Weglinsky, 1997). For a more general discussion, see Hanushek, 1997.

2 Annual Financial Reports for the 1999/2000 school year are due to PDE by the end of October. Onceagain, we do not have data on the one Pennsylvania charter school that closed � Creative Education Concepts.

50

Chapter FiveCharter School Finance

The previous chapter examined the persons, organizations, and processes that led to the developmentand start-up of charter schools. We suggested that founding a charter school requires a great dealof fiscal, human, and political resources. This chapter focuses on fiscal resources. It also moves ourdiscussion from charter start-up to operations. Fiscal resources are, indeed, a crucial input in theeducational process.1 In this chapter we provide answers to three questions:

! What are the sources of charter schools� operating revenues?

! How do charter schools spend their money? How do these expenditure patterns compare withthose of other public schools?

! Are charter schools fiscally healthy?

Most of our analysis in this chapter focuses on the 30 schools open during the 1998/99 school year.2

5.1 Revenue Sources

Just as LEAs are the primary overseers of charter schools, they are also the primary funding sourcefor most charter schools. With the exception of federal and state funds, and funds raised locally, allcharter school funds are funneled through host districts. This distinguishes Pennsylvania from otherstates such as Michigan, in which charter school funds come directly from the state.

An important part of the charter school and school choice concept is that public revenues followstudents, whether they choose to attend district schools or charter schools. In theory, then, studentsare able to choose whether to attend a charter school solely on educational criteria, withoutsignificant concern for finance. As the law is written, however, charter school students inPennsylvania take with them less than their full per-pupil allotment. For most students, a charter

3 Data came from the May 3, 2000, version of PDE�s Pennsylvania Charter Schools Operators� Manual,produced by the Office of Educational Initiatives.

4 The positive skewness is also evident in the fact that the mean per-pupil payment is considerably higherthan the median: $5,849.

51

school receives from the students� LEA the total budgeted per-pupil amount minus expenditures onthe following:

! nonpublic programs! adult education programs! community and junior college programs! transportation

! special education! facilities acquisition! construction and improvement services! debt service and fund transfers

For each special education student, a charter school receives from the student�s LEA the sameamount plus the district per-pupil allotment for special education. For both special education andnon-special education, funding levels are based on the previous year�s enrollment. LEAs arerequired to make payments to charter schools in 12 equal monthly installments. Charter schools mayappeal to the Secretary of Education if the LEA fails to make timely payments. The Secretary maythen withhold the amount of the missing payment from state payments to the district.

Although relatively few districts actually host charter schools, hundreds send at least one student toa charter school. Thus, estimating precisely the magnitude of the LEA subsidy to charter schoolswould be an enormous task. We did, however, estimate the typical non-special education subsidyby obtaining from PDE documents showing payments by host districts during the 1999/2000 schoolyear.3 Since host districts can also send students to charter schools in other districts, these estimatesinclude both subsidies payed to charter schools hosted by the district and to out-of-district charterschools. A PDE official warned us that these estimates are tentative and should be used withcaution. In the absence of better estimates, we present summary statistics to give policymakers ageneral sense of the magnitude of the subsidy. We refrain, however, from publishing district-by-district estimates.

According to the estimates, the median per-pupil payment from host districts to charter schools was$5,493. Interestingly, the distribution of payments across host districts is positively skewed,meaning that there are a few districts that pay considerably more per charter school pupil thanothers.4 Indeed, payments ranged from a minimum of $4,885 to a maximum of $7,870, with astandard deviation of $762. These variations reflect the fact that LEA subsidies to charter schoolsare based on the amount districts spend on their own students.

Unfortunately, charter schools� financial reports for the 1999/2000 school year were not availableat this writing. Thus, we are unable to provide 1999/2000 estimates of LEA transfers as a percentageof total revenues. We can, however, provide such estimates for the 1998/99 school year usingcharter schools� Annual Financial Reports. These estimates are found in Appendix C. The mediancharter school received $880,843 for this year, accounting for approximately 81 percent of all

5 Once again, these estimates are taken from the 2000 Charter School Handbook. The same caveats apply.

6 Somewhat surprisingly, there appears to be little relationship between the federal funding levels (eitherTitle I or charter school grants) and concentration of low income students in a school. Unfortunately, there aremany missing values on the low income variable in the 1998/99 data. Hence, this analysis correlated 1999/2000income data with 1998/99 fiscal data. While this certainly has an �apples and oranges� quality to it, we judgedthat imperfect data were better than none. Moreover, it seems unlikely that the concentration of low incomestudents in most charter schools would change dramatically from year to year.

52

revenues. There was, however, considerable variation among the charter schools. The schoolrelying least on LEA payments received just 17 percent of its total fiscal year 1999 revenues fromLEA payments, while the school relying most on them received 89 percent.

LEA subsidies for special education students are, of course, larger. Given the importance of specialeducation, we defer discussion of these issues to Chapter 10.

Since LEA transfers to charter schools are calculated based on districts� per-pupil expenditures, itis interesting to compare what LEAs pay charter schools with what they spend on students attendingtheir own schools.5 For fiscal year 2000, the median per-pupil LEA transfer to charter schools was$2,317 (71 percent) lower than the district�s per-pupil expenditures. The differences ranged acrossdistricts from a minimum of $1,372 less than the district per pupil expenditure to a maximum of$3,545 less than the district expenditure. In percentage terms the lowest paying district provided acharter school subsidy of 63 percent of its per-pupil expenditures on district students, while thehighest paying district provided an 80 percent subsidy.

The fact that LEA transfers comprise only a part of charter schools� revenues leads us to look forother revenue sources. Heavy reliance by charter schools on external funds might raise seriousquestions for policymakers. Indeed, many foundation funders are happy to provide money for start-up projects but are reluctant to provide operating revenues on a long-term basis. Moreover, if charterschools must rely heavily on outside funders for their operations, this might create a natural limit onthe number of potentially viable charter schools and, thus, the range of choices available toPennsylvania students.

After LEA transfers, the revenue source charters rely upon most is federal funds. The most commonfederal sources are Title I monies and special charter school grants. Indeed, as illustrated inAppendix C, the median charter school in Pennsylvania received approximately 10 percent of itsrevenues from federal sources. There is, however, considerable variation among schools, with someschools receiving no federal funds in fiscal year 1999 and others receiving as much as 28 percent oftheir revenues from federal sources.6

The next most important non-LEA source of revenue for charter schools is state funds. Most suchfunds come in the form of start-up grants. As Appendix C shows, the median charter school relieson state funds for approximately 3 percent of its total revenues. Once again, there is considerablevariation among schools, with some schools receiving virtually no state funds and others as muchas 12 percent of their total revenues from state sources.

7 Since the distribution was highly and positively skewed, differences between the two groups show upin means (which are more sensitive to outliers) more than in medians. Specifically, the percentage of total schoolrevenues derived from charitable donations for the mean charter school was 5.3 percent for schools with anongovernmental (NGO) base and 2.1 percent for other schools. The medians for the two groups were 0.006 and0.4 percent respectively. Given the skewness of the distribution, we relied on a nonparametric test of statisticalsignificance. The difference was not discernible at conventional levels (the Mann-Whitney p-value was 0.19).

53

A third source of non-LEA revenues is what state budget officials refer to as �local� revenue sources.These sources include earnings on investments, charitable donations (including foundation grants),and revenues from student activities such as candy sales, car washes, and so on. In the 1998/99 fiscalyear the median charter school received just one percent of its total revenues from such sources.Once again, there was much variation among schools�much more than with other revenue sources.Indeed, while some schools received virtually none of their revenues from local sources, severalschools received more than 30 percent of their revenues in this way. One school received fully 57percent of its revenues from local sources. In these latter schools, the lion�s share of local revenuescame in the form of charitable donations. These estimates are also found in Appendix C.

For most charter schools, LEA, federal, state, and local sources comprise 100 percent of the school�srevenue. In a few cases, however, schools relied significantly on proceeds from extended termfinancing for revenues during the 1998/99 fiscal year. These revenues make up most of the �other�category in Appendix C.

Given the discussion of the resources of charter school founding coalitions in the previous chapter,we performed a number of statistical tests to check for relationships between resources at foundingand subsequent revenue flows. Specifically, we found that charter schools with strong backing fromnonprofits and other nongovernmental organizations relied slightly more on revenues from charitabledonations and from investment activities than other schools. However, the relationship was quiteweak.7 This finding provides some evidence, however, that charter schools whose foundingcoalitions include nongovernmental organizations have some fiscal advantages over other schools.

5.2 Expenditure Patterns

Under Act 22, each individual charter school is considered a separate local education agency (LEA).As such, they are separate nonprofit entities with the authority to determine their own budgets andexpenditure priorities. This distinguishes Pennsylvania�s charter school law from many other statesin which charter schools legally remain a part of their host districts and in which host districts retainlegal liability for debts incurred by charter schools. Given this relative fiscal autonomy, it isimportant to examine charter schools� spending patterns and whether they differ from those of otherpublic schools. In short, what are Pennsylvania charter schools doing with their fiscal autonomy?

Before focusing on spending priorities, however, we examined the total amount charter schoolsspend. Table 5:1 shows that the median charter school spent $7,010 per pupil during the 1998/99academic year. While most charter schools spent between $5,000 and $10,000 per pupil, some spent

8 The charter school mean and median are inflated by one extremely high outlier. Exclusion of this caseyields a mean difference of -92 and a median difference of -355. Thus, excluding this unusual school, charterschools as a group spent somewhat less than their host districts.

9 We calculated total per pupil expenditures by dividing total reported expenditures by average dailymembership. Average daily membership estimates for several charter schools are missing from the School Profilesdatabases linked to the PDE website. We received the missing values, and updated estimates for all charterschools, from Barbara Nelson of PDE.

10 The coefficient of variation is the standard deviation divided by the mean of a distribution. It is usedto compare the amount of variation in samples with different central tendencies.

54

as little as $4,000. One school reported spending nearly $30,000 per pupil. This school providesservices to troubled teenagers, which might explain the unusually high per-pupil expenditure.

In order to interpret these numbers, however, we need a comparison group. The best readilyavailable comparison is each charter school�s host district. Since host districts exist in the samemarket for labor and other school services as their charter schools, we need not make adjustmentsfor differences in cost of living. Table 5:1 shows that as a group charter schools spent slightly more($88) more than host districts per pupil. However, when we examine expenditures on a school-by-school basis, we find that the median charter school spent $253 less than its host district. Thedifference in the two estimates accounts for the presence of extreme values (outliers), mainly in thecharter school distribution. On balance, then, it appears that charter schools and host districts spentapproximately the same amount per pupil.8 This is somewhat remarkable given that, as indicatedabove, the typical charter school receives only 71 percent of its district per pupil expenditure in LEAtransfers.

Table 5:1 Charter School Per Pupil Expenditures Compared with Host Districts and AllPennsylvania Public Schools, 1998/99 (Medians)9

Schools Per Pupil Expenditure, 1998/99 Coefficient of Variation10

Charter Schools $7,010 0.59

Host Districts 6,922 0.14

Charter School minus Host District -253* 22.8Source: Annual Financial Report (PDE-2057) & School Profiles data.* Exclusion of highest (outlier) charter school yields a median difference of -92 and a mean difference of -355.

The estimates in Table 5:1, however, do not allow us to assess whether charter schools and hostdistricts have comparable student bodies and, therefore, comparable demands on their resources.For instance, charter schools might have a higher concentration of special needs students than theirhost districts. One must also bear in mind that charter schools must absorb any number of start upcosts (not the least of which is facilities) that their host districts do not. Hence, to say that they spendapproximately the same amount per pupil as their host districts might imply that they use theirresources more efficiently than host districts. Efficiency, however, involves the relationship between

11 Throughout the report, we use expenditure category definitions as set out by PDE. �Instructional�expenditures include the following subcategories: regular elementary and secondary programs, special elementaryand secondary programs, vocational education programs, other elementary and secondary instructional programs,adult education programs, and community and junior college education programs. Essentially, includes instruct-ional expenditures include salaries and instructional materials.

12 Such variations raise questions about how consistently accounting categories are applied from onecharter school to the next. Some charter schools appear to have limited human resources for budgeting and otheradministrative functions. Unfortunately, we have no way to independently verify the contents of schools� AnnualFinancial Reports.

13 This is a histogram table. Thus, the top row indicates that 8 charter schools spent 20 percent or lessthan their host districts on instructional items; the second row shows that another 7 charter schools spent between10 and 20 percent less than their host districts, and so on. For regional charter schools (i.e., those with more thanone host district/sponsor), we created composite host district comparisons by averaging the values for all of thehost districts. Ideally, one would weight this average by the number of students in the charter school from eachdistrict. Given the limits of the project, we were unable to obtain such information on a broad scale.

55

fiscal (and other) input and various student outcomes. Thus, we defer discussion of these issues untilChapter 11, which deals with student outcomes and their relationship to charter school inputs.

Beyond the total amount of expenditures per pupil, the most important characteristic of a school�soverall expenditure patterns is the proportion of its funds spent on instruction versus other functions.In order to estimate the percentage of charter school expenditures devoted to instruction, weexamined Annual Financial Reports from the 1998-1999 school year, the most recent year for whichsuch reports are available. Results of this analysis are presented in Table 5:2. We found tremendousvariation among charter schools.11 Three charter schools, for instance, spent less than 40 percent oftheir total expenditures on instructional items, while four charter schools spent more than 90 percenton such items. One school reports spending 100 percent of its total expenditures on instructionalitems.12 Taken as a group, the average charter school spent 59 percent of its total 1998-1999expenditures on instructional items. The mean value for all Pennsylvania public schools, bycomparison, was 66 percent while the mean value for all districts sponsoring charter schools was 65percent. Taken as a group, then, charter schools spent fewer of their resources on instructional itemsthan other schools during the 1998-1999 school year.

Table 5:2 Percentage of Total Expenditures Devoted to Instructional Items, 1998-1999

Group Mean Minimum Maximum

Charter Schools 59% 35% 100%

Host Districts 66% 61% 73%

All Pennsylvania Public Schools 65% 46% 73%Source: Annual Financial Report (PDE-2057) & School Profiles data.

Continuing with the charter school-host district comparisons, Table 5:3 provides the frequencydistribution for the percentage of charter school expenditures devoted to instruction minus the samenumber for each school host district.13 The table shows that two-thirds (20) of the 30 charter schools

14 Most charter schools spent little or nothing on such services during the 1998/99 school year. Of the5 schools that reported such expenditures, however, they comprised up to 37 percent of total expenditures. Thus,for some schools, the need to acquire, construct, or improve facilities constitutes a large burden. While Act 22prohibits charter schools from using public funds to construct new facilities, they may use other funds to do so.

56

Table 5:3 Instructional Expenditures as a Percent-age of Total for Charter Schools Minus the Same Percentage for Host Districts

Difference Number of Schools

>-20% 8

-20 to -10% 7

-9 to 0% 7

1 to 9% 3

10 to 20% 3

>20% 2

Total 30Source: Analysis by the Evaluation Center

under examination spent a lower percentage oftheir total expenditures on instructional itemsthan their host districts�some by more than 20percentage points.

This finding leads to two interrelated questions.First, why do charter schools as a group devotea smaller share of their total expenditures toinstructional items? Second, why is there suchwide variation among charter schools? Webegin with the first question.

One possible answer to the first question is that,as start-up organizations, charter schools mustbear many one-time and fixed costs (e.g.,renting and/or renovating facilities) thatestablished districts have either covered in thepast or have spread out over time by means ofcapital budgets. If this is true, then the fact thatcharter schools spend a lower percentage of their total expenditures on instructional items is notnecessarily an indication of inefficiency or waste. Rather, it might simply reflect the relative highand fixed start-up costs associated with founding a new school. While available data do not allowus to fully dispose of this issue, we examined the percentage of total expenditures charter schoolsuse for support services. This category includes instructional staff, administration, health, business,maintenance, transportation, and rent payments for school facilities. (The �facilities� category, bycontrast, includes expenditures to acquire and improve buildings14).

Table 5:4 shows that for the average charter school, �support services�� which include rent andfacilities maintenance�comprised 38 percent of total expenditures, compared with 32 percent of hostdistricts and 33 percent for all Pennsylvania public schools. Hence, there is at least some evidencethat for charter schools, instructional expenditures are �crowded out� to a certain degree by the needto cover start-up and other related expenses. Another possible explanation is that charter schoolstypically pay their teachers less than other similar schools (see Chapter 8). This, in turn, wouldreduce the demand on their instructional budgets. We emphasize, however, that this is only apreliminary answer to the question, and that it deserves further attention. A fuller accounting ofcharter school expenditure patterns can be found in Appendix C.

15 As with the other variables, the distributions were skewed. We depart from the normal practice ofreporting medians, however, because many of the medians are zero. Readers should bear in mind that means areparticularly sensitive to outliers.

16 The mean proportion of expenditures devoted to instruction for schools with NGO backing was 60.7percent, compared with 58.5 percent for other schools. The mean is more appropriate than the median in this case,since the distributions are reasonably symmetrical. Looking at the medians, however, produces the opposite result,with NGO-backed schools devoting 55.8 percent and other schools 59.3 percent of their total expenditures toinstructional items.

57

Table 5:4 Expenditure Patterns in Charter School and Other Public Schools (means)15

Expenditure Category Charter Schools Host Districts All PA Public SchoolsInstruction 59% 66% 65%Support Services 38 32 33Noninstructional Services 0 2 2Facilities Acquisition & Improvement 2 n/a n/aDebt Service 1 n/a n/an/a = data not availableSource: Annual Financial Report (PDE-2057) & School Profiles data.

Turning next to the question of why there is so much variation in expenditure patterns among charterschools, once again the data are limited. Nonetheless, there is some evidence that schools withstrong backing from nongovernmental organizations devote a larger share of their total expendituresto instruction. These groups might provide facilities and other resources that allow charter schoolsto devote a larger share of their operating budgets to instructional items. The relationship, however,is weak and somewhat sensitive to choice of statistical technique.16 We hope that others willexamine in greater detail the reasons behind these apparent patterns.

5.3 Fiscal Viability

For charter schools to operate successfully in the long run, they must be fiscally viable. While it isbeyond the scope of this report to provide a full fiscal audit of Pennsylvania charter schools, weexamined two indicators of fiscal viability. First, we examined charter schools� capacity to developand execute budgets by looking at variances between budgeted and actual revenues and expenditures.Second, we examined charter schools� end-of-year balance to see if there were any deficits. Finally,as with the earlier sections of this chapter, we tested explanations for observed variations in charterschools� fiscal health.

In order to provide a composite picture of schools� capacity to budget, we calculated the variancebetween budgeted and actual revenues and between budgeted and actual expenditures. Highvariances can cast doubt on a school�s ability to effectively plan and execute educational andorganizational strategies. In operational terms, a revenue variance is the amount of funds actually

17 Readers will note that there are large differences between the mean and median values. This isbecause each distribution is highly skewed toward the high end. In other words, in each case there are a smallnumber of schools with large positive variances, even though most schools have lower variances. Since the meanis more sensitive to outliers, it is substantially higher than the median in each variable. In such cases, statisticiansgenerally use the median so as to avoid overrepresenting outlying values.

58

received minus the amount the school budgeted for. The same holds true for expenditures. In orderto facilitate comparisons among high- and low-budget schools, we converted these variances intopercentages of the total revenues (expenditures) budgeted for. Thus, for instance, a revenue varianceof 20 percent means that the school received revenues that were 20 percent higher than thosebudgeted for, whereas a revenue variance of -20 percent means that the school received revenues thatwere 20 percent less than those budgeted for. Naturally, it is better for schools to have positivevariances in revenue and negative variances in expenditures.

On the revenue side, the median school received 1.7 percent more than it budgeted for (see Table5:5). Thus, taken as a group, Pennsylvania charter schools budgeted conservatively and receivedmore than expected. As with the other fiscal variables discussed in this chapter, there was greatvariation among charter schools. Indeed, revenue variances ranged from -47.5 percent (meaning thatthe school received much less revenue than expected) to 96.7 percent (meaning that the schoolreceived much more than it budgeted for). Two schools reported no budgeted revenues for the1998/99 school year. These were excluded when calculating summary statistics.

Table 5:5 Variances in Revenues and Expenditures as a Percentage of Budgeted Revenues andExpenditures, 1998-1999

Variable Median17 Mean Minimum Maximum

Revenue Variance 1.7% 7.4% -47.5% 96.7%

Expenditure Variance -11.4 -1.4 -46.1 63.2

Source: Annual Financial Records (PDE-2057)

On the expenditure side, Pennsylvania charter schools appear to have been similarly conservative.Indeed, the median charter school spent 11.3 percent less than it budgeted for. Once again, there wasgreat variation among charter schools. Expenditure variances ranged from -46.1 percent (meaningthe school spent much less than it expected) to 63.2 percent (meaning the school spent much morethan it expected). Once again, two schools reported no budgeted expenditures and we excluded thesewhen calculating summary statistics. A complete listing of revenue and expenditure variances canbe found in Appendix C.

In spite of the wide variation in charter schools budgetary precision, we find generally that schoolsthat underestimate their revenues also underestimate their expenditures, and vice versa. Figure 5:1illustrates the relationship between revenue variances and expenditure variances by means of ascatterplot. The horizontal axis represents the percent variance in revenues while the vertical axis

59

Va

ria

nce

in e

xp. a

s %

of b

ud

ge

t

Variance in rev. as % of budget-.5 0 .5 1

-.5

0

.5

1

Figure 5:1 Revenue and Expenditure Variance Compared, 1998/99

represents the percent variance in expenditures. The graph divides the two-dimensional space intofour quadrants. The lower-left quadrant represents schools that underestimated both their revenuesand their expenditures. In such cases, even high variances should provide little cause for alarm.Similarly, the upper-right quadrant represents schools that overestimated both their revenues andtheir expenditures. Once again, these cases provide little cause for alarm. The lower-right quadrantrepresents schools that overestimated their revenues and underestimated their expenditures � onceagain, no cause for concern. The upper-left quadrant, however, might provide cause for concern,because such schools would have underestimated their revenues and overestimated theirexpenditures. Fortunately, there are only two schools in this quadrant. Both schools, moreover, lievery close to the intersection of the �zero� lines, indicating that the magnitude of both revenue andexpenditure variances was very small.

Pennsylvania charter schools, in sum, did a remarkably good job of budgeting, both on the revenueand expenditure sides of the fiscal equation. Readers should bear in mind, however, that these datarepresent only one year of expenditures.

While providing thorough explanations for these findings lies beyond the scope of this report, wedid test a number of explanations that derived from the discussion in Chapter 4. First, we

18 We used Mann-Whitney rank sum tests to assess statistical significance. 19 In particular, we compared each school�s end-of-year balance with the highest grade level it included.

60

hypothesized that charter schools in their second year of operation would have more experience atbudgeting and thus would have lower variances. The extra year of experience appears to have hadno notable impact on budgeting for revenues. However, we did find a modest effect on theexpenditure side. Whereas the median expenditure variance for schools in their first year ofoperation in 1998/99 was -12.7 percent, it was only -2.1 percent for schools in their second year ofoperation. We tested for similar differences in schools according to whether they were built uponthe organizational foundations of preexisting schools. Once again, we found a difference on theexpenditure side but not the revenue side. The median expenditure variance for charter schoolsbased on preexisting schools was -6.8 percent, compared with -11.1 for other schools. Finally, wetested whether schools with backing from nongovernmental organizations had lower variances thanother schools. Surprisingly, we found that schools with NGO backing had higher variances thanother schools. In all the analyses, however, the small number of cases prevented us from controllingsufficiently for variances in target population, community characteristics, and so on. Thus, wecannot rule out the possibility that second year schools were stronger in some unobserved respectthan first schools. Indeed, one might theorize that the second year group of schools was strongersince applicants had more time to observe and learn from the experiences of the first round of charterschools. None of the differences reached conventional levels of statistical significance.18

The ability to accurately anticipate revenues and expenditures in the budgeting process should enablecharter schools to avoid running deficits. Given that charter schools appear to do a good job ofbudgeting, we should expect that few, if any, would run deficits. Examination of charter schools�1998/99 Annual Financial Reports (AFRs) reveals that the median balance for all charter schools was$84,380, or 11.4 percent of total expenditures. However, there was significant variation amongschools. Seven of the 30 schools (23 percent) we examined showed negative end-of-year balances,the largest of which was over $400,000, or 10.7 percent of that school�s total expenditures. On thepositive side, some schools showed positive balances of up to 58 percent of their total expendituresfor the year. Thus, while most charter schools appeared to be fiscally healthy in the 1998/99 schoolyear, there were a number of trouble spots. A complete listing of end-of-year balances can be foundin Appendix C.

Once again, we investigated a number of explanations for variations in end-of-year balances. First,recognizing that secondary schools are generally more expensive to run than elementary schools, wetested whether schools with older students were more likely to run deficits,19 but found no cleardifference. Second, we tested the hypothesis that schools backed by nongovernmental organizationswould be more likely to run positive balances, but again found no clear difference. Finally, we testedfor an �experience� effect by comparing the balances of first year schools with those in their secondyear of operation. Surprisingly, we found that schools in their first year of operation were morelikely to run positive balances than schools in their second year of operation. This difference

20 The median balance for first year schools was positive at 14.2 percent of total expenditures, while thesame value for second year schools was negative 1.3 percent. A Mann-Whitney rank sum test showed that thedifference was statistically discernible at the .02 level.

61

surpassed conventional levels of statistical significance.20 One might theorize that second yearschools had to deplete many of their start-up resources in the first year, or that the first year schoolswere stronger because they were able to observe and learn from the experiences of the 1997/98cohort of charter schools. This issue certainly deserves further exploration as policymakers studyways to modify Pennsylvania�s charter school law.

5.4 Summary and Conclusions

Like other school choice policies, Act 22 mandates that funding follows students. Thus, schoolshave an incentive to work to satisfy students and their parents. Under the terms of the Act, thisfunding is funneled through host districts and other LEAs sending students to a given charter school.The size of the LEA subsidy is based on the LEA�s per-pupil expenditure on its own students anddiffers for special education and non-special education students. We defer detailed discussion ofspecial education finance to chapter10.

The first section of the chapter examined sources of charter schools� revenue. Using PDE data, weestimate that the median charter school received approximately 81 percent of its total revenues fromLEAs during the 1998/99 school year. There is, however, a large amount of variation among charterschools. Next to LEA transfers, the largest revenue source for charter schools is the federalgovernment, mostly through Title I monies and special charter school grants. While there isconsiderable school-by-school variation, the median charter school received approximately 7 percentof its total revenues from the federal government. The remainder of charter school revenues camefrom state grants (e.g., start-up grants) and �local� sources, such as earnings on investments,charitable donations, and revenues from student activities (e.g., candy sales, car washes, and so on).In addition, a few schools relied on proceeds from extended term financing during the 1998/99school year. While the scope of the study prevents us from attempting to provide detailedexplanations for variations in charter school revenue sources, we did find that schools connected tonongovernmental organizations rely on charitable donations and investment activities slightly morethan other schools. This suggests that the advantages of group affiliations in the context of start-upmight continue into the early years of the schools� operations. However, the evidence to date is quiteweak. Nonetheless, the role of NGOs in charter school operations bears closer examination.

An important policy issue follows from the fact that the typical charter school relies on non-LEAsources for approximately one-fifth of its total revenue. From one point of view, such reliance onnondistrict sources is good inasmuch as it tends to favor schools that bring a great deal oforganizational, fiscal, and social capital to the table. In this view, these schools are more likely torealize some of the major goals of privatization�to leverage community resources so thatgovernments can do less with more and to build a sense of collective responsibility for schools andstudents. From another perspective, charter schools� reliance on non-LEA sources is worrisome,

62

since it raises questions about the sustainability and scalability of the reform. There are, after all,a limited number of organizations willing and able to sponsor charter schools. Foundations andother charitable organizations, moreover, are often more sanguine about providing start-up moniesthan about covering long-term operating expenses. This dependency on externals funds might beespecially troubling, from this point of view, if charter schools must use them for operating costs aswell as one-time start-up expenses. Thus, what at first blush appears to be a strength of manyPennsylvania charter schools might turn out to place limits on the range of charter-related choicesavailable to students. Ultimately, resolution of this debate depends on how efficiently charterschools are spending their revenues.

The second section of the chapter examined charter schools� expenditure patterns. We estimate thatcharter schools spent approximately the same amount per pupil as their host districts during 1998/99.Of that total amount, we found that charter schools typically spend a smaller percentage oninstructional items than their host districts, and a large percentage on support services (which includerenting and maintaining facilities) and on other noninstructional items. The reasons for thesedifferences in expenditure patterns might lie more in the exigencies of starting new schools (e.g.,acquiring and maintaining physical facilities) than in any inherent inefficiency in charter schools.However, these are questions that can be answered only with the passage of more time.

The final section of the chapter examined charter schools� fiscal viability using two indicators. First,we found that charter schools appear to be relatively conservative in budgeting, taking in more thanexpected on the revenue side and spending less than expected on the expenditure side of the ledger.Moreover, there is some evidence that schools benefit from experience, as second year schools hadslightly lower expenditure variances than first year schools (there was no discernible relationship onthe revenue side). Second, we examined charter schools� end-of-year balances for the 1998/99school year. We found that 7 schools (23 percent) ran negative balances (deficits), the largest ofwhich constituted 10.7 percent of total expenditures.

In the final analysis, the most important question about charter school finance is how efficiently theyspend their resources. Answers to questions about efficiency, however, require evaluators toexamine the relationship between fiscal inputs and student outcomes. Chapters 11 and 12 examinestudent outcomes in charter schools. Unfortunately, the data are insufficient to yield reliableestimates relating costs to outcomes.

Before examining outcomes, however, we turn in the remainder of this section to an evaluation of twoimportant types of nonfiscal inputs and resources. The next chapter provides an overview of thecharacteristics of charter school students and their families. The final chapter of this section assessesthe characteristics of charter school teachers and staff.

As we noted earlier in the chapter, the accuracy of some of the data may not be as high as we wouldwish. However, compared with other states where we have evaluated charter schools, particularlyMichigan, the completeness and accuracy of the Pennsylvania financial data seems quite good. Inpart this is due the technical assistance and support provided by the Office of Educational Initiatives.

63

Chapter SixStudent and Family Characteristics

Act 22 states that charter schools should be open to all students. Hence, in this chapter we comparecharter students’ characteristics with those of noncharter students. Also, the charter concept (likeall policies) makes certain assumptions about the behavior and attitudes of its target population(students and families). In this chapter, we attempt to test some of these “coproductive”assumptions. Questions we will address include the following:

F How do charter students’ and families’ background characteristics compare with those ofnoncharter students, particularly in the host district?

F What kinds of schools did students attend before coming to the charter school?

F How much student turnover is there in charter schools?

F Why did students and their families choose their charter school?

F Are students and parents aware of the school’s mission?

6.1 Sampling of Students and Parents

Student Sample

For the purpose of this study, a sample of students completed questionnaires in two consecutiveschool years: 1998-99 and 1999-00. Only students in grades 5 and above were included in thesurveys. Twenty-three schools and 923 students were included in the 1998-99 sample (8 of the 31schools were not included in 1999 since they did not have any students enrolled in Grades 5 orhigher). In the second sampling during spring 2000, 25 of the original 31 schools and a total of1,106 students were included in the sample (only 5 schools were not included in 2000 because theydid not provide instruction at Grade 5 and above; also, one school was closed after the 1999-00school year).

In the selected classes, all students were targeted, but a number of students were either absent or notpresent during the administration of the surveys and, therefore, were not included. The response rate,which was calculated on the number of students who completed the surveys in the sampled classes,

64

was high for both samples (90.4 percent in 1999 and 90.3 percent in 2000). One cyber school hada very low response rate in both years due to difficulties in corresponding with students who had tobe sampled though the mail.

The aim of the sampling was to select at least 3 classes and at least 40 students at each school. Insome cases the number of students was fewer than 40. In these cases, we sampled all the studentsenrolled at the school in Grade 5 and above. To the extent possible, an effort was also made to selectclasses instructed by different teachers and classes that represented different subject areas. Onlyclasses required by most/all students at a particular grade level were sampled in order to avoidselection bias.

All members of the evaluation team received a handbook with instructions and support materials forthe data collection before the work started. Members of the evaluation team administered thesurveys to students. Evaluation team members read general instructions and then explained theitems that some students typically have trouble with. For classes in Grades 5 and 6, the surveyadministrator read through the survey item by item as the students completed the work. Additionally, in-service training was provided to new members of the evaluation team before theyadministered surveys to students on their own.

Parent Sample

Depending on the size of the school, between 25 and 35 parents from each school were randomlyselected during a visit by a member of the evaluation team in May 1999 and in the Spring of 2000.The random sample was drawn from a roster of all students. Survey packages containing a coverletter, questionnaire, and a return envelope were prepared for each selected family. Each charterschool was encouraged to prepare and include its own cover letter in order to make the contact withthe families more familiar. The questionnaires were sent home with the students or delivered to theparents by mail.

Upon completion, the questionnaires were returned to the designated person at the school, and thename of the family was checked off the list to denote that the survey was completed and returned.The designated person at each school was instructed not to open any of the sealed envelopes, sincethis would compromise the confidentiality of the respondents. A summary of the results for eachschool was returned to each school after the data analysis was completed.

This component of the evaluation was optional during the first year. It was not initially scheduledin the evaluation proposal, but representatives of a number of charter schools, along with staff at thePennsylvania Department of Education, thought parent feedback was important. This issuewarranted inclusion during the first round of data collection, even though there would not besufficient time for follow-up before the close of the school year. During the second round ofsurveying, in the spring of 2000, more time was available to conduct a follow-up. Two dollars wereenclosed with each survey. As many as three follow-up surveys were sent to some families and, insome schools, as many as four follow-up calls were made to non responding families. The effort put

65

into the follow up increased the overall response rate, even though the response rate was below ouranticipated goal.

In 1999, 11 schools did not conduct parent surveys or were unable to obtain a sufficient responserate to have their parent responses included in the state totals. In 2000, 5 schools did not receive asufficient response rate and were therefore not included in the totals for the state. Altogether, theresponse rate was 51 percent.

Typically, a 40 percent response rate has been used as a cutoff point for decisions about includingschools in the total sample. A number of schools had parent response rates between 25 percent and40 percent. Since these schools had large numbers of families targeted and because the results fromthis subset of schools did not exhibit any noticeable differences from the remainder of the schools,a decision was made to use a 25 percent response rate as a cutoff point. Four schools had responserates below 25 percent, so they were excluded from the sample. Therefore, 26 of the 30 participatingcharter schools were included in the analysis of total parent results. Within these 26 schools, 364of the 777 families sampled returned a completed survey, making the total response rate 46.8percent. Appendix B includes specific information about the survey results for parents and students.

6.2 Description of Charter School Students and Families

In this section, we describe the students enrolled in the charter schools. We have relied on a varietyof sources, including data reported to the Pennsylvania Department of Education by the charterschools. Other sources of data included were the annual reports prepared by the charter schools, anddata collected directly from students that took part of our sample. An attempt is made to refer to thespecific source of the information as it is presented.

For the 2000-01 school year, enrollment in Pennsylvania charter schools is expected to exceed20,000 students. This is a rapid increase from the 1,179 students that were enrolled in the first yearof the reform (1997/98). Chapter 3 contains more specific information about the growth of thecharter school reform in terms of the number of schools and enrolled students.

Grade and Age of Students

Students taking part in the survey were rather evenly divided between the middle and high schoollevels as well across the specific grade levels. However, Grades 8 and 12 were somewhatunderrepresented, with each having fewer than 92 students. Figure 6:1 depicts the distribution ofstudents by grade for both samples. The students ranged from 10-20 years in age. They were ratherevenly distributed by age, although the majority of the students fell between the ages of 11 and 17.Only 75 students were between the ages of 18 and 20. The average age of the students sampled in2000 was 13.8 years, which was the same as for the sample in 1999.

66

Distribution by Grade

0255075

100125150175200225

5th 6th 7th 8th 9th 10th 11th 12th

1999 (N=923)

2000 (N=1,106)

Figure 6:1 Distribution of Sampled Students by Grade

Distribution by Grade, All Charter Schools in 1998-99

0

100

200

300

400

500

600

700

800

K 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 10th 11th 12th

Figure 6:2 Distribution of All Pennsylvania Charter School Students by Grade

The children of the sampled parents were rather evenly distributed across the various grade levels.Again, Grades 8 and 12 were underrepresented, each with less than 4 percent of all parents. Justover 50 percent of the parents had children enrolled at the elementary level (an average of 8.58percent at each grade between Kindergarten and Grade 5), 24.6 percent had children enrolled at themiddle school level (an average of 8.2 percent for each of Grades 6-8), and parents with studentsenrolled at the high school level accounted for 23.9 percent of all parents returning a questionnaire(an average of 5.97 percent at each grade level, 9-12).

67

Distribution by Race/Ethnicity in 1999-2000

0%

20%

40%

60%

80%

100%

Asian/PacificIs lander

Black Hispanic Native Ameri-can Indian

W hite

Sampled students (N =1,073)Sampled parents (N = 347)All students in 30 participating schoolsAll students in 44 schools with data

Figure 6:3 Distribution of Charter School Students by Race/Ethnicity, 1999-00

Gender and Ethnicity

Students were fairly evenly distributed by gender, with 51.3 percent females and 48.7 percent males.It is interesting to point out, however, that there was considerable variance among schools. Someschools had high proportions of either males or females, depending on the nature and profile of theschool. Minorities, which account for 73.2 percent of the sampled charter school enrollments, areobviously highly represented in Pennsylvania’s charter schools. Among the sampled students thelargest group by race were Blacks with 51.4 percent, followed by Whites with 26.8 percent,Hispanics with 15.8 percent, Native Americans with 3.6 percent, and Asian/Pacific Islanders with2.4 percent.

Figure 6:3 illustrates the breakdown of students by Race and ethnicity and highlights differencesbetween the sample of students we drew from grades 5-12 officially reported figures for all studentsin the 30 participating schools as well as for all schools that were operating in 1999-2000. Theyellow bars in the figure illustrate the most accurate breakdown for all charter schools. It isinteresting to note that the schools that were opened in 1999-2000 have a lower proportion ofHispanics and a higher proportion of White students than the schools that were already open.

Because the charter schools are highly concentrated in urban areas, and particularly in Philadelphia,we can expect a large proportion of students from minority backgrounds. In contrast, the NationalStudy of Charter Schools Report (RPP, 2000) gives the demographic breakdown for charter schoolsnationally with 48.2 percent white, 23.5 percent black, 21.1 percent Hispanic, 3.4 percent Asian orPacific Islander, 2.6 percent American Indian or Alaska Native, 1.2 percent other. From thesefigures we can see that Pennsylvania charter schools are attracting more minorities than are charterschools in other states.

68

Some studies of charter schools have indicated that charter schools are leading to greater segregationbased upon race. Cobb and Glass (1999) found that charter schools in Arizona had fewer minoritiesthan the districts in which they lie and that half of the charter schools exhibited evidence ofsubstantial ethnic separation, meanwhile, Horn and Miron (1999) found that the proportion ofminority students was dropping sharply in Michigan charter schools.

By contrast, the proportion of minorities in Pennsylvania charter schools is substantially higher thanthe state average and comparable to the school districts in which the charter schools are established.Among the 30 participating schools in our study, 79.6 percent of the students are minoritiescompared with 57 percent minorities in the host districts of these charter schools. With the additionof the schools that opened for the 1999-00 school year, the proportion of minorities dropped 4percent to approximately 75.6 minorities for the 44 charter schools for which we have data. At thesame time, the proportion of minorities across the host districts also dropped 4 percent toapproximately 52.9 percent. The drop in the proportion of minorities in the charter schools was dueto the addition of new schools. These new schools were located in districts that had fewerminorities. This indicates that the schools that opened in 1999-00 have fewer minorities. Whencomparing the proportion of minorities school by school, we find that all but 8 charter schools enrolla higher proportion of minorities than their host districts.

While the student sample included only schools that had at least one class between Grades 5 and 12,the parent sample included all schools, even those with only lower elementary grades. Eighty-sixpercent of the parents responding were female.

The proportion of minority parents in the parent sample should reflect the proportion of enrolledstudents. Figure 6:2 illustrates that the sample of parents had more White families and fewerHispanic families than the sample of students. The differences could be due to the fact that theminority families have more siblings enrolled in the charter schools; thus, the proportion of minorityfamilies is not as high as the proportion of minority students. Another partial explanation is that the5 schools excluded due to a low response rate had a considerably higher proportion of minoritystudents than the 25 schools included in this component of the study. It is also possible that theminority families were underrepresented in the sample because they did not complete and return thesurvey. It is important to point out that, although some schools have a high proportion of parentswho do not have English or Spanish as their first language, the parent survey was made available inEnglish and Spanish only.

Proportion of Low Income Students

Family income is an important determinant of students’ achievement (see chapter 11). For thisreason we thought it was of particular interest to compare the proportion of low income students, asdefined by eligibility for free/reduced lunch, in charter schools with that population of students inthe districts from which they come. For the 30 schools participating in our study in 1999-00, 68percent of their students came from low income households compared to 66 percent of students in

69

Average Household Income

0%

5%

10%

15%

20%

25%

Less than$10,000

$10,000-$19,999

$20,000-$29,999

$30,000-$39,999

$40,000-$59,999

$60,000-$99,999

$100,000or more

Figure 6:4 Average Household Income of Charter School Families

the sending districts. While these figures are very similar, it is important to point out that there isconsiderable variation both across the charter schools as well as across the sending districts.

When we include the schools that were started in 1999-00, we find that the charter schools have alower proportion of students that are classified as low income compared to the sending districts (i.e.,61 percent low income for the 44 charter schools we had data on in 1999-00 compared with 63percent for the sending districts).

Table 6:1 Proportion of Low Income Students in Charter Schools, Sending Districts, andNonsending Districts

Proportion Low Income, 30 Participating CharterSchools Compared to 30 Host Districts

Proportion Low Income, 44 CharterSchools Compared to 40 Host Districts

Charter Schools 67.8% 61.0%

Districts WithCharter Schools 66.0% 63.5%

Districts WithoutCharter Schools 24.8% 23.8%

Table 6:1 illustrates the differences in the proportion of low income students for charter schools, aswell as for the districts that have charter schools within their boundaries or from which charterschool students originated. While the differences are small between charter schools and the sendingdistricts, the differences are great between school districts that are losing students and districts thathave no charter schools and that are losing no students to charter schools.

Comparison of average household income provides another way to examine the distribution ofstudents in charter schools. The annual family income reported by the sampled parents indicated that

70

about 66 percent of the parents had annual family incomes between $20,000 and $100,00, with 2.8percent of the families over $100,000. One-third of the families had incomes under $20,000 (21.4percent between $10,000 and $19,999 and 9.7 percent below $10,000), which clearly indicates thatthese charter schools are providing educational services to a large number of low-income families.

Fifty-seven percent of the families were in two-parent homes, while 41.7 percent were in singlefamily homes and 1.7 percent lived in other types of households (e.g., student living with otherrelatives or residing in state facility).

Highest Level of Education Charter School Students Plan to Complete

The students were asked about the highest level of education they planned to complete. They wereasked to choose between the following categories: High School, 2-Years of College, 4-Years ofCollege, Graduate School and Not Sure Yet. These categories were explained to the students andexamples of careers requiring the specific level of schooling were provided. Seven percent of thesampled students expected to stop their schooling after high school, while 9.4 percent planned tocomplete a 2-year degree, 28 percent planned to complete a 4-year degree, and approximately 38percent, planned to go on to graduate school, which was explained to the students to include trainingafter the bachelors degree that included a masters, doctoral, or professional degree. Not surprisingly,nearly a fifth of the students were still not sure about their future school/training plans beyond highschool. The students’ reponses regarding their further educational plans was noticeably more diversein 2000 than in 1999. The proportions of students either not planning to go to college or elseplanning to go to graduate school increased noticeably in 2000, while the proportion of students withplans to attend four years of college decreased.

Level of Parents’ Formal Education

The level of formal education of the sampled parents varied considerably: slightly more than 7percent indicated that they had not completed high school, 33.5 percent of the parents ended theirformal schooling after graduating from high school, 32.1 percent completed less than 4 years ofcollege, 9.2 percent obtained a bachelors degree, 5.6 percent had a BA plus some graduate courses,and 12.3 percent, a slight increase from last year, completed a graduate or professional degree.

Length of Enrollment at Charter School

Nearly 47 percent of the students reported that they had been enrolled in their charter school for oneyear or less, a logical decrease from 88 percent in the previous year when most of the schools werestarted. Forty-six percent of the sampled students had been enrolled for 2 years. Just over 7 percentof the students indicated that they were enrolled for 3 or more years. Since the charter school reformis only three years old, the maximum number of years a student could be enrolled in a charter schoolis 3. However, some of the schools were private conversions and the students indicated that theywere enrolled more than three years.

71

Previous School Attended

Nearly 80 percent of the sampled students in the Pennsylvania charter schools had previouslyattended public schools. Of the remaining 20 percent, 16.2 attended either a private or a parochialschool, and 1.2 percent were home schooled. The remaining students either did not attend school,attended some other form of school, or else they did not respond to the question. The proportion ofstudents transferring from private/parochial to charter schools increased from 16 percent in 1999 to17.4 percent in 2000. Because our sample only includes students in grades 5-12, and because alarger proportion of students move from private schools in the lower elementary grades, it is likelythat our sample underestimates the proportion of students moving from private/parochial school tocharter school.

The enrollment in charter schools of students who had not previously attended a public schoolrepresents an additional burden for the host school districts. Districts must divert a portion of theirresources to educate these students yet the students do not represent a decrease the number ofstudents left in the district. Representatives from Philadelphia Public Schools indicated that studentswho move from nonpublic schools to charter schools represent a heavy financial burden on theirbudget. The district indicated that 25 percent of the students enrolled in charter schools are movingfrom nonpublic schools. Our sample of students from Philadelphia indicated that 18.6 percent ofthe students were previously enrolled in nonpublic schools, which is an increase from 16.8 percentin 1999. Our sample of schools does not include the charter schools that opened in 1999-00 and insome of these schools there appear to be substantially more students moving from private/parochialschools.

Among the schools in our sample, there was considerable variation in terms of the types of schoolsfrom which the charter schools were attracting their students. While several of the charter schoolshave no students that previously attended a private/parochial school, a few schools have close to halfof their students coming from nonpublic school backgrounds.

The three schools with the highest proportion of students coming from nonpublic backgroundsinclude one private school conversion, one operated by a for-profit educational managementorganization, and one charter school with college prep profile located in Philadelphia. These schoolsalso had a low proportion of minority students and low-income students.

Figure 6:5 illustrates the differences between our student and parent samples in terms of what they reported as the previous school attended. We have grouped these various categories into three areas:public school, nonpublic school, or other. The sample of parents covered grades K-12, while thesample of students covered only grades 5-12. It is common that after the first year of charter schooloperation the largest group of new students is at the entry grade level for a particular school. A moredetailed look at the parent responses indicates that more than 5 percent of the parents selected“other” because their child was just entering school in kindergarten and did not previously attend aschool. Surprisingly, 7 percent of the parents also indicated that they were switching from anothercharter school.

72

Type of School Attended Before the Charter School79.6%

17.6%

2.9%

69.8%

17.5% 12.7%

0%

20%

40%

60%

80%

100%

Public school Nonpublic School(private, parochial, or home

schooled)

Other(other charter school, other

school type, or did notattend school)

Sampled Students

Sampled Parents

Figure 6:5 Type of School Attended Before Charter School: Responses from Sampled Students and Parents

The movement of students from nonpublic schools into charter schools results in local districtshaving to pay for educating children that were not previously enrolled in the public schools. Wehave seen in other states that the local districts report a large movement of students to charter schoolsin their first year of operation; but over time, the transfer of students back and forth between thecharter school and the local school district equalizes (Horn & Miron, 2000). In reality, the numberof students moving to the charter schools is still high because elementary-level charter schools willcontinue to enroll large groups of students but only at the kindergarten level. Since the local districtsnever enrolled these students, they do not appear in their records as transfers.

Amount of Time Volunteering at Charter Schools

It was interesting to find that 76.9 percent of the parents reported that they did not volunteer at schoolat all, or to a very limited degree (i.e., less than 3 hours per month). A much smaller proportion ofthe parents reported volunteering quite extensively. Just over 12 percent of the parents volunteeredbetween 4 and 6 hours per month, 4.7 percent volunteered between 7 and 9 hours per month, 1.9percent volunteered between 10 and 12 hours, and 4.2 percent volunteered more than 12 hours permonth. One would expect even more extensive parent voluntary involvement, since 53 percent ofthe parents believed that voluntary work was required at their charter school. Compared to theprevious year, the proportion of parents who believed voluntary work was required dropped from64. Likewise, the level of volunteering in the schools had dropped noticeably. This may be due tothe fact that the schools were better established and required less volunteer work from parents.

73

Distance to Charter School

The average distance from home to the charter school was 4.9 miles (4.3 in 1999), while the averagedistance to the nearest applicable traditional public school was 2.3 miles. Distance to the schoolshould not be a large deterrent for parents, since transportation is supposed to be provided by localschool districts in the same manner as transportation is provided to other schools in the district.Nevertheless, since a number of charter schools are still wrestling with arrangements fortransportation, the considerably longer distance to the charter schools indicates a high level ofcommitment on the part of parents.

6.3 Reasons for Choosing Charter Schools, Responses fromSampled Students and Parents

Parents’ Reasons for Choosing Their Charter School

In the most recent parent surveys, 87.5 percent reported that they were aware of the charter school’smission. Table 6:2 contains the rank-ordered reasons for choosing a charter school provided by oursample of 364 parents in 2000. The parents were asked to rate each factor on a 1 to 5 scale (1 = NotImportant to 5 = Very Important) according to how important they were in choosing their charterschool. The order and relative rating of the responses for the 1999/2000 school year did not differmuch from the previous sampling in 1998/99.

As one can see from these results, some parents chose a charter school because of what they did notlike at their local traditional public school, while others chose a charter school because of what wasbeing promised. A number of state and national charter school studies suggest that the reasonsparents choose a charter school are equal to what exists at the charter school. It is important,however, to distinguish between reasons for choosing and what actually exists at the charter school,especially since most of the parents chose their charter school before it was even open and wouldhave had limited information about the quality of instruction that would be offered. In order todifferentiate between what parents expected and what the school actually provides, a portion of theparent questionnaire was designed to have parents rate their initial expectations and contrast this withwhat they currently perceived or were experiencing at the charter school. These findings arediscussed further in Chapter 14, which deals with process and market accountability.

According to the National Study of Charter Schools (RPP, 1998) parents choose charter schoolsbecause they are dissatisfied with the public schools and/or are attracted to charters. Thepredominant areas of parental dissatisfaction with public schools are: (i) low academic expectations,(ii) poor instructional practices, (iii) environment and culture, (iv) safety, and (v) sense that parentsare not welcome at school. The top six reasons parents were attracted to charter schools include thefollowing: (i) nurturing environment, (ii) safe environment, (iii) value system, (iv) quality ofacademic program, (v) high standards for achievement, and (vi) small class size.

74

Table 6:2 Parents’ Reasons for Choosing Their Charter School, Rank Ordered by Mean ScoresNotimportant

Very important Mean STD Med-

ian1 2 3 4 5

Good teachers and high quality of instruction 1.7% 1.7% 11.0% 16.7% 68.9% 4.49 0.88 5Safety for my child 2.3% 3.1% 10.0% 12.9% 71.7% 4.49 0.96 5Academic reputation (high standards) of this sch’l 1.4% 3.7% 13.0% 24.3% 57.6% 4.33 0.93 5I prefer the emphasis and educationalphilosophy of this school 2.3% 2.3% 12.7% 30.6% 52.0% 4.28 0.93 5Promises made by charter school’sspokespersons

6.6% 4.3% 15.9% 25.9% 47.3% 4.03 1.18 4

My interest in an educational reform effort 6.8% 9.3% 21.4% 22.0% 40.6% 3.80 1.25 4My child wanted to attend this school 16.8% 10.3% 25.9% 14.0% 33.0% 3.36 1.45 3I was unhappy with the curriculum andinstruction at previous school 23.5% 12.2% 14.2% 13.9% 36.2% 3.27 1.61 4My child has special needs that were not metat previous school 25.9% 12.5% 16.1% 10.1% 35.4% 3.17 1.63 3I prefer a private school but could not afford it 28.7% 7.8% 17.8% 15.5% 30.2% 3.11 1.61 3Convenient location 24.2% 11.5% 25.6% 9.9% 28.7% 3.07 1.53 3My child was performing poorly at previousschool 34.8% 10.7% 18.0% 9.9% 26.7% 2.83 1.63 3Recommendations of teacher/official at mychild’s previous school 46.1% 11.0% 15.9% 8.1% 18.8% 2.43 1.57 2

The 6 lowest rated factors, among the 20 options to which the parents could respond (based on amean rating from 0 = feature not applicable, 1= not powerful to 5= very powerful), are as follows:(i) support for home-schooling, (ii) longer school year, (iii) focus on cultural/ethnic needs, (iv)extensive community service, (v) flexible school schedule, and (vi) services for disabled.

The findings from the national study are similar to what we have found in our Pennsylvania parentsurveys. In particular there is agreement with the high ranking of “quality of instruction,” “safety,”and “academic reputation” as factors instrumental in the choosing a charter school.

On our surveys parents were asked to list some other factors that motivated them to enroll theirchildren in a charter school. The responses fell into four broad categories: (1) class size; (2) culturalissues; (3) teacher attitudes and relationships with parents and students; and (4) length of the schoolday and/or year.

Class size was mentioned most often. Parents indicated the desire for smaller class sizes, for greaterindividual attention for students, and for better behavior management. Examples of commentsinclude “Class size afforded my child added individual instruction”; “belief that a controlledenvironment is conducive to the education of our children”; “Discipline One on One”; “Closeknitted, less children. I thought it would be less children in classroom-sizes I thought there wouldbe 18-20 children;” “My child could get private school attention in a public school setting.”

75

Cultural issues focused on teaching of cultural/ethnic customs as well as the inclusion of foreignlanguage instruction. Examples of comments include “Bilingual school education”; “This schooloffers different language so that my son can be bilingual in the future”; “ African American Staff,foreign language”; “Racial/ethnic/cultural diversity”; “The Afro-centric Awareness. Blacks takingthe responsibility to teach their own.”

Teacher attitudes and relationships with parents and students revealed the importance that parentsput on positive communication with school personnel and on knowing that teachers are taking apersonal interest in their children. Examples of comments include “The faculty at the charter schoolhave a better sense of how a child should be treated.(Respect)”; “The children’s education is veryimportant to the teachers”; “The fact that the classes had fewer students allowed more one on onetime from the teachers along with a chance for a teacher-student-parent relationship to be formedin a positive manner”; “This school is concerned about the whole well being of any child”; “Thestaff really care about your children, help to stay in school and enjoy school”; “I believe that the staffcares about my children.”

Length of school day/year was also mentioned as a motivation for choosing a charter schoolparticularly the availability of full-time Kindergarten programs and extended number of school days.Examples of comments include “The length of time they attend throughout the year”; “all daykindergarten.”

Student’s Reasons for Choosing their Charter School

The majority of student respondents indicated that their parents preference was the primary reasonfor choosing the charter school. Nearly 50 percent of the students rated as “Very Important” thattheir parents thought the charter school would be better for them. Following this was the itemdealing with the perceived improved safety at the charter school. Though it was the lowest rankedreason for choosing a charter school, friends have an important impact on parents’ decisions. It isa matter of parent and student trust (Meister & Schuh, 2000).

An analysis of responses was conducted to compare the students in Grades 5-8 and the studentsenrolled at the high school level. The differences on four items were significant. Students at thelower levels were more likely than high school students to rate the following reasons as moreimportant: “My parents thought this school is better for me,” “This school has better computers andother equipment,” and “My friends were attending this school.” Smaller school size was seen to bea more important factor for high school students than for students at the lower levels.

Two survey items dealt with the quality of the previous school the students attended: “Teachers atmy previous school did not help me enough” and “I was not doing very well at the previous school.”Important factors inducing families to choose alternatives to their local public schools are the poorquality of the local school or the lack of appropriate services for students. Nevertheless, these factorswere not highly rated as reasons for choosing the charter school among the sampled students.

76

The two least important reasons for choosing a charter school–convenient location and friends wereattending this school–indicate that the students believed that an active choice was being madebecause of quality and safety rather than convenience or social concerns. The general pattern ofresponses from students regarding reasons for choosing their school was the same in both years thesurveys were administered. Table 6:3 shows the order in which students rated reasons for choosingtheir school.

Table 6:3 Students’ Reasons for Choosing Their Charter School, Rank Ordered by Mean ScoresNotimportant

Very important Mean STD Med-

ian1 2 3 4 5

My parents think this school is better for me 9.3% 6.1% 16.6% 19.4% 48.6% 3.92 1.31 4This school is safer 20.9% 9.7% 21.2% 14.9% 33.4% 3.30 1.53 3We heard that teachers were better inthis school 27.4% 9.8% 20.0% 16.3% 26.5% 3.05 1.55 3

Teachers at previous school did not helpme enough 30.1% 10.9% 18.3% 16.1% 24.6% 2.94 1.57 3

This school has small classes 36.3% 10.9% 17.8% 12.8% 22.2% 2.74 1.58 3This school has better computers &other equipment 37.1% 11.2% 17.3% 10.9% 23.5% 2.72 1.60 3

This school is smaller 40.6% 9.7% 15.4% 10.8% 23.5% 2.67 1.63 2I was not doing very well at the previous school 40.2% 10.4% 16.9% 10.9% 21.6% 2.63 1.60 2This school has a convenient location 35.5% 11.6% 26.0% 11.2% 15.8% 2.60 1.46 3My friends were attending this school 57.7% 11.4% 11.9% 7.9% 11.0% 2.03 1.42 1.

When students were asked to list some other factors that motivated them or their families to choosea charter school, the responses tended to fall into four broad categories: (1) learning environment,(2) personal relationship to school, (3) prior discipline or academic difficulties, and (4) specificcurriculum offerings.

Learning environments include smaller class sizes, school structure, and overall school climate.Examples of comments include “I needed a smaller environment”; “Friends were here, goodteachers”;”That it was smaller and I can work better”; “Uniforms and discipline motivated myfamily.”

Personal relationship to school includes family member(s) or friend(s) attending, family memberon staff, and school close to extended family/caregiver. Examples of comments include “my motherbecame the Spanish teacher here”; “Friends were here”; “that my Mom work here”; “My sister wenthere and she liked it.”

Prior discipline or academic difficulties include behavior and learning problems experienced bystudents at their previous school(s). Examples of comments include “The teachers and I hadconsistent conflicts”; “To get more help at this school”; “no other school would accept me”;”I got

77

expelled so I got put here”; “ I was put here due to permanent expulsion;” “. . .County Courts arevery persuasive.”

Specific curriculum offerings refers to any specialized curriculum in academic, artistic,vocational, or athletic area. Examples of comments include “I am interested in law so we chose thisschool;” “The Economics part because I want to own my own business;” “Get more credits than apublic school to get into college;” “Because I can learn more Spanish and my culture;” “You get yourdiploma in 10 months and you get paid.”

6.4 Awareness of School Mission

Students are aware of their schools’ missions. The mean score was 3.52 on a 5 point scale (1 waslow and 5 was high) with a standard deviation of 1.41 (n = 934). About 20 percent disagreed orstrongly disagreed that they were aware of their school’s mission; about 20 percent neither agreednor disagreed; and about 60 percent agreed or strongly agreed. The following year, students whoagreed or strongly agreed dropped 3 percent; those who neither agreed nor disagreed rose about 1percent; and those who disagreed or strongly disagreed rose about 2 percent. The 1999-00 parentdata have not yet been entered.

Nearly nine-tenths of the parents reported they were aware of their school’s mission. Just over 76percent of parents responding felt their school followed the mission well or very well. A number wassurprising, since nearly the same amount of parents do not participate at all or very little in theirschool. Last year results were similar: just over nine-tenths of the parents were aware of the school’smission, and about 80 percent of parents thought their school followed the mission well or very well.The 1999 data also reflected a smaller group that did not volunteer at all or very little.

6.5 Summary

In this chapter, we have reported descriptive data about charter school students and parents. Muchof the data was collected from our two rounds of surveys, in May 1999 and then a second timeduring the spring of 2000. Supplemental data that PDE collects was also used to provide comparisondata with noncharter public schools through the state as well as in the host districts.

The age and grade level of students sampled was fairly evenly distributed from Grades 5-12, as werethe grade levels of the children of the parents being sampled (Grades K-12). Minorities wereobviously highly represented in Pennsylvania’s charter schools, with all but 8 charter schoolsenrolling a higher proportion of minorities than their host districts. Charter school students werefairly evenly distributed by gender; however, there were considerable differences among schools,some with high proportions of either males or females, depending on the nature and profile of theschool.

78

The number of low income students was also compared to the host district schools. The proportionsare rather even although there is a considerable amount of variation among charter schools as wellas among host districts.

When surveyed on the level of education they planned to achieve, many students were not yet sure.However, of those who planned to continue their education after high school, most aspire tocomplete a 4-year degree or graduate level study. The length of formal education of parents wasalso surveyed, and this varied considerably depending on the location and profile of the school.

Eighty percent of the students surveyed had previously attended a traditional public school. Just over17 percent had attended a nonpublic school before attending the charter school (i.e., private,parochial or home schooling) and the remainder either did not attend school or attended some othertype of school, or another charter school. While many of the schools had few students moving fromnonpublic schools to the charter school, a few schools really stood out with close to half of theirstudents coming from private or parochial schools. This transfer of students from nonpublic schoolsinto the charter schools provides an extra burden in terms of costs on the sending districts and hasbecome of point of contention between charter schools and districts.

The primary reason parents chose the charter school related to the school’s mission, or to problemsthey were having with the traditional public school. Other issues noted were class size and teachers.Students chose the charter school because of parent preference, friends attending, not doing well intheir previous school, and class size. Also, both parents and students seemed to be very aware of theschool’s mission and what it meant.

79

Chapter SevenTeacher and Staff Characteristics

One of the stated goals of Act 22 is to provide enhanced professional development opportunities forteachers. Moreover, the charter concept makes certain assumptions about the attitudes and behaviorsof teachers and staff. In this chapter we provide a general profile of charter school teachers and staffand in the following chapter we will examine opportunities for professional development of teachersand staff. Questions that will be addressed in this chapter include the following:

! How many teachers and staff do charter schools employ?

! What are the demographic characteristics of charter school teachers (gender, race/ethnicity, age)?

! What proportion of teachers and staff are devoted to instruction? What role do other staff play?

! How much and what kinds of experience/education and training do charter school teachers have?How does this compare with other Pennsylvania teachers?

! What proportion of teachers are certified to teach in their area? What are the credentials of thosewho do not hold PA teaching licenses?

! Why do teachers/staff choose to join a charter school?

! Are teachers/staff aware of the school�s mission?

7.1 Description of Charter School Teachers and Staff

There are two main sources of data for the findings reported in this section. One is from thequestionnaires we collected from a sample of teachers and staff in May 1999 and again in the springof 2000. The other main source is data collected and reported by the Pennsylvania Department ofEducation. The source of the data and information is included when reporting findings.

Survey Sampling of Teachers and Staff

In sampling teachers and staff, we included all instructional staff and key administrators at each ofthe 30 schools participating in our study. Thirty-one schools were open at the onset of the study inMay 1999. Since that time, 1 school closed its doors, leaving 30 schools that we visited andadministered surveys. An additional 17 schools opened their doors in the autumn of 1999. The onlydata we have from these schools was provided by PDE.

In our most recent survey, a total of 537 teachers and staff completed and returned surveys from the649 that were targeted (82.7 percent response rate). All schools took part in this component of the

80

Distribution by Race/Ethnicity in 1999-2000

0%

20%

40%

60%

80%

100%

Asian/PacificIslander

Black Hispanic Native Amer-ican Indian

White

Only Sampled Teachers (N =269)

All Sampled Teachers & Staff (N=518)

All teachers in 30 participating schools (N=392)

All teachers in 47 schools with data (N=664)

Figure 7:1 Distribution of Charter School Teachers and Staff by Ethnicity

evaluation, and all schools are included in the analysis. One school had a response rate just under40 percent; but 8 schools had a 100 percent response rate, 12 schools had response rates between 80percent and 99 percent, and the remaining schools had response rates between 60 and 80 percent.The overall response rate was 83 percent, which was 10 percent higher than in the 1999 sample. Theitem response for each survey question was typically 95 percent or higher. Appendix B includesspecific information about the achieved sample.

At the end of the 1998-99 school year, the sample included 447 teachers and staff from a total of609 that were targeted. The overall response rate was 73.4 percent. Given the large number ofsurveys and given that all charter schools were sampled, we think these two samples can provide arepresentative picture of the teachers and staff at the charter schools.

Gender

In terms of gender differences, 69 percent of the teachers and staff were female, and 31 percent weremale. Among the principals/directors, there was an even balance between males and females. Therewas little change in the distribution of teachers and staff between the two years that surveys wereconducted. Female teachers and staff are still the majority in charter schools, just as they are in otherpublic schools.

Race/Ethnicity

From the data we collected from the charter school survey, we determined that 54.8 percent ofteachers/staff were white, 34.6 percent African American, 6.8 percent Hispanic, 2.1 percent Asianor Pacific Islanders, and 1.7 percent Native American. In comparison with the previous sample

81

(1998-99 school year), there are slightly increased percentages for white staff as well as Asian andNative American staff. The proportion of African-American and Hispanic staff dropped a fewpercentage points since the previous year, however. Figure 7:1 illustrates the relative distributionof staff by race/ethnicity in 1999/00. Separate bars illustrate the distribution for sampled teachers,all sampled staff, PDE data collected from 30 participating schools, plus PDE data collected fromthe 47 charter schools that were in operation during the 1999-00 school year.

Age

The age distribution among the Pennsylvania charter school teachers indicates that they are youngerthan teachers in the traditional public schools. Among classroom teachers in 1999/00 (n=277), 52percent were in their 20s, 26 percent were in their 30s, 16 percent were in their 40s, and 5.4 percentwere 50 or older. The classroom teachers were the youngest among the various groups of staff,while the principals/directors were considerably older.

Table 7:1 Age Distribution of Charter School Teachers Compared with National DistributionAge Group Pennsylvania

Charter School Teachers Percentage of National Public School Teachers

1998/99 1999/00 1993

under 30 50.5% 52.3% 11%

30-49 43.5% 42.3% 64%

older than 49 6.1% 5.4% 14%

In comparison with the 1998-1999 school year, charter school teachers in the 1999/2000 school yearare slightly younger. It was hard to find comparison data for regular public schools in Pennsylvania,but comparisons with national data indicate that the charter school teachers are younger than theirregular public school counterparts (see Table 7:1).

Role and Proportion of Staff Devoted to Instruction

Among the 537 teachers and staff sampled in May 2000, 52.3 percent indicated that they wereteachers, 11.6 percent teaching assistants, and 7.3 percent specialists. Approximately 8.2 percentindicated that they were Chief Administrative Officers (CAOs), principals, or school directors, and21 percent indicated that they had some other title or position. The number of key administratorsincreased greatly from the previous year. In part this is due to a better follow-up on the surveys.This is also likely to be due to employment of additional key administrators at a number of the largerschools during the past year to help with the operation of the school. It is important to point out thatmany of the teachers and staff have more than one role at their schools.

In comparison with data from 1998-1999, the percentage of teachers, specialists, and principals inthe 30 charter schools increased 3.9 percent, 2.2 percent, and 3.7 percent, respectively, in 1999-2000.

82

Table 7:2 Percentage of Classroom Teachers in Pennsylvania Charter Schools Comparedwith Regular Public Schools

Percentage of classroomteachers in Pennsylvania

charter schools

Percentage of classroomteachers in

Pennsylvaniapublic schools

1999 2000 1999

48.4% 52.3% 49.2%

Distribution of Teachers and Other Staff by Grade,All Charter Schools in 1990-2000

0%

5%

10%

15%

20%

25%

K 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 10th 11th 12th

Teachers 0ther staff

Figure 7:2 Distribution of Teachers and Other Staff by Grade, 1999-2000 Note: 23 teachers and 116 other staff indicated that grade level was not applicable for their position. The

large proportion of �other staff� at Grade 12 is due to the large number of non teaching staff atYouth Build, which only caters to 12th grade students. This school accounted for 22 of the 23 staffcompleting a survey in this category.

According to 1998-1999 state-level data,classroom teachers comprised 49.2 percent ofall public school staff in Pennsylvania.According to 1996-1997 national data,classroom teachers comprised 52.4 percent ofall public school staff in the United States. Table 7:2 illustrates the change in charterschools over the last two years and comparesthis with the proportion of classroom teachersin all Pennsylvania public schools. One cansee that charter schools have a higherproportion of classroom teachers in the mostrecent year; however, this estimate is likely to be high since our sample did not include custodial,maintenance, and cafeteria staff in the larger schools.

Distribution of Teachers and Staff by Grade Level

Teachers and staff were asked to indicate which grade they work with most. Teachers and staffappear to be somewhat evenly distributed by school level (i.e., elementary, middle, or high school),but larger differences exist by particular grade levels. Figure 7:2 illustrates the distribution of all

83

teachers and staff by grade level as well as the distribution of teachers only across the various gradelevels. Just over 26 percent of the staff indicated that the grade level with which they were workingwas not applicable because they worked in administration, in the provision of support services, orbecause they worked with a number of grade levels.

7.2 Educational Background and Years of Experience ofPennsylvania Charter School Teachers and Staff

Certification of Teachers

Of the 280 staff who indicated they were teachers in the May 2000 sample, 75 percent reported thatthey are currently certified to teach in Pennsylvania. This is a decrease from 81.8 percent in theprevious year. On the other hand, the percentage of staff who were certified in another stateincreased from 2.3 percent in 1999 to 4.3 percent in 2000. The percentage of staff who were workingto obtain certification also increased from 12.1 percent in 1999 to 17.9 percent in 2000. Furthermore,the percentage of teachers who were not certified and were not working to obtain certification hasdeclined from 3.7 percent to 2.9 percent between 1999 and 2000. This information should beconsidered indicative and not conclusive. For example, among the 50 teachers who reported thatthey are working to obtain certification, many may be working for a second certification. It may alsobe the case that the �teachers� who are working to obtain certification are, in fact, only teachingassistants and did not answer the question on role in school correctly.

Most teachers reported that they were teaching in a subject area in which they are certified to teach,although approximately 11.5 percent of the teachers indicated that they were not certified in thesubjects they taught. This is a slight improvement from the previous year. Since the regulationsrequire only 75 percent of the charter school staff to be certified, one can assume that charter schoolshave more uncertified teachers than traditional public schools. It is important to point out, however,that even though Pennsylvania has an oversupply of teachers, many school districts are have a largenumber of teachers on emergency certification.

Teacher Education

In terms of formal education, the charter school staff appear to be well qualified (see Tables 7:3 and7:4). Among those 275 staff who had completed a university degree, 75.3 percent had a B.A. as theirhighest college degree, 20.7 percent had an M.A., 1.5 percent had a 5- or 6-year certificate, and 2.5percent had a Ph.D. Thirty-three percent of the staff were working toward another degree. For themost part (i.e., 70.3 percent) they were working toward an M.A.

84

Table 7:5 Highest Academic Degree of Pennsylvania Charter School Teachers as Compared to Noncharter School Classroom Teachers

Percentage of Charter School Teachers

Percentage of AllPenn. Classroom

Teachers 1999 2000 2000

Bachelor�s degree 70.8% 75.3% 55.6%

Master�s degree & 5-or 6- year certificate 26.4% 22.2% 43.2%

Doctorate 2.8% 2.5% 0.5% Source of charter school data comes from our surveys, while the noncharter data were obtained from PDE website.

Table 7:3 Role and Amount of Formal Education

RoleDid not

completehigh school

Completedhigh

school

Less than 4years ofcollege

Collegegraduate BA/BS

Graduatecourses, no

degree

Graduate/professional

degreeTeacher 0.0% 0.0% 1.1% 44.1% 30.8% 24.0%Teaching assistant 0.0% 35.5% 32.3% 22.6% 8.1% 1.6%Specialist 0.0% 0.0% 15.4% 33.3% 10.3% 41.0%CAO/Principal 0.0% 6.8% 15.9% 29.5% 9.1% 38.6%Other 1.9% 26.7% 24.8% 10.5% 12.4% 23.8%Total (N=529) 0.4% 10.0% 11.7% 32.9% 21.2% 23.8%

Table 7:4 Role and Highest Academic DegreeRole Bachelors Masters 5- or 6-year cert. Doctorate

Teacher 75.3% 20.7% 1.5% 2.5%Teaching assistant 94.1 5.9% 0.0% 0.0%Specialist 50.0% 41.2% 2.9% 5.9%CAO/Principal 45.7% 31.4% 5.7% 17.1%Other 49.0% 42.9% 4.1% 4.1%Total (N=410) 68.3% 25.4% 2.2% 4.1%

Note: Figures based upon sample of teachers and staff. There were 127 surveys with missing data,because a considerable portion of the teaching assistants and staff in the �other�category didnot complete a B.A. degree.

In comparison with lastschool year (1998-1999),the percentage of teacherswho had a B.A. increasedfrom 71 percent in 1999 to75 percent in 2000.However, the percentagesof teachers who had anM.A. and a 5- or 6-yearcertificate decreased from26.4 percent to 22.2 percentbetween 1999 and 2000.The percentage of teacherswho had a doctorate

85

decreased from 2.8 percent to 2.5 percent between1999 and 2000. Table 7:5 highlights thesechanges between the two samples.

One can also see from this table the classroom teachers in noncharter public school tend to havemore formal training with nearly twice as many of them receiving masters degrees than the charterschool teachers. The charter schools had, however, a larger percent of the charter school teachersobtained a Ph.D. than the noncharter teachers (2.5% vs. 0.5%).

Table 7:6 Years of Experience by Role and in Various Types of School, 1999-00 PrivateSchool

ParochialSchool

CharterSchool

PublicSchool

Total Yrs. ofExperience*

Years atCurrent School

Mean STD Mean STD Mean STD Mean STD Mean STD Mean STDTeacher 0.95 2.54 0.31 1.37 1.69 0.82 1.84 3.28 4.80 4.36 1.72 0.94

Teaching assistant 0.39 1.32 0.19 1.08 1.56 0.74 0.89 1.99 3.03 2.42 1.74 1.21

Specialist 1.10 2.78 0.56 1.43 1.62 0.59 2.79 4.46 6.08 5.44 1.62 0.59

CAOs/directors 2.43 4.41 0.70 2.49 1.98 0.88 5.77 9.64 10.89 10.23 1.98 0.88

Other staff 0.62 2.26 0.05 0.28 2.00 2.24 1.84 3.94 4.50 5.66 2.00 2.24

Total, all teacher/staff (N=536) 0.93 2.62 0.29 1.35 1.76 1.24 2.12 4.41 5.11 5.57 1.78 1.29

* Total years of experience as an educator in the school types/roles listed in the table

Most of the years of experience for the charter school staff was in public schools. About 24 percentof the accrued experience of the charter school teachers and was in private and/or parochial schools.

On average, the teachers had 4.80 years of experience as educators (this does not include relatedexperience outside of the school types noted in Table 7:6). This is an increase from 4.75 years ofexperience reported by teachers in the 1999 sample. Overall, the levels of formal education andamount of working experience of the charter school staff is similar to charter school teachers in otherstates we have studied. In terms of years at current school, we can see that the CAOs and keyadministrators have, on average, spent more years at the charter school than the teachers (1.98 yearsversus 1.69 years, respectively).

Table 7:7 contains data on teachers collected and reported by the Pennsylvania Department ofEducation. The first 30 schools in the table are those that opened in 1997-98 or in 1998-99 that wevisited to collect survey data. The last 17 are the schools that opened in 1999-00 and which we havenot visited for data collection. The indicators include gender, status, certification type,race/ethnicity, education level, and totals for classroom teachers. At the bottom of the table we havetabulated totals and percentages for the 28 schools participating in the study as well as for all 45schools that were in operation during the 1999-00 school year.

Table 7:7 Charter School Classroom Teachers by Gender, Employment Status, Certification Type, Race/Ethnicity, Level of Education, and Totals for 1999-2000 (www.pde.psu.edu/esstats.html)

Gender Status Type Race Level of Education

Charter SchoolMale Female Full-

timePart-time

Ele-men-tary

Second-ary

K-12 orMiddleSchool

SpecialEd.

Speech Am.Indian/ Alaskan

Native

Asian/PacificIslander

Black His-panic

White LessthanBA

BA MA PhD Total

Manchester Academic CS 4 10 14 0 6 5 3 0 0 0 0 7 0 7 0 6 8 0 14Northside Urban Pathways CS 3 9 11 1 4 5 1 2 0 0 1 2 0 9 0 10 2 0 12Urban League of Pittsburgh CS 3 3 6 0 6 0 0 0 0 0 0 3 0 3 0 5 1 0 6Mosaica Academy CS 5 24 29 0 24 2 1 2 0 0 0 0 0 29 0 21 8 0 29Centre Learning Comm. CS 3 1 3 1 3 1 0 0 0 0 0 0 0 4 0 0 3 1 4Nittany Valley CS 0 4 4 0 4 0 0 0 0 0 0 0 0 4 0 4 0 0 4Chester Co. Fam. Acad. CS 1 2 3 0 3 0 0 0 0 0 0 0 0 3 0 3 0 0 3Sylvan Heights Science CS 3 4 7 0 6 0 1 0 0 0 0 1 1 5 0 6 1 0 7Chester Community CS 0 11 11 0 11 0 0 0 0 0 0 3 0 8 0 10 1 0 11Chester Charter School 2 18 20 0 17 2 0 1 0 0 0 1 0 19 0 18 2 0 20Village CS of Chester-Upland 6 8 14 0 9 3 1 1 0 0 0 10 0 4 0 11 2 1 14GECAC Community CS 2 11 13 0 12 0 0 1 0 0 0 3 0 10 0 8 5 0 13Northeast Charter School 2 1 3 0 0 3 0 0 0 0 0 0 0 3 0 1 2 0 3Keystone Educ. Center CS 14 7 21 0 3 14 0 4 0 0 0 0 1 20 0 19 2 0 21Center for Econ. & Law CS 11 5 15 1 0 15 0 1 0 0 0 9 1 6 0 10 5 1 16Eugenio Maria de Hostos CS 3 5 8 0 7 1 0 0 0 0 0 1 6 1 0 6 2 0 8Family Charter School 3 7 10 0 9 0 1 0 0 0 1 4 0 5 0 7 3 0 10Imhotep Institute CS 6 9 14 1 0 14 0 1 0 0 0 14 1 0 0 10 5 0 15Multi-cultural Academy CS 4 1 5 0 0 4 0 1 0 0 0 1 0 4 0 3 2 0 5Phil. Community Acad CS 5 8 13 0 1 9 2 1 0 0 0 1 1 11 0 8 5 0 13Phil. Harambee Inst CS 4 8 12 0 12 0 0 0 0 0 0 12 0 0 0 9 2 1 12Alliance for Progress CS 1 10 11 0 11 0 0 0 0 0 0 6 0 5 0 9 2 0 11The Laboratory School CS 6 19 20 5 21 2 2 0 0 0 1 9 0 15 0 17 8 0 25The Preparatory CS 2 5 7 0 0 7 0 0 0 0 0 1 0 6 0 5 2 0 7West Oak Lane CS 2 19 21 0 17 2 0 2 0 0 0 13 0 8 0 18 2 1 21World Communications CS 12 15 27 0 0 24 2 1 0 0 0 21 0 6 0 20 7 0 27Youth Build Phila CS 5 3 8 0 0 7 0 1 0 0 1 3 0 4 0 1 7 0 8Ridgeview Academy CS 14 39 51 2 7 41 3 2 0 0 0 0 0 53 0 50 3 0 53

Gender Status Type Race Level of Education

Charter SchoolMale Female Full-

timePart-time

Ele-men-tary

Second-ary

K-12 orMiddleSchool

SpecialEd.

Speech Am.Indian/ Alaskan

Native

Asian/PacificIslander

Black His-panic

White LessthanBA

BA MA PhD Total

Career Connections CS 3 2 5 0 0 5 0 0 0 0 0 0 0 5 0 3 2 0 5Thurgood Marshall Acad. CS 5 10 15 0 9 3 3 0 0 0 0 11 0 4 0 11 4 0 15Wonderland CS 0 2 2 0 2 0 0 0 0 0 0 0 0 2 0 2 0 0 2Collegium CS 0 8 8 0 7 0 0 1 0 0 0 0 0 8 0 3 5 0 8Architecture & Design CS 14 6 16 4 0 20 0 0 0 0 0 3 3 14 0 8 10 2 20Christopher Columbus CS 3 18 21 0 20 1 0 0 0 0 2 0 10 9 0 15 6 0 21Freire CS 0 5 5 0 0 5 0 0 0 0 0 1 1 3 0 4 1 0 5Germantown Settlement CS 7 13 20 0 15 5 0 0 0 0 0 14 0 6 1 11 7 1 20Imani Education Circle CS 6 13 18 1 15 4 0 0 0 0 0 18 0 1 1 14 4 0 19Mast Community CS 4 19 23 0 5 17 0 1 0 0 2 1 0 20 0 21 2 0 23Math Civics and Sciences CS 16 24 39 1 24 14 2 0 0 0 1 24 0 15 0 25 14 1 40Philadelphia Academy CS 6 24 30 0 27 1 0 1 1 1 1 1 0 27 0 20 10 0 30Renaissance Advantage CS 6 18 24 0 21 1 1 1 0 0 0 15 0 9 0 21 3 0 24Renaissance CS 9 10 16 3 2 11 5 1 0 0 1 15 1 2 0 12 7 0 19Universal Institute CS 2 11 13 0 13 0 0 0 0 0 0 9 0 4 0 8 5 0 13Young Scholars CS 0 3 3 0 2 1 0 0 0 0 0 1 0 2 0 3 0 0 3Crispus Attucks Youth Build CS 1 4 5 0 0 4 0 1 0 0 0 0 0 5 1 3 1 0 5Total for 30 participatingschools opened in �97 or �98 126 266 381 11 193 161 17 21 0 0 4 125 11 252 0 295 92 5 392

Percentages for 30 schools 32% 67.9% 97.2% 2.8% 49.2% 41.1% 4.3% 5.4% 0.0% 0.0% 1.0% 31.9% 2.8% 64.3% 0.0% 75.3% 23.5% 1.3%

Totals for all 47 charterschools 208 456 644 20 355 253 28 27 1 1 11 238 26 388 3 479 173 9 664

Percentages for 47 schools 31% 68.7% 97.0% 3.0% 53.5% 38.1% 4.2% 4.1% 0.2% 0.2% 1.7% 35.8% 3.9% 58.4% 0.5% 72.1% 26.1% 1.4%

Source: Pennsylvania Department of Education, 2000 (www.pde.psu.edu/esstats.html)

88

7.3 Reasons to Seek Employment at a Charter School

A number of possible reasons for the teachers and staff to seek employment at a charter school werelisted, and the staff were asked to rate each reason on a 5-point scale according to how relevant eachreason was in influencing their decision to seek employment at the charter school. (Table 7:8includes a rank-ordered list of the results on this question.)

Table 7:8 Reasons for Seeking Employment at This School (Rank Ordered According to Means)Notimportant

Very important Mean STD Median N Miss-

ing1 2 3 4 5My interest in being involved in aneducational reform effort 3.1% 3.5% 20.7% 31.5% 41.3% 4.04 1.02 4 518 19

Opportunity to work with like-minded educators 2.9% 3.5% 20.4% 33.1% 40.1% 4.04 1.00 4 514 23

This school has small class sizes 3.5% 3.7% 21.0% 31.8% 40.0% 4.01 1.04 4 518 19Safety at school 6.3% 6.3% 19.4% 24.4% 43.5% 3.92 1.20 4 520 17Academic reputation (highstandards) of this school 7.8% 7.6% 28.3% 28.9% 27.3% 3.60 1.19 4 512 25

Promises made by charter school�sspokespersons 10.4% 9.8% 26.5% 27.9% 25.4% 3.48 1.26 4 520 17

More emphasis on academics asopposed to extracurricular activities 5.8% 10.4% 33.4% 29.9% 20.5% 3.49 1.10 4 521 16

Parents are committed 13.4% 10.3% 25.5% 24.1% 26.6% 3.40 1.34 4 522 15Convenient location 18.9% 13.0% 28.0% 18.7% 21.4% 3.11 1.39 3 529 8Difficult to find other positions 36.8% 15.0% 23.7% 13.0% 11.5% 2.47 1.39 2 506 31

Two intrinsic factors were among the most important factors for seeking employment in charterschools were: the interest in an educational reform effort and the opportunity to work with like-minded educators. Other factors that influence teachers/staff to join a charter school are safety atschool, academic reputation, promises made by charter school�s spokespersons, more emphasis onacademics, committed parents, and convenient location. The least important factor rated byteachers/staff was difficulty in finding other positions. Nevertheless, just over 24.5 percent of theteachers agreed or strongly agreed that �difficulty in finding other positions� was an important factorin seeking their position.

Many factors that were important in influencing decisions to seek employment at the charter schoolswere related to a better working environment; for example, working with small classes, safety atschool, and high academic standards.

It is interesting to note that the rank order and relative strength of these factors was largelyunchanged between 1998-99 and 1999-00. Two factors, however, dropped in terms of theirimportance in choosing to work at a charter school. One was promises made by the charter school�sspokespersons, and the other was the academic reputation of the school.

89

7.4 School Mission

Charter schools provide choices for parents. Additionally, charter schools allow teachers to chooselearning communities in which to work that match their interests and skills. Note that the choicepremise of the charter concept assumes that teachers choose schools according to mission and thatthis, in turn, makes them more likely to work harder for student outcomes. In this section, we shallexplore 3 general questions related to school mission, including: (i) how familiar are teachers andstaff with the mission of their school? (ii) do teachers and staff believe the mission of their schoolis being met? and (iii) are charter schools able to fulfill their mission?

All but 16 staff members (3.1 percent) indicated that they were aware of the school�s mission. Ofthose who indicated they were familiar with the mission of the school, 27.6 percent thought themission was being followed �very well,� while 44.8 percent thought it was being followed �well,�23.2 percent �fair,� and 4.4 percent �not very well.� These figures indicate that most teachers andstaff thought their school was living up to its mission.

In comparison with last school year (1998-99), teachers and staff in 1999-00 were slightly moreaware of their school�s mission. Likewise, a slightly larger proportion of the teachers and staffthought the mission of their school was being followed reasonably well (i.e., the percentage ofteachers and staff who thought the mission was being followed �very well� and �well� increasedfrom 70.4 percent in 1999 to 72.4 percent in 2000). As in the previous year, these figures indicatea general satisfaction among the teachers and staff in terms of their school�s ability to live up to itsmission.

In another section of the questionnaire, the staff were asked to rate their level of satisfaction with anumber of aspects of the school, including school mission statement. Here, 32.5 percent of the staffindicated that they were �very satisfied� with the mission of their school, while another 37.7 percentindicated that they were �satisfied� with it. While the teachers and staff were generally quitesatisfied with the schools� missions, they were not equally convinced that the schools could fulfillthem. Nearly 14.2 percent of the staff indicated that they were dissatisfied or very dissatisfied withtheir school�s ability to fulfill its mission, while 29.4 percent were uncertain. Still, 37.7 percent ofthe staff indicated that their school could fulfill its mission, and 18.7 percent were very convincedthat their school could do this.

Table 7:9 Level of Satisfaction with the Mission of the School (N=507)Verydissatisfied Very

satisfied Mean STD

1 2 3 4 5School mission statement 2.8% 3.2% 23.9% 37.7% 32.5% 3.94 0.97

Ability of school to fulfillits stated mission 6.5% 7.7% 29.5% 37.7% 18.7% 3.54 1.08

1 Wilcoxon signed ranks test was used to analyze the difference between these two variables. Thisnonparametric procedure tests the hypothesis that the two related variables have the same distribution. It makesno assumptions about the shapes of the distributions of the two variables.

90

When comparing the two items in Table 7:9, note a significant difference in level of satisfaction interms of the school�s ability to fulfill its mission (Z= -10.349, p=0.00).1 Hence, there is a significantdifference between the �ideal school� represented by the school mission and the �actual school�represented by the perceived ability of the school to fulfill its stated mission. The difference betweenthese two items increased slightly from last year. Over time, it will be interesting to continue to tracethe difference between these two indicators. Since the schools are all relatively new, and all but 4schools are in their second year of operation, these results are not surprising. After three or moreyears of operation, however, one should expect that the staff�s belief in their school�s ability to fulfillits mission would increase substantially.

7.5 SummaryThis chapter contained descriptive information about teachers and staff in Pennsylvania charterschools. It was found that a majority of the teachers in charter schools are female, while principalsand directors were evenly distributed between males and females. This generally measures up to thefigures for traditional public schools in Pennsylvania. Also, in comparison with traditional publicschools, the teachers of charter schools are generally younger. It was found that the majority ofteachers in charter schools are white, with African-American teachers second. There are fewHispanic, Asian/Pacific Islander, or Native American teachers in Pennsylvania charter schools.

The survey also sought to identify the roles of staff and the proportion of those directly involved ininstruction. The survey found that there is a slightly higher percentage of teachers on staff in thecharter schools than in the traditional schools. Many of the principals and directors noted that theyalso hold an additional title/role. The teachers are distributed relatively equally by level ofelementary, middle, and high school, but the distribution among certain grade levels is not as equal.

Background information and data on staff experience were also collected. The percentage ofteachers certified in Pennsylvania dropped from the previous year�s study, but the percentage ofteachers certified in other states rose slightly. There was also an increase in the percentage ofteachers working toward certification from the previous year. The vast majority of the teachers withuniversity degrees had attained a B.A. as their highest level of education. However, a third of theteachers stated they are working toward another degree. The average years of experience levelamong surveyed Pennsylvania charter school teachers was just under five years.

When teachers were asked why they chose to seek employment in the charter school, the responsethat was highest rated was �to be part of an education reform movement.� Teachers also sought towork with like-minded educators and to work in a safe environment. The least popular reason givenwas that a teacher had difficulty finding work elsewhere.

Teachers were also asked about the mission statement of their school. The teachers seemed to bequite familiar with their school�s mission statement. However, they were not as confident in theschool�s ability to fulfill the mission.

91

Chapter EightWorking Conditions, Professional

Development, and Levels of Satisfaction for Charter School Teachers

A common notion about charter schools is that they provide an opportunity for professionals tochoose a school that matches their interests. Additionally, charter schools are commonly expectedto allow educators an opportunity to innovate and at the same time be held accountable for theirwork. It is also expected that charter schools will provide new professional developmentopportunities to teachers. We will explore these topics in this chapter as well as describe theworking conditions of teachers and their levels of satisfaction with their schools and the particularconditions under which they work.

The questions that are addressed in this chapter include the following:

� What are the working conditions of charter school teachers and staff, and how satisfied are theteachers with these conditions?

� What are the initial expectations of teachers, and how do these compare with their currentexperiences?

� What kind of teacher induction plans do charter schools have for new teachers?

� How much and what kind of professional development time/opportunities do charter schoolteachers have?

� Do charter school teachers have opportunities to work collaboratively, select texts, designcourses?

� Do teachers think they have enough time to develop innovative instructional practices?

� How much teacher/staff turnover is there in charter schools? Does this vary with salaries, scopeof professional development opportunities, and other factors?

� Are there any innovative practices in the area of professional development that other schoolsmight emulate?

92

8.1 Working Conditions for Teachers and Staff and Levels ofSatisfaction

School Facilities and Available Resources

The quality of school facilities varied extensively among the charter schools. Therefore, it was notsurprising to see an even split in the responses from teachers and staff concerning the quality of theirschool’s facilities. Approximately 30 percent of the staff were satisfied or very satisfied with theschool buildings and facilities. On a related item, 26.3 percent of the teachers and staff agreed orstrongly agreed that the physical facilities were good, while the rest were either not satisfied with thefacilities or were uncertain.

Survey results indicate that the schools vary widely in the quality of their facilities and theavailability of resources. This was also confirmed in site visits and interviews. Just over 35 percentof the teachers and staff indicated that they thought their school had sufficient financial resources.On a related item, 46.9 percent of the teachers and staff indicated that they were satisfied with theresources available for instruction.

A number of items in the questionnaire addressed class size. It was clear that this was an importantreason for seeking employment at a charter school and an aspect of the schools with which theteachers were particularly interested but not yet necessarily satisfied.

While a majority of staff indicated they were not insecure about their future at their particular school,35 percent of the teachers and staff indicated otherwise. It is not clear if this insecurity is due touncertainty about the charter school reform or due to the role of the particular school in itscommunity and its ability to live up to its mission. Similarly, 22.6 percent of the teachers and staffindicated that they did not plan/hope to be teaching in that particular school next year, as comparedwith 77.4 percent who intended to return. Unlike findings in Michigan and Connecticut, mostteachers and staff reported that they did not have many noninstructional duties in addition to theirteaching load.

Autonomy of Pennsylvania Charter School Teachers

We found that–on the whole–the teachers indicated that they have autonomy and can use their ideasand creativity in designing the curriculum at their schools. While we did not have a clear responseto this from many teachers, we found that a number of the schools were exemplary in regard to thisissue. Some of our findings particular to individual schools are included in the following examples:

� At one of the middle schools, we learned that the day-to-day operations and any decisions relatedto curriculum and instruction are the responsibility of the staff.

� At one elementary school, all of the surveys from teachers indicate that they are autonomous andcreative in their classrooms; at a few other schools, many of the teachers indicated likewise.

93

� At one high school, the teachers are asked to experiment with specific techniques. Responsesfrom teachers in interviews or the open-ended questions on the survey confirmed this was thecase. To sum up what the teachers told us, we include the following comment: “I have beenable to design and implement curriculum.”

� In terms of professional development, the teachers at one school are encouraged to explore areasthat are relevant to them and the community they serve.

� Teachers at an urban school noted that they design the curriculum based on history, traditions,customs, and culture of African Americans.

� At one school, the teachers told us they were able to be innovative, but only after approval fromthe chief administrative officer (CAO).

� At one of the Philadelphia schools, the lead teachers are responsible for coordinating thedepartment, but all teachers are encouraged to design and create new strategies and programs;

In the annual reports submitted by the schools, we found evidence that around 35 percent of theteachers are heavily involved in the development of the curriculum. In an additional 15 percent ofthe schools, we found that the teachers were autonomous in terms of curriculum- and instruction-related matters. At a few schools there was clear evidence that decisions about curriculum andinstruction were in the hands of the CAO, another administrator, or the board.

We found evidence at a number of charter schools that there was a conscious effort to involveteachers in developing curriculum. Examples of this include the scheduling of weekly meetings forteachers to meet with a curriculum and technology specialist and the use of teachers’ meetings toestablish goals and objectives for the year.

Collaborative Work of Charter School Teachers

Forty-five percent of teachers did not comment about a collaborative working environment. Forty-two percent of teachers indicated that they have opportunities to work collaboratively. Someteachers reported that they have a team-teaching plan that pairs a teacher who has strengths in aspecific area with a teacher who has strengths in another area. Others said that teachers arecontinually discussing new trends in education. Several indicated that they have mentors andcontinue to learn from one another. One teacher expressed that they work “within nations” andcollaborate in this manner. Thirteen percent of teachers indicated that they do not workcollaboratively. Several teachers indicated that teachers only work collaboratively at monthly staffmeetings.

We found clear evidence in slightly more than half of the charter schools that the teachers in theschools work collaboratively. At many of these schools, the teachers were encouraged to workcollaboratively. In 70 percent of the schools from which we collected information regarding teachercollaboration, we found that collaboration was promoted through mentoring and team teaching.

94

Although this is not innovative to public schools, we think it is a positive program and that it seemedto be implemented well.

� A few schools pair teachers who complement each others’ skills and abilities.

� At one school the teacher mentoring program, staff workshops, and presentations about what isbeing learned are used to encourage collaboration.

� We were told at another school that they use the mentoring program to promote learning fromone another as well as for continuous evaluation.

� One school reported that it intended to hire more experienced teachers in the future to mentorthe newer teachers.

� In the responses to the open-ended questions, a number of teachers noted that teamwork withother teachers and staff was the greatest strength of the school.

� One school organized “nations” within the school to promote teacher collaboration.

It is important to point out that while collaboration of teachers was being encouraged and wasworking in many schools, in a few instances teachers indicated that there was little collaborationaside from occasional or monthly staff meetings.

Time for the Development of Innovative Instructional Practices

The teacher survey asked about teachers’ initial expectations and compared these with what teachersare currently experiencing in their schools. They were asked whether their schools’ support/aresupporting innovative practices and whether they will be/are autonomous and creative in theirclassrooms. As indicated in Table 8:1, there is a 15 percent discrepancy between expectation andcurrent experience in the area of innovations and a 13 percent discrepancy between their expectationand current experience in the area of autonomy and creativity.

Table 8:1 Teacher Expectations and Current Experience With Regard to Innovative Practicesand Autonomy

Initial Expectation Current Experience False Partly

trueTrue Mean STD False Partly

trueTrue Mean STD

The school will support/is supportinginnovative practices 1.1% 25.2% 73.8% 2.73 0.47 6.8% 34.2% 59.0% 2.52 0.62

Teachers will be/are autonomous andcreative in their classrooms 1.7% 19.9% 78.4% 2.77 0.46 3.6% 31.5% 64.8% 2.61 0.56

Teachers submitted a variety of responses in terms of their autonomy. A large proportion reportedthat they are autonomous and creative in their classrooms. Others expressed that they areempowered in decisions related to curriculum, instruction, and day-to-day operation of the school.Several said that their working conditions are very flexible compared with work in previous schools.

1 This estimate was derived by estimating an ordinary least squares regression model on teacher salarydata for all noncharter school districts in Pennsylvania. The model regressed average teacher salary at the districtlevel against variables representing teacher education and experience, along with per pupil expenditures. Togenerate the predicted value for charter schools, we enter mean values on each of the predictor variables in orderto generate the predicted value for teacher salary conditional upon these mean values. The analysis is based on allPennsylvania public schools and uses 1998-99 salary values. Readers are invited to contact the authors for detailson this analysis.

2 Readers are invited to contact the authors regarding details on the regression analysis.

95

Less common but still prevalent in some teachers’ perceptions were that they were now aware ofinnovative practices or opportunities to be innovative at their schools.

Teachers’ Salaries

Charter school teachers had average annual salaries of $30,048.1 This can be contrasted with annualteacher salaries of $48,457 in the state. The teachers at charter schools have fewer years ofexperience than their counterparts in traditional public schools; therefore, their salaries are lower.We used a regression model to estimate the relationship between teacher salaries and a number ofpredictor variables including educational level, average years of experience, and per pupilexpenditures at the noncharter public schools. With this analysis we were able to predict that charterschool teachers should be earning, on average, $39,831a year.2 In other words, comparablenoncharter public school teachers with these characteristics would have an average annual teachersalary of $39,831.

From this analysis we can still see that charter school teachers have salaries that are substantiallylower than what one would expect, and what these charter school teachers might receive in regularpublic schools. These differences can be explained in part by the charter schools’ need to divertresources to purchase or renovate facilities. Tentative findings reported in the chapter on financehighlight that charter schools are in fact devoting a lower proportion of their expenditures to capitalinvestments than traditional public schools.

Just under 30 percent of the teachers and staff were satisfied or very satisfied with the salaries theyreceived, while 32 percent were either dissatisfied or very dissatisfied with their salaries. A largeproportion of the staff (38 percent) indicated that they were neither satisfied nor dissatisfied withtheir salaries. The teachers and staff were generally more satisfied with the fringe benefits than withsalary. The level of satisfaction with salaries decreased noticeably from the previous year, whileteachers’ satisfaction with fringe benefits increased slightly between 1999 and 2000.

Later in this chapter, the rate of attrition among teachers is described. One important factor behindthe large proportion of teachers leaving charter schools is likely to be due to the higher salaries theycan receive in regular public schools.

Table 8:2 includes data on how the teachers and staff at Pennsylvania charter schools rated theirlevels of satisfaction with various aspects of their current working conditions.

3 Because these questions are actually nonparametric in nature and the variables are ordinal, the marginalhomogeneity test was used to compare the paired distribution of responses. This also found significant reductionsin expectations on all items (p = .001).

96

Table 8:2 Levels of Teacher and Staff Satisfaction with Working ConditionsNot verysatisfied Very

satisfied Mean STD Median N Don'tknow

Miss-ing1 2 3 4 5

Salary level 12.8% 19.2% 38.6% 22.7% 6.8% 2.91 1.09 3.00 516 8 13Fringe benefits 7.0% 13.8% 30.9% 31.1% 17.1% 3.38 1.13 3.00 485 34 18Resources available forinstruction 7.5% 17.2% 28.5% 26.9% 19.8% 3.34 1.19 3.00 494 22 21

School buildings andfacilities 13.9% 23.4% 31.3% 20.7% 10.8% 2.91 1.19 3.00 518 2 17

Availability ofcomputers and othertechnology

7.9% 14.5% 20.5% 24.8% 32.3% 3.59 1.29 4.00 517 7 13

School governance 9.9% 10.8% 32.5% 28.4% 18.5% 3.35 1.19 3.00 493 22 22Administrativeleadership of school 7.0% 11.4% 24.0% 28.1% 29.5% 3.62 1.21 4.00 516 5 16

Evaluation or assessmentof your performance 5.7% 9.5% 22.5% 37.0% 25.4% 3.67 1.12 4.00 476 41 20

Overall, 70.2 percent of teachers/staff were very or somewhat satisfied with their school’s missionstatement, and 56.4 percent were very or somewhat satisfied with the ability of the school to fulfillits stated mission. Section 7.4 contains more details about teachers’ satisfaction with their school’smission. Just over 62 percent of the teachers and staff indicated they were very or somewhatsatisfied with evaluation of their performance; 57.6 percent were very or somewhat satisfied withschool leadership; and 57.1 percent were very or somewhat satisfied with availability of computersand technology. Teachers/staff were least satisfied with school buildings and facilities (31.5 percentwere satisfied with facilities).

8.2 Initial Expectations and Current Experiences of Teachersand Staff

A number of identical items were used in the surveys to examine and compare the charter schoolstaffs’ “initial expectations” as opposed to “current experience” (See Appendix B, Teacher/StaffResults, Question 16). In general, it is clear that the teachers and other staff were content with theirschools and satisfied with the services they provide. It is interested to note, however, that there werestatistically significant differences between what was initially expected and what the educators werecurrently experiencing on all variables. What the staff were reporting as “current experience” wassignificantly less positive than their “initial expectations.”3

97

The biggest differences between initial expectations and current experience were on the followingitems: 1. The school will have/has effective leadership and administration.2. Parents will be/are able to influence the direction and activities at the school. 3. Support services will be/are available to students.4. Students will receive/receive sufficient individual attention.5. There will be/is good communication between the school and parents/guardians

This does not imply that teachers and staff were not satisfied with these aspects of their school.Rather, it infers that they had high expectations in these areas that did not correspond with what theywere currently experiencing.

While these findings are rather striking, it is important to consider their educational significance.Likewise, it is important to consider likely explanations for these findings. Given the feedback wereceived from teachers and staff, it seems that teachers simply expected too much. A large portionof the teachers were seeking jobs at schools that were not yet in operation. Given such a situation,expectations are understandably high. Since many of the teachers are also very young, theirexpectations may be higher than normal. Unfortunately, we do not have comparable data fromregular public schools.

Between the 1998-99 survey and the 1999-00 survey there were many similarities and somedifferences. For instance, on the items dealing with initial expectations and current experience werenearly the same. In a number of items, the difference actually decreased between the two years,which suggests areas where improvements were being made, even if the initial expectations werestill not met. These areas of improvement are listed below and ranked according to improvements.1. Students will have/have access to computers and other new technologies2. School personnel will be/are accountable for the achievement/performance of students3. Teachers will be able to influence the steering and direction of the school4. The school will support/is supporting innovative practices

Ratings on three items actually decreased, which suggests areas where things are getting worse.These are listed below:1. Teachers will be/are autonomous and creative in their classrooms.2. The school will have/has small class sizes.3. The quality of instruction will be/is high.

It is interesting to note that there are discrepancies in the factors that influence teachers/staff to joina charter school. Teachers/staff were asked about their initial expectations and current experiencein the quality of instruction and teachers’ empowerment. There was a 21.4 percent differencebetween their expectations (74.1 percent) and current experience (52.7 percent) that the quality of

4 All but 5 charter schools provided information on their professional development plans in the 1999-00school year.

98

instruction will be/is high. In terms of teacher empowerment, there is a large difference betweenteachers’ expectations and current experience that they are able to influence the steering anddirection of the school.

The gap between teachers’ expectations and their current experiences is a warning sign for charterschools. Although there are differences between teachers/staff’s initial expectations and currentexperience, teachers/staff generally are still positive about their schools.

8.3 Professional Development in Pennsylvania Charter Schools

Our main source of information on professional development plans was annual reports submittedby the schools in August 1999 and August 2000.4 We also collected information on professionaldevelopment opportunities during interviews with charter school staff. Teacher/staff surveysincluded still other items regarding professional development. The annual reports included thenumber of days or hours devoted to professional development as well as some details about thenature of the training. In addition to this information we used data on professional developmentreported by the PDE.

One widely used, though imperfect measure of a school’s commitment to professional developmentis the number of days its teachers are engaged in various in-service activities. PDE collects data forall schools on the number of teacher absence days for professional development activities. Theobvious limitation to this indicator is that it fails to capture professional development activities thatdo not require teacher absences. Thus, it excludes such activities as teachers reading books andengaging in discussions on their own; teachers taking evening, weekend, or summer courses; and soon. Also, teacher shortages in many districts are compelling administrators to find professionaldevelopment activities that do not require teacher absences and hence the need for scarce substitutes.Many administrators schedule professional activities during the afternoon hours–hours during whichteachers are normally in the school building.

Schools apparently interpret the survey question differently. Some schools, it appears, read it ascovering only those professional development days enumerated in teacher contracts. Other schoolsappear to interpret the question as covering all such activities, whether called for by contract or not.In spite of these limitations, this indicator provides a useful picture of at least one aspect ofprofessional development in Pennsylvania charter schools.

We calculated the number of professional development days by dividing the number of teacherabsences per school by the total number of teachers. This yields an estimate of the number of daysfor the “typical” teacher. Unfortunately, it does not allow us to observe variations among teachers.Next, we compared the number of professional development days in charter schools and noncharterschools.

5 According to the national School and Staffing Survey (Choy, Chen, & Ross, 1998) conducted by TheNational Center for Education Statistics, public school teachers across the nation take part in a variety ofprofessional development activities. A majority (64 percent) of public school teachers participated in sessionsdealing with methods of teaching in their field, 51.4 percent with student assessment, 50.9 percent with cooperativelearning in the classroom, 47 percent with use of educational technology for instruction, and 30 percent conductedin-depth study in their subject.

99

From this analysis we found that, on average, charter school teachers have 7 days of professionaldevelopment each year, compared with 5 days for noncharter public schools. It is important to pointout, however, that the charter school totals were influenced by 4 schools that reported between 17and 50 days of professional development per year. These outliers clearly weighted the charter schooltotal. At the same time there were apparently 10 charter schools that did not report this data to PDEand were not included in these figures.

It was clear from the documentation and interview data that a number of schools had highexpectations in terms of professional development for their classroom teachers. The charter schoolteachers seemed to have support for professional development opportunities from their schools. Thesupport included release time from teaching, scheduled time built into teacher’s schedules, andtuition reimbursement. One school noted that it had established a professional developmentcommittee that helps to plan and arrange for professional development activities. Graduate levelclasses were emphasized by many as an important and sometimes required form of professionaldevelopment. In one urban school, the teachers are expected to enroll in graduate classes approvedby the Board of Education of the local school district. At another school the teachers develop theirown professional development goals and discuss these with the CAO. At this particular school, theteachers are expected to be taking Ph.D. or M.A. courses and working toward a degree.

The format for professional development opportunities include workshops, conferences, in-servicetraining, outside training, and graduate courses. The content of professional developmentopportunities includes methods of teaching, technology, student assessment, and classroommanagement. For example, some teachers attended conferences dealing with reading programs andcurriculum development. Several teachers indicated that they attended in-service training sessionsdealing with at-risk students, classroom management, and discipline. Others reported that thetraining sessions were focused on software and computers.5

Most of the cited professional development consists of weekly or monthly staff meetings at theschool. Some schools allot only a few hours a month or week for professional development whileothers allot a number of days. More than half of the schools reported that they have professionaldevelopment activities/opportunities in their schools with a duration from 30 minutes to 4 and a halfhours per week. While the schools were not always very specific about how they use these days orhours, it appears that the topics addressed reflect the changing needs of the staff. At the school thatwas closed, we were informed that no money was available for professional development.

Below we have included examples of how the schools described the amount and scheduling of theprofessional development activities at their schools:

100

� Teachers work year-round schedules with schools providing opportunities for attending summerprofessional development programs away from school.

� Teachers attend mandatory professional development activities, monthly workshops.

� Weekly staff meetings with professional development sessions. Also all teachers are encouragedto participate in other learning activities outside of the school

� All school training sessions are hosted by the school plus release time is given for outsideworkshops.

� Ten school days set aside for professional development. Also, 60 minute staff meetings once aweek after school that focus on trends in education and educational research

� At least 2 hours per week

� Monthly staff meetings cover discipline, classroom management, curriculum, conflict and peermediation, assessment, attendance, and grade recording.

� 31 total days set aside for professional development (during August and throughout school year)

� Training for two weeks prior to school, plus six full days during the school year

� Three weeks of staff development before school year along with 8-10 days during the schoolyear, early dismissal days, conferences, workshops, and presentations

� Two hours a week

� 4.5 hours a month, and 3 workshops during the year

� Required participation in two week preservice session, one week session at the end of the year,and a two-hour in-service seminar each week; teachers are also sent to conferences.

� Weekly three-hour meetings

� Monthly half-day staff meetings

� Self improvement through courses; reimbursement for half the cost of approved courses

� Minimum of 12 full days set aside for professional development utilizing outsideagencies/consultant groups to provide the training

� funds are provided for teachers to attend at least two required training sessions in addition tohalf-day in-service during the school year.

� 40 hours of professional training a year in such topics as first-aid/CPR, working with at-riskstudents, etc.

� Teacher meets with the curriculum specialist, plus 1.5 hour monthly staff meetings

� Funds are set aside for staff to attend local, regional, and state conferences.

� Teachers are asked to do a self-assessment of their professional development needs and proposeoptions for growth.

101

8.4 Teacher Induction Plans

Generally speaking, the teacher induction plan is for new teachers and includes training andorientation activities at the beginning of their first year of teaching. Some schools reported that theteacher induction plan covered ongoing training throughout the teachers’ first year.

Many schools conduct training for all their staff, not just new staff, while others have separateorientations, workshops, and mentoring for the new teachers. Two-thirds of the schools providedinformation on their induction plans in the 1999-2000 annual reports. Many of the schools did notprovide very specific details about their induction programs but, coupled with interview data anddocumentation collected from the schools, we were able to assemble an overview of the practicesand trends in terms of teacher induction plans. A handful of schools noted that they held meetingsat the school to cover orientation issues. Most of the schools considered the induction plan to be anorientation to the school and a program to help teachers become more effective and develop theirgoals for the year.

The purpose of the teacher induction activities was most often to familiarize new teachers with theschool and to better prepare them for their work in the classrooms. Topics covered in the orientationor initial meetings at the school included such things as the history of school, school mission andgoals, policies and procedures, personal attributes, meetings with students and staff, overview ofclassroom duties, curriculum development, state academic standards, classroom observation, policiesand procedures training, and strategies to improve teaching performance.

The most common element of the induction plan was mentoring of new teachers. Several teachersindicated that the school assigned a mentor to help them increase their general professionalknowledge, instructional techniques and practices, classroom management, and student assessment.The induction coordinator or mentor was often the CAO, curriculum specialist, director ofinstruction, other experienced teachers, or a teacher education professor from a local college oruniversity.

Some schools required participation in workshops while others provided optional workshops. A fewschools expected the new teachers to develop their own professional plan during the inductionperiod. One school expected the new teachers to work with director of instruction to buildprofessional competence, while at another school the new teachers worked with a curriculumspecialist on a weekly basis and attended other meetings/workshops to help them improveinstruction. Not uncommon was to find that a number of schools required the new teachers to haveextra days/weeks of preparation and training before the start of the school year (this ranged from 15to 96 hours). One school required a month-long training session for new teachers.

One criticism we heard from a number of CAOs is that teacher education programs at the universitiesand colleges are not preparing teachers to work in urban classroom settings. One schooladministrator said that this was very important for their own teachers as well as for teachers in

102

surrounding schools. Therefore, they were exploring options with the district to conduct morelengthy preservice training programs for new teachers with no experience in urban schools.

8.5 Turnover of Teachers and Staff in Pennsylvania CharterSchools

One factor that is limiting the charter schools is the relatively high rates of attrition among teachersand staff. Among the 30 schools participating in our study, nearly 40 percent of the teachers leftduring or between the 1998-99 and 1999-00 school years. During this same time period less than10 percent of the CAOs left or were replaced.

An item on the teacher/staff questionnaire that provided a related indicator of attrition was thequestion, “Do you plan (hope) to teach here next year?” Three-quarters of the staff indicated thatthey wished/intended to return the next year. This was essentially the same proportion that theprevious year reported that they did not intend to return. Based on conversations with charter schoolCAOs and teachers and based on an examination of the attrition figures on a school-by-school basis,it is clear that the level of attrition at some of the charter schools was extremely high, while at otherschools nearly all professional staff were returning.

For those teachers/staff who planned not to return next year (22 percent), their dissatisfaction waswith school governance, administrative leadership, resources available for instruction, ability of theschool to fulfill its stated mission, and evaluation of their performance.

Further study is needed to examine the factors behind teacher attrition. For now, however, it is safeto speculate that the substantially lower salaries that charter schools can pay teachers are animportant factor behind these high levels of attrition.

8.6 Summary

The atmosphere of a school greatly influences a teacher’s job satisfaction. Factors such as facilities,autonomy, and salary are important to teachers. Both site visits and teacher surveys indicated thatthe facilities and resources of charter schools varies widely. Naturally, teacher satisfaction with theseissues varies widely as well. Many teachers have come to charter schools seeking autonomy increating and implementing curriculum. We found evidence at a number of charter schools that therewas a conscious effort to involve teachers in developing curriculum. Teachers indicated that theythought they had autonomy in curriculum decisions and freedom to utilize creative approaches tocurriculum. Indeed, many teachers report that they have considerable flexibility and opportunitiesfor creativity in their day-to-day activities. We found clear evidence in slightly more than half of thecharter schools that the teachers work collaboratively. The teachers are encouraged in theircollaborative work efforts through programs of team teaching, mentoring, and staff memberscreating presentations. While charter school teachers make considerably less than their public schoolcounterparts, not all are dissatisfied with their salary, with some 30 percent reporting they were

103

satisfied or very satisfied. It is worth noting that salaries may be lower due to the need to divertfunding to the renovation and development of facilities.

There was a measurable difference between initial teacher expectations and current experience onmany topics. Statements relating to topics such as the effectiveness of the leadership andadministration, good communication between parents/guardians and school, availability of supportservices to students, parental ability to influence the direction and activity of at the school, andstudents receiving sufficient individual attention had the largest decline from initial expectations tocurrent experience. There is also a large decrease between teachers’ expectations and currentexperience regarding the degree to which they are able to influence the steering and direction of theschool. Other areas that teachers currently felt less positive about were class size, emphasis onacademics, and parental involvement. One must note that there are some explanations for thesefigures including that many teachers were hired before the school opened, some teachers simplyexpected too much, and many were young teachers. Despite these figures, many teacher are stillsatisfied with their teaching environment, and about 75 percent planned to returned to the school thefollowing year.

Charter schools devote considerably more time to teacher professional development activities thannoncharter schools. A strong emphasis on graduate study was frequently reported, with someprograms requiring it. The format for professional development opportunities included workshops,conferences, in-service training, and graduate courses. Some schools allot only a few hours amonth or week for professional development while others allot a number of days. Theseopportunities often take the shape of summer workshops, release time for conferences, and frequentmeetings and workshops within the school year. Most of the cited professional developmentconsisted of weekly or monthly staff meetings at the school. The content of professionaldevelopment opportunities included methods of teaching, technology, student assessment, andclassroom management. A school’s teacher induction program also falls under the category ofprofessional development. Schools generally held meetings to orient new teachers to the school,help them become more effective, and teach them to set goals. The induction plans often includeda mentoring program.

1 See Perrow (1986) and Wilson (1989) for approaches to analyzing organizational tasks and fordesigning decision structures around them.

2 This argument is made in a more general form in studies of management and organizational structure.

3 This is similar to the Tiebout model from public finance theory and public administration. See Loweryet al. (1992) for an accessible discussion of the model and its applications to the delivery of local public services.

104

Chapter NineInnovations in Pennsylvania Charter Schools:

Governance, Curriculum, Instruction

The charter school concept, as we have stated, is predicated on a trade-off between autonomy andaccountability: charter schools will receive more autonomy than other public schools in exchangefor being held more accountable for student outcomes and outcomes related to the mission of theschool. In the previous chapter we explored what charter schools are doing with their autonomy inthe area of teacher professional development and working conditions. In this chapter, we continueour analysis of the uses to which charters put their autonomy by focusing on charter schoolorganization and governance, curriculum, and instruction.

Like other site-based management reforms, the charter concept grows, in part, out of a critique ofthe organizational principles of traditional public schools. The charter school concept makes threeorganizational assumptions about education. The first assumption involves the nature of theeducational task.1 Like advocates of other forms of decentralization (e.g., site-based decisionmaking, magnet schools), charter school advocates argue that education according to �one-size-fits-all� approaches is unlikely to work well. In the language of organizational theory, education on thisview is nonroutine and labor intensive. What works for one group of students, moreover, might notwork well for other groups of students. Accordingly, charter school advocates argue that those�closest to the students� are best qualified to make most important decisions.

The second assumption addresses the motivations of stakeholders. In general, charter schooladvocates think that teachers, administrators, and students (as well as parents) are more likely tomake heavy personal investments in schools if they have a high degree of influence in schooloperations, if they have opportunities for professional development, and if they agree on a clearlyarticulated school mission.2 Here the argument for autonomy dovetails with the argument for choice.Chubb and Moe (1990) and Hill et al. (1997) have argued that stakeholders are more likely to cometo agreement on school mission and methods if they can sort themselves into schools by choice,instead of being compelled to attend schools by virtue of living within a district.3

105

The third premise of the charter school advocates� argument for autonomy is a claim about how todesign organizations around particular tasks. They point out that the traditional hierarchical structureof schools (e.g., clear division of labor, specialization, vertical integration, and standard operatingprocedures) makes sense only if one believes that education is an essentially routine task in whichone can prescribe one-size-fits-all educational interventions for large groups of students. This �massproduction� metaphor, however, conflicts with charter advocates� assessment of the educational taskas nonroutine and highly variable. Accordingly, charter school advocates argue that by allowingcharter schools to be smaller and nimbler, they will be more likely to achieve their goals.

In this chapter we explore the extent to which Pennsylvania charter schools have actualized theseorganizational principles. Throughout, we seek to highlight innovative practices that other publicschools might adopt. The data used in this chapter come from two main sources. First, staff at TheEvaluation Center compiled archives of documentation about charter schools garnered from annualreports submitted to PDE and from documents gathered during site visits to the schools. Second,we relied on data from the charter school surveys developed by The Evaluation Center and on theNational Association of Secondary School Principals� School Climate Survey. The former has theadvantage of being keyed to charter school issues, while the latter has the advantage of providingcomparisons with national norms. Surveys were administered to all teachers and staff and to samplesof parents and students (Grade 5 and higher). Details on the surveys and sampling procedures maybe found in Chapter 2. Finally, we collected information through interviews with charter schoolpersonnel, representatives of local school districts, and from the two charter school resource centers.

9.1 A Framework for Thinking About Innovation

On the face of it, the concept of innovation is quite straightforward. The root of the word derivesfrom the Latin novus, which means new. An innovative educational practice, therefore, is any suchpractice that is new. However, a little thought reveals that the concept is fraught with ambiguities.Indeed, how new must the practice be to be considered innovative? Must it be truly unique or mayit build on other practices? Perhaps innovations can consist of combining existing program elementsin new ways or in finding new ways to implement and deliver existing program ideas. Finally, whatis the frame of reference for assessing a practice�s innovative character: all schools or schools in aparticular district?

Whether a given practice is innovative depends, in part, on the purposes at hand. Charter schoolproponents variously trumpet a number of purposes or ultimate goals for charter schools, each ofwhich has different implications for innovation. In one view, charter schools are to be publiceducation�s �R&D.� This statement of charter schools� goals is consistent with a fairly ambitiousview of innovation�the creation of truly unique practices that can be shared and perhaps emulatedby a large group of schools. A more modest view of innovation is that charter schools exist primarilyto provide choice. If this is the case, then innovations must simply provide new options for studentsand parents in a particular geographic area. Hence, the bar is high on the former view and somewhatlower on the latter view.

4 See, e.g., in re Souderton Charter School Collaborative, CAB 1992-2; in re Phoenix Academy CharterSchool, CAB 1999-10; in re Sugar Valley Rural Charter School, CAB 1999-4; in re William Bradford AcademyCharter School, CAB 1999-8; in re Vitalistic Therapeutic Center Charter School, CAB 1999-6; in re HillsAcademy Charter School, CAB 1999-12.

106

Figure 9:1 Charter Appeals Board on �Innovation�

The Charter Appeals Board (seeChapter 4) has taken the narrower,second view of innovation. In anumber of decisions4, the Board hasstated that a charter school innovationis any practice or service not providedby the charter school�s host district.Figure 9:1 illustrates this definition ofinnovation using a Venn diagram. Inthe diagram, the range of programsoffered by the host district isrepresented by the circle on the right,while the programs offered by thecharter school are represented by thecircle on the left. Innovation, in thisdiagram, consists in the part of the charter school circle that does not intersect with the host districtcircle. Morever, the Board has ruled that whether a practice is innovative does not depend on howmuch overall overlap there is between host district and charter school practices (the intersectionbetween the two sets). Neither is the extent of charter school offerings relevant in identifyinginnovations. In terms of the diagram, this means that the size of the charter school circle that doesnot overlap with the host district circle is irrelevant.

Why Charter Schools Are Able to Innovate

There are a number of factors that suggest that charter schools should be more able to innovate thantraditional public schools. Among these factors, the following can be noted:

F Charter schools have most rules and regulations waived.

F Charter schools are not bound by seniority rules and union contracts so they can more easily hirelike-minded educators committed to the school�s particular vision.

F Charter schools have more flexibility in implementing reforms because they are not bound byunion work rules governing how personnel spend their time.

F Charter schools can selectively hire teachers and staff in building a team that matches and iscommitted to the unique mission of the school.

F Charter schools have more flexibility to hire people with different credentials. In some casesthey hire non-native English speakers to serve a particular population or Ph.Ds. to teach thesubject in which they specialize.

107

The Pittsburgh resource center representative�s perception on why charter schools can be moreinnovative focused on the positive benefits of empowering teachers by involving them ingovernance. He used a metaphor of �simultaneous engineering� in which �everyone who makes thecar is at the table when it is being designed� so that the design and implementation of innovation aremerged�the same people who are dreaming up the innovation are the ones who are putting it inplace, evaluating its impact, and refining it, all in a seamless way where ideas and information don�tget bogged down in layers of clearance and review far away from the classroom. He noted thatcharter schools have an easier time doing this both because �they are starting with a clean slate� andbecause their small size �tends to force everyone to serve in multiple capacities--you can�t justspecialize and look at your little piece of things and not the connections�wearing multiple hats leadsto creativity because you see opportunities to innovate from the ground up and you don�t have towait for clearance to try it.� Although he admitted that charter schools had certain structuraladvantages in achieving it, he thought that this �egalitarian, nonhierarchical culture� in whichteachers were involved in the governance decisions that shape their work and felt empowered to tryout new ideas was one of the most important things charter schools could model for traditionalschools.

In Pittsburgh, the public school representative said that although the district has been working onmany of the same issues as charter schools, �a small single school is better able to focus on them andimplement them more thoroughly,� especially because they are able to put together from scratch ateam of teachers committed to these practices. He hopes that they will demonstrate taking thesepractices farther and deeper than the district has been able to, although he admits it will be achallenge for the district to figure out �how to scale up those lessons across so many schools andmuch bigger schools and without such a focused team.� In general, he offered several reasons whycharter schools are able to be more innovative:

F Their small size fosters flexibility, good communication.

F They are starting from a clean slate, especially in terms of building their team.

F They have much more latitude over their schedule, both within the school day and school year.

F They can hire staff with unique credentials that are appropriate to their mission.

F They are able to do much more professional development with their staff: �My impression is thatteachers at charter schools see learning as what they do�there is not the punch-the-clock attitude.We have to pay for every extra minute of training. My guess is when they do voluntary trainingin the evening or weekends, they get much better attendance than we do.�

Many of these freedoms are the envy of traditional public schools, and some of the district officialswe spoke with expressed that they wished district schools had this options.

Obstacles To Innovation

It is also important to recall that charter schools have many disadvantages. For example, they appearto have less money for instruction and teachers receive less money, so hiring and retaining effectiveteachers is difficult when districts can pay substantially more.

5 The difference does not quite reach conventional levels of statistical significance (Mann-Whitneyp=0.18).

108

Also, the fact that the charter schools are in a start-up phase limits their ability to innovate. CAOsand teachers are struggling with establishing the school and have little time left over forimplementing the often unique or innovative curriculum suggested in their applications.

One cannot judge the degree of innovation from the charter school applications, however, sincemany schools are not able to implement fully what they have proposed. The Philadelphia resourcecenter representative echoed this point, stating that in her experience, �most charter schools are moreinnovative on paper than in practice at this point. It takes time to grow into implementation, andstarting a school from scratch is tough. With so many issues to worry about, it is not surprising thatthey stick to what they know at first.�

9.2 Organization and Governance

The terms and extent of a chief administrative officer�s (CAO) authority are difficult to observewithout detailed and in-depth fieldwork (which was beyond the scope of this project). However, wewere able to glean information about CAOs through the perceptions of others as expressed onsurveys. Both the parent and teacher surveys asked respondents to express their level of agreementwith the statement, �This school has good administrative leadership.� Some 67 percent of parentssurveyed in 1999/2000 indicated that this statement was true, while 27 percent indicated that it waspartly true. This represents a slight decline from the previous year, when 72 percent judged thestatement true and 21 percent judged it partly true.5 Teachers were somewhat less sanguine abouttheir administrators, with 53 percent of those surveyed in 1999/2000 indicating that the statementwas true and 36 percent indicating that it was partly true. This represents virtually no change fromthe previous year. Both groups� initial expectations, however, were greater than their currentexperience, with approximately 95 percent initially expecting the statement to be either true or partlytrue.

Without a comparison group, however, it is difficult to interpret these findings. It was well beyondthe scope of this study to survey parents and teachers in comparable noncharter schools. Instead, weadministered the National Association of Secondary School Principals� nationally-normed SchoolClimate Survey (SCS) in order to get some sense of how charter schools measure up against anational sample of noncharter schools. The SCS includes a subscale of several items assess teacherperceptions of administrator effectiveness. Some 70 percent of teachers, for instance, agreed thatadministrators set high standards and communicate them effectively to others. Similarly, 69 percentof teachers agreed that administrators set a good example by working hard themselves. Taken as awhole, teachers� responses to the administration subscale on the SCS yielded a standard score of 52,just above the national norm of 50. Charter school teachers in Pennsylvania, therefore, appear torank their administrators a little higher than did a large sample of noncharter teachers. Completeresults of the School Climate Survey are presented at the end of Chapter 12.

6 A Mann-Whitney rank sum test yielded a p-value of 0.05.

109

We also considered the role of teachers in charter school governance. A review of documentaryevidence and responses to open-ended survey questions suggests that teachers have a number ofopportunities to influence school policy, including those listed below:

F Involvement in the strategic planning process, including attendance at planning meetings

F Attendance at meetings to plan and evaluate curricula

F Attendance at school board meetings and otherwise providing input to board members

F Participation in staff evaluations

F Input via school surveys and interviews

F Formal appeals process for teacher grievances

Opportunities, however, do not guarantee influence. Unfortunately, it is difficult to observeinfluence empirically without extensive and in-depth analysis. We did not observe these activitiesourselves, given the limitations of the project. However, the teacher surveys asked teachers to assesstheir own influence in school decisions. Of those teachers surveyed in 1999/2000, 54 percent eitheragreed or strongly agreed with the statement, �Teachers are involved in decision making at thisschool.� This is up from 48 percent for 1998/99, a statistically significant increase.6

However, we also found evidence that would suggest that there are a number of barriers to teacherparticipation in school decision making. Most of these barriers are related to time. Indeed, manyteachers reported that their instructional and other duties leave them little time to participate activelyin school decisions. Teachers in one school expressed concern that the CAO did not allow forenough teacher participation and did not communicate well with staff. However, we found noevidence that this is a general problem across all charter schools.

It appears that parents and students also have a number of opportunities to influence schooldecisions. One source of opportunities for parents comes through charter schools� boards ofdirectors. Some schools reserve a number of slots on these boards for parent representatives. In atleast one school, there is a formal parent advisory board, the chair of which has a seat on the boardof directors. In another school there is a formal parent appeal board designed to hear complaintsabout decisions made by the board of school directors. Several schools allow attendance at schoolboard meetings to count toward parent participation requirements.

In addition to participation on formal boards, some schools give parents credit toward participationrequirements for attending parent-teacher conferences. Other schools have included parents onsearch committees formed to hire new teachers and staff. A number of schools periodicallyadminister parent surveys to solicit input on school policies. Finally, some schools have a general�open door� policy, whereby parents are free to drop in at the school at any time.

110

Subjects on which parents often have influence include discipline policies, dress codes, personnelsearches, business and financial issues, and school facilities.

In many charter schools, students have similar opportunities to influence school decisions. In someschools, students are invited to attend meetings of the school board directors. In at least one school,a student representative sits on the school�s advisory board; at another school a student sits on theschool�s board of trustees. Other schools report that they conduct regular round-table discussionsthat provide students an opportunity to comment on school policies and governance. And, as withparents, many charter schools report conducting surveys to solicit student input.

However, opportunities for influence do not always lead to actual influence. As with the teacheropportunities, project limitations prevented us from directly observing students� and parents�involvement in school decision processes. The parent and student surveys, however, included itemsdesigned to tap respondents� perceptions of their influence. Approximately 83 percent of parentssurveyed in 1999/2000, for instance, said that it was true or partly true that �I am able to influencethe direction and activities in the school.� This, however, is down from 89 percent the previous year,a statistically significant decline.

The charter school student surveys included no similar items. However, the School Climate Survey(SCS) asked students the extent to which they agreed with the statement, �Administrators in thisschool listen to student ideas.� Some 54 percent of students either agreed or strongly agreed withthis statement. Similarly, 56 percent of students either agreed or strongly agreed with the statement,�Teachers and students help to decide what happens in this school.� Hence, it appears that studentsare less likely than their parents to perceive that they have influence in charter schools.

Responses to open-ended questions on the charter school surveys identified barriers to greater parentparticipation. These included completing work and family obligations, especially among those inlow-income neighborhoods. Teacher responses to similar questions often reflected the sameobservation.

Hard evidence of community influence was elusive. However, it does appear that communitymembers� influence in charter schools comes in two forms. First, inasmuch as many charter schoolswere established by (or with the assistance of) community groups, community actors often have greatinfluence over the initial design of charter schools. The role of community and other groups incharter school formation is discussed in Chapter 4.

Community groups have other opportunities to influence charter schools after they are up andrunning. As with parents and students, some charter schools reserve spots on important boards andcommittees for community members. Indeed, in one school, community members reportedlyparticipate on committees formed to address issues such as dress codes, discipline, curriculum,faculty searches, and decisions about the school�s physical facilities.

111

Once again, opportunities for influence do not necessarily lead to actual influence. One way toassess community influence on charter schools is to query community members themselves. Sincethis was beyond the scope and budget of the project, we relied on teacher assessments of communityinfluence as expressed in surveys. Approximately 47 percent of teachers, for instance, reported thatthey were either satisfied or very satisfied with their school�s relations with the community atlarge�virtually the same percentage as in the previous year. A somewhat smaller percentage (38percent) of teachers, however, agreed or strongly agreed with the statement, �Most people in thecommunity help the school in one way or another.� Even fewer teachers (25 percent) agreed orstrongly agreed with the statement, �Community attendance at school meetings and programs isgood.�

Innovations in terms of the operation and governance that were reported by charter school CAOsincluded, among others, the following:! Sibling preference! Extended hours (longer school year and/or school day)! Extended school day plus Saturdays ! Small class size! Continuity between middle school and high school! Mandatory parental participation! Involvement of parents! Continuous staff development programs! A school within a school�a school without walls�distance learning program! Full service providing for the whole student: social services, medical services, and day care! Utilization of token economy system

Discipline Policies

Most of the schools have developed or implemented whole school policies, with three actuallystating that teachers have the responsibility to discipline. Most of the whole school policies focuson preventative measures or behavior modification techniques to give students the responsibility fortheir behavior and teach them how to change certain behaviors appropriately. Four schools mirrorpart or all of their host district�s policy, while the rest of those for which we have data use thetraditional forms of disciplinary action such as time-outs, referrals, detentions, in and out of schoolsuspensions, expulsions, and parent teacher conferences. Some schools have student conferences.

Examples of some of the whole school policies include the following:! Control Theory- where students learn to plan new and more acceptable behavior. Students may

question techniques, be isolated within the classroom, and ultimately be removed. A time-outintervention room helps students plan more effective behavior.

112

! Focus on internalizing appropriate behaviors/behavior modification. Each class follows theserules: �Respect yourself, others, and school property.� Teachers manage most issues. There isdue process for suspensions and expulsions.

! Parents and students receive a handbook. One Philadelphia school does not expel students;rather, there is a program for behavior modification that is created with parents, teacher, student,and school psychologist.

! Provide safe learning environment, discipline. Self-discipline is an integral part of each child�seducation program.

! Positive reinforcement of good behavior; statement of expected behavior, time-out to refocus,parental involvement by phone or note, extended time-out if pattern persists, sending a childhome if s/he puts her/himself or another in danger.

! Overall policy is preventative, positive, and encourages pro social skill development (workedon with police department).

! Level system/behavior management plan. Daily points for good behavior = access to incentives.! Intervention techniques�MA�AT principles: harmony, balance, reciprocity, justice, truth,

respect for self and others. Teachers �collaborate with students to create a caring and cooperativeclassroom where everyone contributes, everyone is responsible for his or her own behavior. . .�

! Rules and regulations are developed by the community with input from parents to fostercooperation, respect, and community collaboration and includes students wearing uniforms andapplying detentions, suspension, and expulsions when necessary.

! Mirroring of school district policy (suspension and expulsion policies)(4 schools).! YouthBuild policies focus on violence, stealing, and drug use.! Students are expected to comply with public and school laws and directions of assigned staff.! Time-outs, conferences with parents, conferences with students, referrals! �Every student will be disciplined based on our knowledge of the current problems and past

experiences with individuals involved. Every situation is different; and after hearing all factsinvolved in the case, an administrator will make a decision.�

! Detention, expulsion, suspension! Respect yourself, respect others, and respect property! Privileges revoked by taking away the school�s card that each student carries! Zero tolerance for violence, weapons, drugs! Due process for expulsions and suspensions! Code of student behavior signed by parents in support! Strict discipline is the hallmark of MACS, and it reserves the right to administer the appropriate

corrective measure such as demerits, restoration of behavior and knowledge, suspension andexpulsion.

! Strict�based on the broken windows philosophy of our police commissioner

113

Descriptions of various teacher/classroom discipline policies include the following:

! Teachers maintain discipline within their own classrooms. An in-school suspension room isavailable. Board of directors has also established a Judiciary Committee to review the mostserious cases.

! Discipline is handled by the teacher who is required to contact a parent�may need to involvean administrator, and student may be suspended or expelled. Parent-teacher conferences areutilized in one school.

! Issues handled first by the employee directly responsible, and if unresolved, by the next levelmanager and eventually the CAO at one school. Where appropriate, communications are sentto the Board president.

Comments from students and staff regarding the implementation and success of discipline policiesinclude the following:

! �No major student conflicts this year� (from field notes) but based on survey, discipline and badbehavior are a problem.

! 89.7 percent of parents do not have concerns about their child�s safety.

! �We have many students with very severe psychiatric problems that seriously disrupt the learningof others, and we have inadequate means of dealing with such disturbances.�

! �Students were out of control.�

! One school has security guards and a good security system.

! one school�s parents reported that there isn�t enough security, there isn�t a security guard andstudents report fights and disrespect.

! �We haven�t had a fight in four years.� (Local newspaper story from a charter school)

! In one school, students say there are fewer fights and less trouble than at their previoustraditional public school.

9.3 Curriculum and Instruction

Unlike many educational reforms, the charter school concept is largely silent on particular curricula,assessment, and instructional methods�except to say that many methods used in traditional publicschools are insufficient. Instead of prescribing a specific reform package, charter laws carve out anopportunity space in which charter schools may exercise autonomy over such matters. The charterschool concept holds that granting schools more autonomy will leave them better able to address theparticular needs of their students. Data collected on the charter school surveys (see Table 8:1)indicated that approximately three-fourths of responding teachers initially expected support forinnovative practices, as well as autonomy and opportunities for creativity in the classroom. This can

114

be contrasted with what they perceived to be true at the time of the survey, when only 59 percent ofteachers responded that they experienced support for innovative practices and 64.8 percent ofteachers responded that they experienced autonomy and opportunities for creativity in theirclassrooms. This seems to indicate that the autonomy granted to charter schools has not yet beenoperationalized into classroom practice.

The charter schools in Pennsylvania tend to categorize their curricula with ethnic or cultural profiles,specific pedagogical approaches, or a focus on academic programs oriented for specific groups ofstudents (i.e., school-to-work, remedial, or college preparation). Below, examples of curricularapproaches are listed to illustrate the diversity of options available in Pennsylvania charter schools.

Ethnic or cultural profiles

! Several charter schools have an ethnic and/or cultural focus and may offer bilingual educationin Swahili or Spanish, for example. Some schools are beginning foreign language instructionas early as kindergarten.

Specific pedagogical approaches/themes

! A few schools reported a curriculum that is based on real world experiences including buildingreal world skills; challenging students to think; and focusing on life and social skills throughcommunity partnerships and utilization of city resources.

! A curriculum is based on parent and student career objectives: Structured job shadowing andinternships develop communication, problem solving, and decision making skills, whilepromoting working independently through the curriculum that includes a school-to-workprogram, internship program, partnerships with businesses, and world languages.

! Many charter schools report that they align their curriculum with state and/or national standards.

! Four schools have a constructivist learning approach integrated into the curriculum; othersreport a back-to-basics curriculum.

! Teaching and learning are integrated in order to promote opportunities for cooperation,teamwork, and continuous learning.

! A few schools base their curriculum on purchased programs like Success for All reading,Everyday Mathematics, or Discovery Works.

! A few schools state that they incorporate the state standards into their curriculum.

Several schools align their curriculum based on state and national standards. Developmentallyappropriate plans are designed for students, with collaboration between parents and teachers todevelop these plans in accordance with the school�s mission. Setting high standards is commonamong most of the charter schools. Approximately 50 percent of teachers indicated in the surveythat an emphasis on academics was important or very important in their decision to seek employmentin a charter school, and about 63 percent of teachers indicated that their school had high standards

115

and expectations for students. Seventy-four percent of teachers initially expected that achievementlevel of students would improve. However, a much lower proportion of teachers believed thatstudent achievement levels were actually improving as much as they had initially expected.

Academic programs oriented for specific groups of students

! Several schools report that students work at their own level or have individualized plans forreaching goals successfully.

! One school was founded as a college preparation school, and a few other schools highlight theirprogram as college prep.

! Several schools serve underachieving, at-risk, or students who may have dropped out of schoolaltogether. Their goals are often to break that trend for students and help them succeed in boththeir community and the workplace. Some activities that promote these goals include activelearning from the student, choice in career paths, a focus on life and social skills, and individuallearning styles. One school focuses on reengaging at-risk inner city youth.

! As highlighted in Chapter 10, a few schools cater to students with disabilities, and nearly halftheir students receive special education services.

Many schools include community partnerships as an integral part of the curriculum (and often partof the instructional program) in one of several different ways: (i) parental participation,volunteering, and/or teaching; (ii) community partnerships with local businesses or universities; (iii)school-to-work activities with students spending time at local workplaces; (iv) use of communitymembers as resources for success; and (v) service learning with students taking part in voluntarycommunity projects.

Below we have included comments made by charter school CAOs and teachers regarding what isunique or innovative about their school. This list illustrates the diversity of practices and alsohighlights the range of what is perceived to be innovative or unique. Though this list focuses oncurricular innovations, many items overlap into instructional innovations such as the learning stylesmethod of teaching or full-day kindergarten.

Home-School Communications! Achievement discussed with parents�no

report cards! Home visits! Ongoing parenting workshops to help

parents understand their kids

Student Accommodations! Personal education plans for every student! After school tutoring program! Full-day kindergarten

! Standards-driven focus! Project-based learning! Technology-supported learning! Long-term contextualized projects! Immediate access to learning tools! Opportunities to complete homework

before end of school day! Opportunity for intellectual curiosity of

students! Well-rounded education to meet all

students� needs

116

Instructional Accommodations! Learning styles methods of teaching! Multiple intelligences! Collaboration with colleagues

Teacher Roles! Teachers with different expertise are

paired together! Autonomy of staff to implement new ideas! Commitment to individualization

Specialized Subjects! Business and entrepreneurial classes! Science is applied to math and language

arts! Specialized instruction in economy! Fencing and chess! Health/physical education! Foreign language, kindergarten and up! Emphasis on world languages! Explicit phonics! Back to basics in math and reading

While many charter schools distinguish themselves by their unique themes, such as ethnic profiles,aviation, or performing arts, it is also important to point out that most of these schools usecurriculum and instructional practices similar to those used by the public schools that surround them.Even so, just over 50 percent of teachers agreed or strongly agreed that they were satisfied with theirschool�s curriculum.

A large number of the charter schools report that they use their sponsoring district�s curriculumframeworks, so one can conclude that, for most of the schools, there was little innovation in termsof either structure or classroom practice. In Pittsburgh, we were informed by district officials andrepresentatives of the charter school resource center that local charter schools generally looked quitetraditional in terms of curriculum and pedagogy, with the exception of one school that is�operationalizing the new standards from top to bottom.� Those standards drive everything fromcurriculum, to pedagogy, to assessment, to professional development, to the organization of time,to communication with parents.

Scheduling Most charter schools adhere to the traditional districts� calendar year and school day. In some casesthey are induced to do so because of the conditions specified in their charter granted by the localdistrict. Additionally, the charter schools are limited in scheduling because of availability oftransportation, which is provided by the local school districts. Nevertheless, we are aware that anumber of schools deviate from the norm in terms of scheduling the day and academic year.Examples of some of these variations are noted below:

! One charter school has a nontraditional instructional schedule where students are in class one dayand work the next.

! A few schools have an all-year programs.

! One school begins later in the day, so that students can associate with the business world byending their school day in the evening.

117

Teaching methods are a fundamental part of instruction. We have found a number of very interestingand unique arrangements in terms of delivering the curriculum. Common arrangements includecollaborative teaching, interdisciplinary teaching, and project-based teaching/learning, which arefound in several schools. Below we have outlined the proportion of schools that reported using thesemethods in 1998-99 when there were a total of 31 charter schools in operation.

Table 9:1 Prevalence of Commonly Reported Instructional Innovations in 31 Charter SchoolsCollaborative Teaching Interdisciplinary Teaching Project-Based Teaching

Yes 35.5% 48.4% 41.9%No 12.9% 3.2% 6.5%No data 51.6% 48.4% 51.6%

Technology and Computers

The presence and usage of computers and technology influence the nature of any given school�scurriculum and instruction. For this reason, we decided to explore this topic more deeply. Whencomparing charter schools within their district in terms of the ratio of students to computers, thecharter schools are doing better. The weighted average of charter school students to one computeris 4.5, whereas the district�s weighted average is 7.5. Below we include a number of findingsregarding the reported presence and usage of computers and technology in charter schools. Abouttwo-thirds of the schools had information regarding technology in their annual reports/folders:

! A number of charter schools reported that computers and technology are integrated intoinstruction in a variety of classes. It appears that many of the other schools have computers andtechnology available, but use it only to supplement curriculum and instruction that is deliveredin the traditional sense.

! Most schools indicated that they also have computer labs. Several schools indicated that theyhave computers available in each classroom, in some cases in addition to the labs.

! In one school, e-Pals are established in foreign language classes so students can communicatewith a person of that language.

! One school maintains student portfolios on-line.

! One school has computers in each class. Additionally, 4 computers are connected to the Internet,and 20 are connected to the home school network.

! One school has computer-software training for teachers.

Table 9:2 highlights data regarding the availability of computers and technology at charter schools.The denominator in this case refers to the number of charter schools in operation and the numeratorindicates the number of charter schools that reported having such computer and technologicalresources.

118

Table 9:2 Availability of Computers and Technology in Pennsylvania Charter Schools

YearWWW Internet in

teacherworkrooms

Internet inclassrooms

Internet incomputer

labs

Internetin

library

Computers CD-ROM

1997/1998 4/6 2/6 1/6 1/6 0/6 4/6 2/6

1998/1999 15/31 7/31 12/31 6/31 6/31 17/31 17/31

Most schools do not specify whether they use technology programs or software. Most schools thatlist information regarding technology only list how many computers they have, if they have acomputer lab, and any grants or monies allocated to technology. Only vague specifics, such asteaching of word processing, spreadsheets, and Internet access are provided by a few schools.

The use of technology and computers in the Pennsylvania charter schools varies considerably. Somepractices, such as students enrolling in Internet courses from home, are unique and are certainlyinnovative in terms of instructional programs for K-12 education. Other schools, whether public orprivate, can certainly learn from programs such as these. While implementation of some of theseprograms has not gone smoothly due to financial restrictions and other obstacles, they still provideexcellent learning opportunities.

9.4 Conceptual Issues Related to Innovation

What are the purposes of innovation? Is it good for its own sake? Or do policymakers hope that itwill lead to other positive spin-offs?

A. Innovation in order to satisfy �customer� preferences. A business school definition ofinnovation is any practice that better satisfies a school�s customers than other alternativepractices. In this view, back-to-basics curricula might be innovative if they do a better job ofsatisfying a given customer base than other curricula.

B. Innovation in the service of choice. One view is that innovation is important insofar as itprovides more choices for parents and students. This view, therefore, is more compatible withCAB�s definition of innovation (innovation exists when charters are doing at least some thingsthat their host districts�i.e., their primary competition�are not). This is a relatively permissivedefinition of innovation, since an innovative school may simply bring practices that are commonelsewhere to a new locale.

C. Innovation in the service of research and development. Another view is that innovation isimportant insofar as it provides new ideas that all schools�charter and noncharter�might adopt.This is a more demanding view of innovation, since it requires schools to develop practices thathave not been used elsewhere.

119

D. Innovation as new combinations of existing practices. A related view is that innovation mightnot only generate wholly new practices, but also develop new ways of combining and/orimplementing existing practices. We might call this the �old wine in new bottles� view ofinnovation.

Diffusion of Innovations

Views C and D above are premised on the notion that innovation is important inasmuch as it canprovide models for other schools to follow. In practice, it is often difficult to determine whetherobserved correspondences between host district and charter school practices suggest that charteractions caused host district change or whether host district changes would have happened anyway.There are at least four scenarios, each of which describes the causal mechanisms that generate anobserved correspondence between charter and host district practices. The first three representmechanisms of genuine diffusion. The fourth model represents a situation where what might appearas diffusion actually is not. These models are illustrated in Figure 9:2.

The collegial model. Noncharter schools change as a result of observing and interacting withcharter schools. Notice that there is only one causal arrow running from the charter school to thehost district. This is called the collegial model since it requires some level of communication and(perhaps) cooperation among charter and noncharter schools.

The competitive model. Noncharter schools change not because they communicate with charterschools directly (note the absence of a causal arrow from the charter to the noncharter school), butbecause charters develop practices that better satisfy parents' preferences. Responses from parentson our surveys provided evidence that parents� choices are most often due to unique features orperceived differences between the charter schools and noncharter schools. Many school districtsconduct exit interviews of parents when they remove their child(ren) from district schools. Thisprovides an opportunity for districts to learn firsthand why parents are leaving. Noncharters adoptsome or all of these practices in order to capture more of the market.

The hybrid model. The above-mentioned models lead naturally to a hybrid model that includesboth cooperation between charter and noncharter schools and the competition mechanism. Frominterviews with charter school and noncharter school staff, we know that changes taking place arereported to be due to a variety of factors. In some cases, the responses from the two groupscontradict each other.

Spurious diffusion model. The fact that a noncharter school �adopts� a charter school practicemay be due not to diffusion but to some third factor, such as parent demand. In other words, charterand noncharter schools may arrive independently at the same practice because of common causalfactors.

One other potential model might be called the random chance model. This would apply whencharters and noncharters arrive at a given practice independently. Hence, there are no causalrelationships in this model.

120

While these models focus largely on the potential for charter schools to influence noncharter schools,it is clear that noncharter schools have influenced charter schools, largely by providing a blueprintfor schools that has evolved over decades. While charter schools have the opportunity to start fromscratch, it is clear that most of them look at ways to adapt current practices. Even though many rulesand regulations are waived, charter schools are still bound by many of the same regulations asnoncharter public schools; and the success of these schools is measured�in part�by their progresson student achievement tests, which are based upon a common set of standards. 9.5 District Impact and the Transfer of Innovations

Based on interview notes, it would seem that both the Philadelphia and Pittsburgh public schoolboards share an implicit definition of innovative as something new and different. They focus heavilyon the �what� of charter practice rather than the �how.� The risk of this interpretative perspectiveis that charter schools may be pushed to do something that is new or unique, but in a rathersuperficial way, rather than being encouraged to develop models of how more common things canbe done better.

One Philadelphia Public Schools� representative said she was surprised by what the charter schoolsproposed in their applications��they were not as unique as I would have thought, given theopportunity they had to start with a clean slate.� She admitted that in Philadelphia the bar for trueinnovation is set pretty high because �with 259 schools in the district, we already have some ofeverything: Slavin, Montessori, reduced class sizes; if there is a reform, we are trying it. So it istough in that context to do something that is truly innovative� if the definition of innovation is doingsomething that the district isn�t doing.

Obviously, it is harder to be innovative in a district like Philadelphia that is already trying a widevariety of different reforms. As a representative of one resource center noted, the charter schoolapplications used buzzwords like �individualized education,� �technology infusion,� and �parentinvolvement,� but that �maybe that would have been innovative a few years back, but everybodytries to do those things now; so charter schools need to go deeper if they are really going to beunique.�

Representatives from the Philadelphia and Pittsburgh resource centers mentioned potential �cyberschools� in the works in their cities as being potentially major innovations, something that districtshaven�t done and maybe can�t do. The SUSQ-Cyber charter school in central Pennsylvania isalready serving as an innovative model for providing instruction to a wide variety of studentsdispersed across several school districts.

It was reported to us that Philadelphia board members were biased against approving charters, sincein order to be approved, �a school had to show they were going to do something district schools werenot doing. Otherwise a board member would ask: �Why do we need a charter to do that?� They hadto show it would be a unique model, a choice parents don�t currently have, and something the districtcould learn from.� In Pittsburgh, one public school representative characterized the board much the

121

Figure 9:2 Models for the Diffusion of Innovations Between Charter Schools and Noncharter Schools

same way, stating that �the number one criteria for charter approval as we read the legislation iswhether they are doing something different from what the district is doing. We only grant a charterif we think it is something the district can learn from. Why else should they exist if they are not onthe cutting edge?�

The resource center representative mentioned that the boards in Philadelphia and Pittsburgh wereinviting applications for charter high schools, specifically to relieve overcrowding and disciplineproblems at district high schools. This is a different goal/criteria for charter schools than innovation.

Although charter school classrooms�for the most part�look very similar to classrooms in districtschools, charter schools are innovating in other ways. From the viewpoint of the Duquesne charterschool resource center, the following exemplify how charter schools are innovating:

122

! Finding ways to increase parental participation, particularly by getting their buy-in to the school�smission and communicating with them more clearly and substantively about what the school istrying to accomplish

! Providing teachers a bigger role in governance and more control over their work, which haspayoffs for innovations in lots of areas

! Implementing longer school days and years (also mentioned by other charter school and districtrepresentatives in other parts of the Commonwealth)

! Being more creative about financing and management: charter schools are �doing more withless� in ways districts should be able to learn from, particularly in terms of outsourcing and moreflexible use of personnel

! Implementing more rigorous accountability processes

A Pittsburgh Public Schools representative�s analysis was that the two elementary-level charterschools in Pittsburgh were quite traditional in most ways: �the extended school day and year wasreally the only thing different about them and the reason they had been granted the charter.� On theother hand, there was more enthusiasm about what the district may be able to learn from the twosecondary schools, one that �has a much more comprehensive school-to-career plan than we havedeveloped. We have really struggled in this area, and some competition and a model of what ispossible here will be good�we really need to learn how to do this better.� The district is working onits own school-to-career institute focused on technology and hopes to be able to learn from thisparticular charter school. �They took an extra planning year and it really showed in their application.Their model is tight, well-defined, ready to implement.� The district representative also said thatone charter school in particular is doing innovative things with project-based learning, portfolios,and standards, although the district is working on these issues. This particular charter school isproviding the district with a more focused and comprehensive test of this model.

One resource representative reports that in Philadelphia there is currently �no mechanism for sharingbest practices. The state says that it is coming, but we haven�t seen it.� The transfer of innovationis something that will have to be encouraged by the state because financial and accountability issuesmake the district/charter relationship too tense for collaboration. �The way the law works, both sideshave a hard time not seeing the other side as robbing them. The money situation is just so dire thatit is hard to have a substantive conversation about educating kids.� There is the potential for transferbut there is no evidence of actual transfer yet, mostly because there is no mechanism.� Like thePhiladelphia Public School representatives, the resource center representative noted that tensionsover finances impeded communication about innovation, but she places the blame more squarely onthe district, going so far as to say that the district �frowns on cooperation. The district wants toapprove as few as possible and help them out as little as possible because they just see them as afinancial drain.� Despite this, the representative reports that �we are seeing the beginning of dialogsystemically and even more so collaboration among individual (district and charter) schools incommon (geographic) clusters.� The example of a district and charter school that teamed up to wina grant for after-school programming was mentioned. Most of these relationships develop from

123

preexisting personal relationships, particularly in the case of community organizations that havehistories of working with schools and are now running charter schools.

As in Philadelphia, the Pittsburgh resource center representative reported that there is currently nomechanism for the transfer of innovation in Pittsburgh. �It is difficult to transfer if the districtdoesn�t know what is going on in those schools. Currently, there is almost no communicationbetween charter schools and the district.� He believes that the district needs a full-time liaison tomanage the give and take of information (it should be both ways) between district and charterschools. He believes the burden is on the authorizer to develop a mechanism and make informationabout innovations more widely available; to fail to do so is missing most of the potential benefit ofcharter schools to districts. He thought in particular that Pittsburgh was missing important lessonsfrom the school that is trying to implement new standards. �This is a valuable model of standardsalignment and accountability. It is exactly what the district is always talking about, and here is thisschool actually doing it and learning how it can be done, and the district isn�t paying any attentionat all.� His perspective is that the district doesn�t take charter schools seriously. �They think theyhave bigger fish to fry. They think these aren�t �real� schools and they are run by idealistic hippies.The scale is just too small in Pittsburgh for the district to respect charter schools and think aboutthem as viable models rather than an irritant.� The Pittsburgh Public Schools� representativedisagreed that there is no mechanism in Pittsburgh for the transfer of innovation. �The state annualreview process serves this function, especially because Pittsburgh takes advantage of �Appendix J�to the annual reports, which allows the addition of district-specific questions. Specifically, they addquestions, requesting information such as this: more detail on the school�s accountability system;how they are measuring their goals for student achievement and what progress is being seen to date;and what unique features they are developing that could benefit the district (he argued that the bestway to get transfer of innovation was to ask the schools directly what they are doing that the districtcould learn from).

The Pittsburgh Public Schools� representative stated that this annual review process gives the districtgood, substantive information on the educational programs in the district�s charter schools. �Havingresponsibility for charter schools located centrally in the chief of staff�s office was critical to thedistrict�s ability to process this information and see where it is applicable to other things that thedistrict is working on.� This suggests an administrative trade-off: an office dedicated to charterschools is likely to be more responsive, but will probably, by design, be far from the administrativecenter. Having responsibility for charter schools located more centrally could mean charter schoolswould have to compete for attention, but to the extent that they get attention, there is more possibilityfor connecting with other district initiatives.

The Philadelphia resource center representative�s recommendations for facilitating the transfer ofinnovation are listed below:

! joint staff development

! working groups of schools trying similar things (these could be on-line)

! conferences where both district and charter schools give workshops about their practices

124

! reciprocal site visits

! electronic bulletin boards or discussion groups that describe charter school practice

�The main goal should be just to get information out there about what charter schools are doing.Currently other educators are not aware at all; how would they know?� According to a PhiladelphiaPublic Schools� representative, without a mechanism for transfer of innovation, charter schools arenot worth the investment. �The reason we approved so many was because we thought we couldlearn from them. This district is massive and hard to move, so maybe the smaller schools could belabs for innovation. Maybe they could prove that the problems are not with the kids but with theadults and the system and that it is possible to do better. But so far we aren�t learning anything andI don�t know how we will. There seems to be little innovation so far; but even if there was, we havelittle way to know.� The district representative pointed out that if charter schools only impact thekids they serve, �that is a drop in the bucket;� so both innovation and its transfer are necessary tojustify the existence of charter schools.

Extent of Perceived Competition

Despite the financial and (to a lesser extent) administrative burdens, a Philadelphia Public Schoolsrepresentative said that �we are so large that we are impervious to competition. The financial pinchis hard but, as much as it is resented, the district doesn�t take the next step and see that as a spur toimprove. We don�t feel the pressure the way we should. So that part of the theory of competitiondoesn�t seem to be working. Maybe a smaller district would feel the impact more urgently� (Rofes,1998). She didn�t know of any examples of direct programmatic changes in response to charterofferings but, again, she thought �the district already has a little of everything.� The Philadelphiaresource center representative also said she was not aware of any programmatic changes in responseto charter offerings. �I don�t hear district people talking about the need to change in response tocompetition from charter schools. Individual principals do; they talk about wishing they could buildtheir own teams and do things as flexibly as charter schools do, but they feel their hands are tied.�This suggests a limit to the competition/market metaphor; traditional schools may be structurallyconstrained from responding to the lessons of charter schools.

In Pittsburgh, the resource representative also reported that they knew of no examples of the districtmaking programmatic changes based on what charter schools are doing. �There are just not enoughof them in Pittsburgh to be considered serious competition. They are not losing enough students tofeel it. I don�t know what that threshold is, but we are way below it.� He did mention that it ishappening in other places. He said State College had added a pre-K to one of its schools in responseto a charter offering. The Pittsburgh Public Schools representative agreed that �we would need tohave critical mass to see major programmatic shifts due to charter offerings, and we aren�t even closeto that level.�

The Pittsburgh resource center representative prefaced all of his comments about district impact witha reminder of how few charter schools there are in the city. He said that the slow growth of themovement is attributable to a combination of three factors:

125

! People are �apprehensive about teacher union opposition and afraid to go up against them.� Hereported that the local union has been explicit about its opposition to charter schools to the extentof threatening schools of education that work with charter schools that they will no longer beable to find placements for their student teachers.

! Pittsburgh is �generally more of a hard-core, loyal union town; socially conservative; slow to riseto many occasions.�

! The condition of public education is not nearly so dire in Pittsburgh as in Philadelphia, so peopleare less willing to confront the system. He argued that �the dissatisfaction here hasn�t reallycoalesced; it is latent, so we need to offer people real choices for them to be able to see a betterway.� Just talking about charter schools in the abstract won�t be enough to overcome the inertia.

Groups had talked with Pittsburgh district officials about starting schools for next year, so they areoptimistic that the movement there isn�t defunct, just slow to grow.

As described above, the representative from Philadelphia Public Schools thinks the state must playa role in facilitating the transfer of innovation, because finances and accountability make thedistrict/charter relationship too tense for willing collaboration. This person shared suggestions thatthe state should consider:

! Do more data collection both on academic achievement and on program (curriculum, instruction,etc.). She said the annual reports are �shallow� in this area and not much help in letting districtsknow what is going on in charter schools that they could possibly learn from.

! Provide forums (on-line, conferences) for sharing information about what charter schools aredoing.

! Encourage charter schools to host site visits and joint professional development to shareinstructional practice.

The state seems to be in a better position than districts to monitor as a neutral observer, but anindependent evaluator might be even less threatening and have a better chance of documenting whatis happening in charter classrooms so it can be shared.

The lack of a substantive communication about educational issues between districts and charterschools is both a product of tension over finances and a reason why that tension continues. Unlesscharter schools are seen as generating transferable innovations and benefits in student achievement,the focus will continue to be on what a financial drain they are.

Most of the factors cited by these respondents as fostering innovation related either to personnel(union) rules or the fact that charter schools are generally much smaller schools. These factors maymake it easier for them to implement reform coherently. So even though both district and charterschools are using the �Success for All� reading program, charter schools may have an advantage inimplementing the reform and thus be more truly innovative. However, one must ask if size and

126

freedom from union rules are the things that are truly distinctive about charter schools (or the factorsthat facilitate substantive innovation).

9.6 Summary

Defining the term �innovation� with regard to charter schools and then identifying innovativepractices occurring in charter schools require looking at the purpose for their creation. Though it isapparently difficult to obtain a charter in Pennsylvania, mainly due to the stipulations from theapproval board that the proposed charter be unique, offer a new choice, or offer something thatdoesn�t already exist in a traditional public school, the main purpose is to provide choice forstudents, parents, and teachers. Though the standards for developing a charter school are high, mostoperating schools are finding it difficult to implement their proposed innovations due to the timeinvested in start-up costs and effort.

Charter schools are still quite similar to traditional public schools. However, there are differencesthat distinguish them, such as parent participation, expanded teacher roles, longer school days andyears, thriftiness in finances (they have the ability to do more with less), community partnerships,and rigorous accountability procedures. Though the invitation to participate is presented to parentsand community members, the actual participation or influence may be superficial. That is, parentsor community members may choose not to participate, or their participation may have little influenceon actual implementation.

Charter schools generally are not as unique or innovative as initially intended or proposed. Theirpotential lies in the flexibility and autonomy inherent in their development and is evident in the highlevel of involvement from teachers in creating and implementing curriculum and instructionalpractices and determining organization. This high level of teacher involvement, though timeintensive, results in a better product for students and the school.

Through their unique themes, charter schools demonstrate their autonomy and flexibility in designinga curriculum that meets the needs of their specific population. Ethnic or cultural profiles, specificpedagogical approaches/themes such as a focus on constructivism, career or service learning,national or state standards, and purchased programs are commonly found in Pennsylvania charterschools. However, many charter schools have scheduling limits due to availability of transportationservices provided by the local school district.

Nearly half of all charter schools utilize collaborative, interdisciplinary, or project-based teaching.The organization of the charter schools varies, though most consider their student population,stakeholder motivation, and beliefs in educational theory. Variance is evident in discipline policiesand implementation, though most schools report having minimal disciplinary problems and areoverall satisfied.

At-risk populations seem to be the most serviced in the Pennsylvania charter schools, although it isimportant to point out that a few of the schools have college prep profiles that do not attract at-risk

127

students. The at-risk population, whose high potential for failure in traditional public schools isaddressed in the charter schools, have the potential to be successful in an appropriate setting. Thesepopulations are serviced mostly through a focus on individualized educational programs, but alsothrough collaborations with community groups that work with these students and their families inother ways.

Self-reported innovations in such specified areas as home-school communications, student andinstructional accommodations, teachers� roles, specialized subjects, and organization and structureindicate an attempt to implement new or unique curricular activities.

Overall, it seems that teachers are generally content with their school�s curriculum and its affect onstudent achievement in relation to higher standards and expectations. In general, charter schools aredoing a better job of getting technology into classrooms by providing more computers per student,as well as integrating technology into classes.

There is an overall feeling among charter school and traditional school employees andrepresentatives that because the school districts are so large, they are impermeable to the supposedcompetition the charter schools are intended to present. One major reason for minimal transfer ofinnovations is the lack of mechanism for transfer. Currently, transfer is thought to be unlikely, notonly because there is no mechanism to promote it, but also because charter schools are busyinvesting time running and developing their new school including responsibilities to the statefinancially and for accountability.

Noncharter schools are being affected by the charter schools and vice versa, but in different ways.Some noncharter schools have begun conducting exit interviews with parents when they removetheir children from a school so as to better market themselves in the future. A few noncharterschools have changed or are changing their curricula, for example, due to charter school practices;in other cases, the changes are reportedly independent of charter schools� practices. Additionally,it is important to note that noncharter schools influence somel charter schools by providing ablueprint from which to operate.

128

Chapter TenSpecial Education and

Pennsylvania Charter Schools

Special education is one of the most complicated and controversial areas that public schools andpolicymakers need to address. Large improvements have been made in the instruction of studentswith special needs within the public school sphere over the last few decades. These improvementshave resulted from often contentious efforts on the part of parents and advocacy groups to pushschools to provide more services, and public schools and their advocacy groups pushing to havemore resources allocated to accommodate these students. A large body of rules and regulations haveevolved to cover the area of special education, and schools are increasingly dependent on lawyersto interpret these regulations and defend their interests.

Charter schools, which are a new form of public schools, are bound by most of the same rules andregulations as regular public schools when it comes to special education. A number of factorssuggest that many of the charter schools will face great difficulties when it comes to specialeducation. First, many staff members and school leaders are new or have relatively little experiencein this area. Second, many of the charter schools�because they are new�do not have procedures androutines in place to screen and provide services to students with special needs. The regular publicschools have had many years to establish their practices, which involve partnering with a numberof outside agencies. Third, the charter schools are already devoting a smaller proportion of theirbudgets to instruction than the regular public schools (see Chapter 5), so students that require extrasupport are harder to accommodate. Finally, there is a shortage of certified special educationteachers; since charter schools are new and because most of them pay considerably less than regularpublic schools, they have a harder time recruiting and retaining special education teachers.

In spite of the obstacles mentioned above, many of the Pennsylvania charter schools have enrolledlarge numbers of students with disabilities and provided them with appropriate instruction. Thereare, however, large variations among the charter schools. Compared with other states where we haveevaluated charter school reforms, we found that the Pennsylvania Department of Education has donemore to both provide guidance for the charter schools in regard to special education and to providetechnical assistance for Chief Administrative Officers (CAOs) and teachers. Nevertheless, someareas still need to be strengthened, and some charter schools have yet to include students withdisabilities.

129

In this chapter, we first provide an overview of the legislative and regulatory framework dealing withspecial education and charter schools (Section 10.1). The material in Section 10.1 focuses on thehistory and development of special education. We then discuss the special education provisions inthe Pennsylvania charter school legislation (Act 22 of 1997) and examine initial and revised rulesand regulations promulgated under Act 22. Finally, we examine how special education policies andpractices are being implemented in Pennsylvania charter schools and attempt to identify areas thatneed further attention.

Data for this chapter came from the following sources. First, analysis of the legal and administrativecontext is based on analysis of Act 22, administrative rules proposed and promulgated under the Act,PDE circulars, and other similar documentation. Second, data used to illustrate trends in specialeducation enrollment in charter and noncharter schools came from the annual reports submitted bycharter schools to PDE�s Office of Educational Initiatives, as well as a number of reports preparedby PDE staff. Finally, data on special education implementation came from documentation in theaforementioned charter school annual reports and from surveys administered to charter schoolteachers, parents, and students. We must emphasize that the data we were able to collect on specialeducation are far from complete. Moreover, it was beyond the scope and budget of the study toconduct intensive case studies of special education implementation. Nonetheless, we believe thatthis chapter highlights a number of important issues.

10.1 Federal and State Legislative Requirements

Although charter schools have evolved out of legislation passed in individual states, certain federalrules and regulations preempt those laws. Primary to these are rules governing civil rights. It isimportant to remember that the United States Constitution does not guarantee the right to a publiceducation. The decision to provide public education is left to each individual state. What theConstitution does guarantee, in the Fourteenth Amendment, is equal protection to all citizens.Therefore, if a state�s constitution guarantees the right to a free public education, that guarantee mustbe extended to all children who choose to attend publicly funded schools. If a state allows a publiclyfunded school to exclude some students, that state is in violation of the federal constitution (Turnbull& Turnbull, 1998).

Specific federal legislation establishing the educational rights of students with disabilities and federalfunding for special education programs originated with P.L. 94-142, the Education of AllHandicapped Children Act (1975). The 1997 amendments and reauthorization of that Act, theIndividuals with Disabilities Education Act (IDEA, P.L. 105-17), also referred to as IDEA97,continues to provide states with guidelines and financial motivation for providing appropriateeducational services to students in publicly funded schools. In addition, civil rights legislation suchas Section 504 of the Rehabilitation Act of 1973 and Title II of the Americans with Disabilities Actenumerates specific penalties for not providing appropriate services for students identified as havinga disability. No exemption from either federal special education or civil rights legislation orregulations can be granted (Lange, 1997).

1 Responsibility for providing special education compliance under the federal regulations lies with LEAs.In states where charter schools have LEA status, they are responsible. In states like Connecticut, where charterschools are not LEAs, the responsibility for compliance rests with the district/LEA.

130

10.2 Pennsylvania Charter School Legislation and Special Education

As states across the country have enacted charter school legislation, little has been included in theregulations to provide clear guidelines for charter school compliance with special education andother federal disability rights requirements. �In their rush to increase choice within the publiceducation system and free schools from any outside regulation, state lawmakers appear to haveneglected to address the impact of unavoidable limitations to their freedom to grant waivers fromfederal laws and regulations to charter schools� (Ahearn, 1999, p. 9). As they moved to implementcharter school legislation, most states did not amend special education regulations to includelanguage pertaining to charter schools. Although all state charter school legislation acknowledgesthat charter schools must comply with federal special education rules, Pennsylvania�s Act 22�likeother states� charter school legislation�provides scant guidance on special education. The onlyspecific directives regarding special education pertain to enrollment policies and fundingmechanisms. Enrollment policies are required to be nondiscriminatory, particularly in the areas ofintelligence, academic performance, athletic ability, or other unique abilities. Funding for specialeducation services is provided through the local public school district. Beyond these minimalguidelines, Act 22 states that charter schools are exempt from state special education regulations,yet responsible for full compliance with IDEA.

Special education, unarguably the most regulated area of public education, was not addressed indetail in Act 22 and was an area of limited debate during the review process prior to passage of theAct (Dale, 1999). Richard Dale, Special Education Advisor for Policy and charter school specialeducation contact for the Pennsylvania Department of Education during the development and initialimplementation of Act 22, completed a study that found four reasons for the lack of attention tospecial education during the framing of the Act: (1) special education experts were not involved inthe development of Act 22; (2) the legislative focus was on choice and efficiency, not on equity;(3) policymakers feared nonpassage of Act 22 if special education was debated at length; and (4)special education regulations would detract from the autonomy goals of charter schools, given theobligations they would face for special education since they have LEA status.1 Dale suggests thatspecial education issues were not neglected purposely in Act 22 in an attempt to create a stronger,more autonomous law, but more in an effort to assure a more rapid passage of charter schoollegislation. �Policymakers feared that the charter school bill might be derailed if much time orexposure were given, in the debate, to the controversial matter of special education� (Dale, 1999, p. 274).

After the passage of Act 22, the state Bureau of Special Education provided preliminary guidelinesto charter schools. Preliminary Guidance for Charter Schools Regarding Special Education (PDE,1997) cautioned that �special education implications of our charter legislation, Act 22 of 1997�needed further study in many areas, and compliance with federal regulations regarding persons with

131

disabilities is mandatory even though Act 22 provides charter schools exemption from state specialeducation regulations. Specifically, this document outlined six areas that Act 22 addresses andprovided a brief summary of the major tenets of IDEA: (1) enrollment policies regarding childrenwith disabilities; (2) provision of services to identified students with disabilities; (3) compliancewith IDEA; (4) how services can be provided�directly or through agreements with the local/intermediate school districts or others; (5) transportation and funding obligations of local schooldistricts for students with disabilities; and (6) the right for charters and local school districts to worktogether to meet the obligations of IDEA.

The Pennsylvania Department of Education, perhaps recognizing the potential problems with thestatutory ambiguity, has provided guidance and technical assistance to charter schools. Comparedwith what we have seen in other states, we think that PDE has been very conscientious in its effortsto support charter schools in implementing special education services and to ensure that schools arein compliance with regulations. PDE considers itself to be a national leader in providing support andguidance to charter school administrators and teachers.

The Office of Educational Initiatives has held a number of workshops with charter school ChiefAdministrative Officers (CAOs) and other personnel in which special education was a top priority.In these workshops, and through documentation shared with the charter schools, PDE officialsinstructed charter schools that they were responsible for providing a free appropriate publiceducation (FAPE) to any enrolled child with a disability. Charter school CAOs were advised thatthe state and local school district were responsible for identifying, locating, and evaluating studentsin need of services (referred to as �Child Find� in IDEA), but that further study was needed in thisarea. Charter schools were advised that they were responsible for evaluating students suspected ofbeing disabled, but that they need only complete reevaluations every three years instead of every twoyears, which is required by local school districts. An exception to this is for students with mentalretardation who must be evaluated every two years. In relation to a number of the componentscovered by IDEA, the Pennsylvania charter schools were advised of the following (PDE, 1997):

! Charter schools must comply with IDEA procedural and implementation requirements forIndividualized Education Programs (IEPs) for students identified for special education services,but they would not be required to use the state format for developing IEPs.

! Charter schools were informed that they must comply with IDEA�s requirements for assuringthat educational placements are in the least restrictive environment, but again were cautionedthat this was an area that required further study.

! Procedural safeguards for students and parents would also need to be provided, as stipulated inIDEA, but charters were advised that some flexibility, within federal requirements, might beafforded to charters since they do not have to comply with state regulations.

! Although charter schools were advised that they must provide personnel who are adequatelytrained to carry out the requirements of IDEA, they were also advised that �much additionalstudy� was needed in this area, especially considering the flexibility given charters in the areaof teacher certification.

2 These basic education circulars can be downloaded from the Pennsylvania Department of Educationweb site <http://www.pde.psu.edu/charter.html>

132

! The documentation provided to charter schools also recommended that charter schools contractwith an experienced special education administrator to oversee the compliance of all aspects ofIDEA.

The Pennsylvania Department of Education has issued numerous �Basic Education Circulars�2 sincethe original preliminary guidance document (PDE, 1997) in an effort to provide charter schools withongoing guidance in all areas of operation including special education. While perhaps helpful to theschools, these circulars are considered guidance, not rules. Representatives of university-basedresource centers for charter schools have expressed concerns that still more specific guidance isneeded, particularly in the areas of implementation and evaluation of special education services.These representatives have also expressed concern that charter schools may be found to be out ofcompliance if investigated by the Office of Civil Rights resulting in legal liability to the state if theydo not help these schools comply. Lack of sufficient guidance on special education was alsoexpressed by local school districts, which stated that �even when required by Act 22 to promulgateregulations�as in the case of special education�the Secretary of Education has not provided anydirection� (cited in Meister & Schuh, 2000, p. 10) and by a state legislator who stated that althoughPDE is �eager to advocate for charter schools, they haven�t provided the level of technical assistancedemanded both by charter schools and school districts� (cited in Meister & Schuh, 2000, p. 9).

In a move to provide more specific guidance to charter schools regarding compliance with specialeducation laws, the Pennsylvania Department of Education published proposed charter schoolregulations for review on July 8, 2000 (PDE, 2000). These regulations are a proposed addition to thestate Public School Code of 1949 and would be added as Chapter 711, �Charter School Services andPrograms for Children with Disabilities.� This addition to the Public School Code attempts toprovide much needed clarification of special education rules and regulations for charter schools.The proposed rules cover four broad areas of special education: (1) general provision andsupervision, (2) identification and evaluation, (3) individual education programs (IEPs), and (4)procedural safeguards.

Within these four areas that the proposed guidelines address, several specific tenets that thepreliminary guidance document indicated as �needing much additional study� are clarified,indicating substantial review by the state regulators. For example, whereas Act 22 allows charterschools considerable flexibility in the area of teacher certification, the proposed rules state thatpersons providing special education or related services must have appropriate certification. Anotherexample involves identification and evaluation of students with disabilities. The preliminaryguidelines (PDE, 1997) suggested that school districts, not charter schools, were responsible foridentifying, locating, and evaluating students with disabilities under IDEA�s �Child Find� mandate.The proposed Chapter 711 rules state, �To enable the Commonwealth to meet its obligations under34 CFR 300.125 (relating to Child Find), each charter school shall establish written policies andprocedures to ensure that all children with disabilities that are enrolled in the charter school, and who

133

are in need of special education and related services, are identified and evaluated� (PDE, 2000, p.3465).

The proposed rules also stipulate that charter schools must have written policies that (1) outline howparents will be notified of special education services and programs and (2) provide for systematicscreening to identify the disabilities present in children enrolled in the schools. Although these rulesdo not require the charter schools to establish outreach identification programs to the generalcommunity, it does go much further than the original guidance document in specifically requiringcharter schools to provide Child Find activities to serve enrolled families. The proposed regulationsalso provide extensive direction regarding charter schools� responsibilities to consider ExtendedSchool Year (ESY) services to qualified students. This requirement of IDEA, which was notaddressed in the preliminary guidance document, normally pertains to students with significantcognitive impairments but cannot be limited to students with these types of impairments. Finally,the proposed regulations address the procedural safeguards that must be followed in the case ofsuspension or expulsion of a student with disabilities (an area of particular concern to schools withstudents displaying extreme behavior problems) and procedures required for due process hearings,record keeping requirements, and enrollment policies. A review of the proposed regulations seemsto indicate PDE�s growing recognition that even though Act 22 attempts to preserve autonomy forcharter schools by making them exempt from state special education regulations, there are few areasof the state regulations that go above and beyond the federal rules to truly allow exemption.

The development of these proposed rules demonstrates the effort that PDE has put into clarifyingcharter schools� responsibilities for students with disabilities and to providing a better frameworkfor evaluating charter school compliance in the area of special education. More work may still beneeded since there are still objections by disability advocates and others. The Education Law Center(2000) in Philadelphia has published comments on the proposed regulations that raise valid questionsand concerns (http://www.elc-pa.org/Charters3.htm). The Center states that it supports PDE in itsefforts to disseminate and enforce special education regulations, but it also suggests that theproposed regulations �have major flaws�including some serious legal problems�and requiresignificant reworking� (p. 1). The comments submitted by the Education Law Center identifyprovisions on enrollment, discipline, and development of a special education plan as the major areasof concern with the proposed regulations. These concerns are summarized in the comments presentedbelow.

! Section 711.7�Enrollment is internally contradictory, overly broad, and inconsistent with theIDEA, the ADA, Section 504, and Act 22.

! Section 711.61�Suspensions and Expulsions misleads charter schools about the scope of thelegal mandates applicable to children with disabilities who are being suspended or expelled.

! Special education plan�The regulations fail to establish clear standards for monitoring,supervision, and accountability of charter schools.

Additional comments submitted by the Education Law Center offer specific recommendations forrewording of these and other subsections, general recommendations for improving assurances that

134

Table 10:1 Proportion of Students with Disabilities Enrolled During the 1999-2000 School YearCommonwealthof Pennsylvania

Charter schoolsopened in

1997 or 1998

Charter schoolsopened in

1999

Totals for allcharter schools

in 99/00Percent of EnrolledStudents with Disabilities 12.5% 12.1% 7.8% 10.5%

students with disabilities are provided a free appropriate public education, and comments regardingpotential legal ramifications that charter schools may face if the regulations do not provide clearguidance. A review of public comments on the proposed regulations, provided by PDE, indicatesthat other participants in the hearings, including representatives from the House of Representatives,the Pennsylvania School Board Association, labor and industry, and private citizens, also expressedconcerns with enrollment and discipline provisions as well as other areas of the regulations. Inresponse, PDE indicated that it will review and revise language regarding enrollment and that�extensive review and discussion is needed� in the area of discipline (specifically section 7.61). Ithas also agreed to extend the hearing period in order to facilitate a wider range of stakeholderinvolvement.

10.3 Pennsylvania Charter Schools and Special Education

In this section we examine how and how well charter schools are implementing special educationservices. In particular, we describe enrollments of students with disabilities in charter schools, theratio of special education teachers to enrolled students with disabilities, the certification ofdesignated special education teachers, training and support provided for special education programs,evidence of innovations or model programs for students with disabilities, and funding of specialeducation in charter schools.

Enrollment of Students with Disabilities

Analysis of the enrollment of students with disabilities in charter schools was based on data takenfrom annual reports that the charter schools submitted to PDE. Of the 47 charter schools thatsubmitted annual reports in 2000, 44 included information pertaining to the enrollment of studentswith disabilities. The proportion of students with disabilities in each of the 44 charter schools rangedfrom 0 percent to 41.8 percent, although the information provided by schools did not indicate thenature or types of students with disabilities they have enrolled. The average percentage of studentswith disabilities for the 44 charter schools with available data was 10.5 percent, which is slightlylower than the statewide average of 12.5 percent. Among the schools that opened during the firsttwo years of the reform, the proportion of students with disabilities was higher (12.1 percent) thanfor the 17 schools that opened for the 1999/2000 school year (7.8 percent). Table 10:1 describesthe proportions of students with disabilities in the various groups of charter schools compared withthe state average. All the charter school data reported in this table are derived from annual reportssubmitted by charter schools in 2000.

135

2

24

10

3 3 20

5

10

15

20

25

30

0% 10% 20% 30% 40% >40%P ercentage o f E nro lled S tudents with D isab ili ties

Num

ber o

f Cha

rter S

choo

ls

Figure 10:1 Histogram Illustrating the Distribution of Charter Schools According to Their Proportion of Students with Disabilities

Some traditional public school district officials disagreed with these figures. For example, thePhiladelphia School District reported that only 4.8 percent of the charter school students in its districthad disabilities, which compared with 10.6 percent for its own schools (Philadelphia Public Schools,2000; Shaffer, 2000). The figures reported by Philadelphia charter schools also suggest that theproportions of students with special educational needs in the Philadelphia charter schools isconsiderably lower than in other charter schools in the state (7.4 percent compared with 18.3percent).

Among the charter schools in Philadelphia as well as among the charter schools across the state,there are large variations in the reported data on special education, with some schools reportingextremely high figures and others reporting no students with disabilities in their school. Five charterschools had surprisingly high percentages, more than twice the state average of students receivingspecial education services: Ridgeview Academy reported 41.8 percent (n=79) of 189 studentsenrolled; YouthBuild Charter School reported 37.6 percent (n=79) of 210 students enrolled;Keystone Education Center reported 36.8 percent (n=78) of 212 students enrolled; LancasterPartnership Charter School reported 36.5 percent (n=31) of 85 students enrolled; GECACCommunity Charter School reported 29.1 percent (n=69) of the 237 students enrolled. By contrast,2 schools reported no students with disabilities in their schools, and another 5 schools reported thatless than 3 percent of their students had disabilities. One of these schools noted that its programincludes individual education plans for all of its enrolled students, which explained why it reportedno special education students identified in its school. Figure 10:1 illustrates the distribution ofstudents with disabilities in 44 of the 47 charter schools for which complete information wasavailable (3 schools had incomplete or inadequate information regarding special education).

136

Although Act 22 and the proposed charter school regulations are intended to assure that studentswith disabilities are afforded equal enrollment opportunities, some observers have expressedconcerns regarding enrollment practices in charter schools. Legal staff from the Education LawCenter suggested that charter schools that have enrollment practices requiring interviews withfamilies of potential students as a prerequisite to being included in the admission lottery may beattempting to �select out� students who they think do not have the necessary skills or interest tocontribute to and fit in with the schools� mission (Shaffer, 2000). At hearings for the proposedregulations, this concern was also expressed by a member of the Pennsylvania School BoardAssociation who stated that several school districts had expressed concerns related to �the effortsby some charter schools to dissuade parents of students with disabilities from enrolling in theirschools� (Public Comment, report provided by PDE). From the data reported by schools and fromwhat we observed in our site visits, a number of charter schools do enroll large numbers of studentswith disabilities. Some of these schools are purposely catering to students that the traditional publicschools were not able to serve. Nonetheless, the data also suggest that some of the charter schoolsare not open to recruiting and enrolling all students.

Improving the monitoring of charter schools and ensuring that the schools are in compliance willrequire improved and increased types of data regarding special education in charter schools. Wehave already been informed by PDE that better data on special education in the charter schools willbe available during the 2000-01 school year. This information will also include more details aboutthe categories of disabilities that are being served in charter schools.

Special Education Teacher Certification

In the annual reports that charter schools are required to submit to PDE each year, they are asked todescribe how they are providing for special education. Most of the charter schools indicate that theyhave special education teacher(s) on staff or have contracted for services with local districts,intermediate units, or private consultants. In addition, all Pennsylvania schools�including charterschools�are required to report information related to teaching staff, including the number of specialeducation teachers (www.pde.psa.edu/esstats.html). Of the 30 charter schools included in theevaluation study, 14 indicated that they had no special education teacher on staff. Of these schools,1 reported no students with disabilities and 4 had unknown numbers of students with disabilities.Fourteen schools reported having one or more special education teachers.

Ratio of Special Education Teachers to Students with Disabilities

The Pennsylvania Department of Education provides schools with recommended maximumcaseloads for special education teachers in each specific disability area, but allows schools someflexibility in determining those caseloads. According to the proposed changes to the Pennsylvaniaspecial education regulations (PDE, Bureau of Special Education, 2000) the ratio of students toteacher can vary from 8:1 for resource services for the more severe disability categories of AutisticSupport and Multiple Disabilities Support to 65:1 for itinerant services for students needing Speechand Language Support. When charter schools use outside consultants to develop IEPs and provide

137

consultation to teachers, it is important that their caseloads are manageable and that they areavailable to both the teachers and the students when needed.

Most charter schools indicate that they are using �inclusion� models, with special education beingintegrated into the regular education program. Since most charter schools are organized to havefairly low class sizes and low teacher/student ratios, inclusion models would seem appropriate,although IDEA requires that a continuum of services be available so that students can receive thespecific services and instructional programs that are appropriate to their needs. The concern is morewith the availability of direct service from a special education teacher, when needed, and the numberof students for whom special education teachers and consultants provide indirect services throughconsultation with regular education teachers. The Education Law Center, in a response to theproposed special education regulations for charter schools, suggests that the regulations pertainingto personnel �include language that makes it clear that the education of a child with a disability ina charter school must be driven by the child�s IEP and supervised by personnel certified in specialeducation, regardless of who is delivering the service� (Hubley, 2000).

Based on a comparison of enrollment data recorded in charter school annual reports and statisticsavailable from the Pennsylvania Department of Education K-12 school statistics data files (PDE,Bureau of Information Services, 2000), the ratio of students with disabilities to special educationteacher for the 30 charter schools included in the evaluation ranged from 8:1 to 69:1.

Training and Support for Special Education Programs

The Office of Educational Initiatives (OEI) at PDE has been very active in providing training andsupport relative to special education for the charter schools. All CAOs are expected to take part inworkshops dealing with special education. Workshops have also been arranged for businessmanagers, special education teachers, and regular education teachers at charter schools. The Bureauof Special Education (BSE) coordinates with the Office of Educational Initiatives (OEI) to supportcharter schools in the area of special education by providing training, technical assistance, onsiteconsultations with staff, individual consultations with CAOs and special education coordinators, andcompliance monitoring visits. This August each charter school received a Charter School ResourceKit containing forms, regulations, question and answer and general guidance documents. The BSEand OEI collaboratively provide monthly regional meetings with CAOs and related staff to addressspecial education issues and questions. In addition, regional training sessions are provided on anquarterly basis with topics based on data taken from school monitoring visits, complaints filed, andrequests for technical assistance. The training sessions planned for the 2000-2001 school yearinclude (1) procedural safeguards/formats (August 2000), (2) evaluation/IEP (November 2000), (3)transition Cordero (March 2001), (4) behavior support/classroom strategies (May 2001).In addition, the OEI conducts yearly onsite observational visits to each charter school to provideindividual consultation regarding special education. The BSE conducts monitoring visits to selectedcharter schools each year to assure compliance with special education regulations. During the springof 2000, 6 charter schools were monitored for compliance with special education regulations andcorrective action plans were developed for each school. During the 1999-2000 school year, 17

3 In addition, the ISSP is a source for adaptive materials and technology, coordination of data collectionsystems, assistance with due process hearings, and special education appeals and mediation. The system liststraining components in a variety of topics related to special education including instructional practices,instructional strategies, prereferral intervention, transition, inclusion, and a variety of topics related to specificdisabilities. CISC also offers on-line courses related to special education including a course titled �SpecialEducation: Effective Practices for my General Education Classroom� (A complete list of topics is available on-lineat http://www.cisc.k12.pa.us/partnerships/issp.htm) (ISSP, 2000).

138

complaints against charter schools were filed with the Bureau of Special Education, with 1 additionalcomplaint filed thus far for the 2000-2001 school year (personal communication BSE).

In summing up their efforts, it appears that the goals of the support and technical assistance providedhave been to achieve the following:

! Prepare charter schools to recruit, enroll, and cater to students with disabilities

! Help ensure that charter schools are in compliance with federal and state regulations dealingwith students with disabilities

! Facilitate relationships between charter schools and intermediate units and between charterschools and special education networks in the Commonwealth

The Pennsylvania State Comprehensive System of Personnel Development (CSPD) also providesprofessional development components that are planned by the Bureau of Special Education andissued through the Instructional Support System of Pennsylvania (ISSP) and the Central InstructionalSupport Center (CISC). The ISSP has a stated mission aimed at providing professional developmentsupport and technical assistance that will result in improved educational experiences for all children,including those with disabilities.3 ISSP offers a number of training options through state andregional centers and makes training materials downloadable from the internet. Inclusion in Charter Schools

As noted above, most of the Pennsylvania charter schools report that they use an inclusion modelfor special education service delivery. Inclusion usually results in increased responsibility forclassroom teachers to provide individual attention to students with disabilities. This approach toservice delivery is encouraged by the least restrictive environment requirements of IDEA but has notbeen well defined operationally. We do not have data on how IEPs are being implemented andmonitored in the Pennsylvania charter schools in order to determine if they are able to provide theneeded support and instruction within an inclusion model. It would be valuable to document howand if charter schools are adapting curriculum and instruction to support students with special needsin the regular education classroom or if they are using unique approaches to education that requireminimal or no modifications for students with disabilities. Comments on surveys from parents andteachers indicate mixed opinions regarding the success of inclusion. Some teachers feeloverwhelmed and undersupported, although other teachers indicate it is working well. Below weinclude some comments from teachers and staff that are critical of the inclusion model used at theirschool.

139

! We are an inclusion setting, and many students are not getting their needs met.! The teachers are doing their best, but we get frustrated because the students need extra help.

Special services aren�t abundant�too many children need them, but do not receive them becauseof the # of students per teacher.

! The school is all inclusive, but does not have enough special education services to meet theneeds of the many special education students

! This is a total inclusion school with only 1 special education teacher!! Special education students are consuming our time and energy.

Some staff, however, think that inclusion is working and that the needs of students with disabilitiesare being well met, as evidenced by the following statements reflecting their schools� unique featuresor areas of strength: ! I feel that special needs of students are addressed fairly rapidly and without discrimination.

Resources are available and services are executed without limitations.! The focus on special needs students is a strength at this school ! We have resources such as case managers and a behavior specialist on site. ! This school works at meeting the total package of special needs students.

Teachers� attitudes and opinions regarding meeting student needs are influenced by the proportionof students with disabilities enrolled at their school as well as by the success of the school�s modelfor delivering special education services. The range of comments included here is not representativeof the teachers comments on special education; rather, they highlight the range of inclusion and thedifferences in terms of success in enrolling and providing instruction for students with disabilities.

Special Education Funding

Funding of special education programs is a challenge for public schools, as well as for charterschools. Additional resources often are needed for alternative curricular materials and adaptivetechnologies, as well as physical accommodations. Pennsylvania has been successful in securingrelatively large amounts of start-up funds for charter schools and in having these resources sent tothe schools in a timely fashion. Even so, many charter schools must deal with the lack of sufficient�start-up�and operational funds when faced with implementing expensive special education andrelated services. Although Pennsylvania charter schools are considered independent school districts,they receive their special education funding from the child�s district of residence, not directly fromthe state. The funds received are determined using a formula that may not always provide charterschools with sufficient reimbursement for the cost of special education services.

Specifically, the funding that charter schools receive for students who qualify for special educationequals the amount provided to the students� district of residence for nonspecial education studentsplus an amount determined by dividing the district of residence total special education reported costsfor the previous school year by 16 percent of the district of residence average daily membership for

140

the previous year (PDE, 1999). For both special education and non-special education, fundinglevels are based on the previous year�s enrollment. LEAs are required to make payments to charterschools in 12 equal monthly installments. Charter schools may appeal to the Secretary of Educationif the LEA fails to make timely payments. The Secretary may then withhold the amount of themissing payment from state payments to the district.

As one can see in the details described in the above paragraph, the funding formula is complex forpersons not working with these figures regularly. In order to describe the relative amount of fundsavailable for students with disabilities, we will provide a few examples. In the Philadelphia CitySchool District expenditures for education during the 1999-2000 were $174,934,635. This amountdivided by 16 percent of the average daily membership of 34,336.48 equals $5,094.72. When addedto non-special education funds (referred to as selected expenditures per daily average membership)of $5,492.57, the total amount of per pupil special education funding available to a charter schoolwith a student whose district of residence is the Philadelphia City School District is $10,587.29. Bycomparison, charter schools in the State College Area School District can expect to receive$11,693.80 for each student with a disability, and charter schools in Pittsburgh School District wouldreceive $16,825.66. Per pupil funding amounts refer to a full 180 days or 900/990 hours school yearand would be prorated for students who enrolled or were identified for services for less than a fullyear. Because the amount of money for each student with a disability is the same, schools that enrollonly students with mild disabilities are likely to receive more resources than needed, while schoolsthat enroll students with moderate or severe disabilities are likely to be underfunded.

CAOs at charter schools reported a mixed picture in terms of the amount of additional funds theyreceived from local districts for the students receiving special education services. Some schoolsreported that they receive sufficient funds for the additional services they provide, while othersclaimed that the funds were not sufficient. A few CAOs noted that the districts may be profitingfrom the students enrolled in special education at the charter school because the district was notpassing on all resources they received for students with disabilities. While we do not have data tosupport this in Pennsylvania, we have noted in other states (Michigan and Connecticut) that studentswith disabilities who are enrolled in charter schools tend to manifest the types of disabilities thatrequire less costly programs and related services, such as speech and language or mild learningdisabilities and may require only one or two hours of special education intervention per day.Traditional public schools, on the other hand, tend to have a higher proportion of students withmoderate or severe disabilities, such as mental retardation, autism, and severe emotional impairmentswith many of the students requiring from half- to full-time special education services. The availabledata and the mixed picture presented by CAOs suggests that there are more complexities dictatinghow resources for students with disabilities follow students. This is a topic that will need to be moreclosely examined in the future.

10.4 Related Survey Results

In addition to the documentation we collected on special education as well as the statistical indicatorsreported by the state and districts, we also used surveys and interviews to collect information

141

My child will receive/receives sufficient individual attention

3.3%

14.5%

82.2%

9.3%

27.5%

63.3%

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation

Current experience

Figure 10:3 Individual Attention for Children: Parents� Initial Expectations vs. Current Experience

Students will receive/receive sufficient individual attention

2.8%

23.2%

74.0%

11.4%

39.8%48.8%

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation

Current experience

Figure 10:2 Individual Attention for Students: Teacher/Staff Initial Expectation vs. Current Experience

regarding special education fromCAOs, teachers, students, andparents. In this section, we reviewand summarize some of the findingsfrom the surveys and interviewsregarding the special needs ofstudents and the ability of the schoolsto accommodate these special needs.

In the surveys we administered, anumber of items were included thatasked the teachers/staff and parents torate and contrast their �initialexpectat ions� and �currentexperience� with the charter school. Some of these items addressed areasor issues related to special education.The findings showed that the staffand parents had high expectations,but the current experience of charterschool staff and parents indicated thatthese expectations were not beingmet.

Figure 10:2 illustrates the responsesfrom teachers and staff when asked torate their initial expectation in termsof individual attention available forstudents. This is contrasted withwhat the teachers and staff currentlyperceive in terms of individual attention for students. Even though their expectations were not met,one can see that most teachers and staff still believed that students were receiving sufficientindividual attention. Just over 11 percent of the charter school teachers and staff reported thatstudents were not receiving sufficient individual attention, while 48.8 percent thought individualneeds were being met. The remainder (39.8 percent) thought the statement about individual needsbeing met was partly true.

Figure 10:3 illustrates a similar pattern for parents. However, we can see that parents� initialexpectations were higher than those of teachers/staff. Likewise, parents� current experiences suggestthat they perceive that their children are receiving sufficient individual attention.

Teachers/staff and parents were also asked to rate initial expectations and current experience in termsof availability of support services. Levels of expectations in terms of availability of support services

4 Some 71 percent of parents who indicated special needs as a reason for choosing the charter schoolindicated that the school was serving student needs not served elsewhere, compared with 57 percent of parents whodid not indicate special needs as a reason for choosing the charter school.

142

were not as high as expectations for individual support. Nevertheless, we found that teachers/staffand parents indicated that their expectations were not being met in terms of support services.Appendix B contains the full set of results on these surveys.

Teachers and staff at the charter schools were also asked whether �class sizes are too large to meetthe individual student�s needs.� Results on that statement indicated 67.4 percent of staff disagreedor strongly disagreed, 15.5 percent agreed, and 17.1 percent neither agreed or disagreed.

The results from our surveys indicate that one of the most important reasons parents cited forchoosing a charter school was that their child�s needs were not being met at a previous school (45percent of parents indicated that this was an important or very important reason for choosing theircharter school). The surveys also included items that asked parents and teachers to assess the extentto which their charter schools are addressing the special needs of students. Just over 65 percent ofparents in 1999/2000, for instance, agreed with the statement that �This school is meeting studentneeds that could not be addressed at other local schools,� down slightly from 67.6 percent during theprevious year. Not surprisingly, parents who indicated that their child�s special needs were animportant factor in choosing the charter school were more likely to agree with this statement thanother parents.4 Thus, charter schools get especially high marks from parents who said that theirstudents� special needs were not well served in other schools. A similar percentage of teachers (64.3percent) agreed that their charter school was serving children with needs not served elsewhere, downslightly from 65.6 percent during the previous year. A few parents, however, did mention thatcharter school need better financial resources for special education:

The results from the nationally-normed School Climate Survey that we administered confirmed ourfindings regarding individual attention for students. This and other related indicators fall under thesubscale entitled Teacher-Student Relationships. On this subscale, the sample of charter schoolteachers and staff rated Teacher-Student Relationships above the national norm for public schools(57th percentile), while results from students on this instrument rated these relationships at the 55th

percentile, which is also above the national norm. Chapter 13 contains more detail about the schoolclimate surveys, and specific details on each indicator and subscale are contained in Appendix B.

Reviewing the results of the surveys and some of the comments provided, particularly from staff,it appears that the biggest concern with special education comes primarily from the teaching staffand concerns adequate staffing and staff training. Some comments from staff that reflect theseconcerns are included below:

! We are not special education teachers, yet we all have students who desperately need smallerclassrooms with special help yet they continue to sit in our classrooms disrupting classroomenvironment and taking an immense chunk of teacher attention.

143

! It would be great to be able to hire more specialized staff to assist teachers with developing anindividualized curriculum for children who are struggling academically.

! Special Education does not exist, although it should.

! We are expected to teach special education students but no one has a special education degree.The special education director doesn�t know what learning support means. He tells us that weare to write IEP goals according to what needs we can meet not what the child�s needs are.

10.5 Summary and Conclusion

There is substantial ambiguity clouding the legal and regulatory framework for special education incharter schools. There is also a conflict between autonomy granted by states to charter schools onone hand and federal regulations on special education on the other hand. Pennsylvania�s charterschool law, like the laws in other states, did not sufficiently address special education prior topassage. To its credit, the Commonwealth has made considerable headway since enactment of Act22 in clarifying charter school obligations and providing guidance to charter schools, in proposingnew regulations in this area, and also in providing support and technical assistance to help charterschools provide special education services that are in compliance with IDEA.

Charter schools in Pennsylvania vary extensively in the degree to which they are enrolling studentseligible for special education services. The average proportion of students with disabilities in thecharter schools was 10.5 percent in 1999-00, which is slightly lower than the statewide average of12.5 percent. Among the schools that opened during the first two years of the reform, the proportionof students with disabilities was higher (12.5 percent) than for the 17 schools that opened for the1999/2000 school year (7.9 percent).

There are large variations in the reported data on special education, with some schools reportingextremely high figures and others reporting no students with disabilities in their school (the rangewas 0 percent to 41.8 percent). Five of the charter schools had surprisingly high percentages ofstudents with disabilities, more than twice the state average, while 2 schools reported no studentswith disabilities and another 5 schools reported that less than 3 percent of their students haddisabilities. Figures reported by Philadelphia charter schools and by the Philadelphia School Districtsuggest that the proportions of students with special educational needs in the Philadelphia charterschools are considerably lower than in other charter schools in the state (i.e., the charter schools inPhiladelphia reported that 7.4 percent of their students had disabilities while the district reported thatcharter schools had 4.8 percent).

The evaluation questions addressed in this chapter are formative in nature, and the findings presentedare intended to provide a description of the current special education services and a framework andbasis for future evaluation of this complex area of public education. While the findings reported inthis chapter suggest that not all is well with special education in the Pennsylvania charter schools,it is important to recognize that regular public schools are also not able to fully satisfy parents whohave students with special needs. One important reason for choosing charters was the inability of

144

their previous school to accommodate the child�s special needs, and a clear majority of the parentsalso thought the charters were able to accommodate students� needs that could not be addressed atother local schools.

Charter schools and traditional public schools alike are faced with the challenge to provide a qualityeducation to all students. The presence of students with disabilities poses strains on availableresources, both human and financial. While charter schools are expected to do more for less, theinclusion of students with disabilities appears�in many cases�to be beyond what some of theseschools are able to manage. We have seen, however, that some charter schools are clearly up to thechallenge. Not only are they able to include and provide quality instruction, they are also exploringnew options in terms of service delivery and in terms of bringing in new resources for studentsrequiring extra support.

The evolving rules and regulations that aim to ensure that all citizens have a free and appropriateeducation have resulted in schools being challenged to find new models to deliver services. Thesmall class sizes that are typical of charter schools facilitate inclusion of students with milddisabilities. In fact, most of the charter schools reported that they pursue full inclusion. This model,however, is not necessarily favorable for the inclusion of students with moderate or severedisabilities.

Public schools are expected to provide a full range of services so that students with disabilities canbe placed in the least restrictive environment that still allows them to function effectively.According to IDEA, charter schools are expected to have a plan to provide a full continuum ofservices. This plan can include cooperative agreements with IUs or LEA, or with private providers.This is one area that appears to need further attention by charter schools.

Better information will facilitate better monitoring of special education in the future. The ongoingwork of PDE to provide guidance and support to charter schools will help to ensure that the qualityof services will improve and hopefully facilitate the enrollment and inclusion of students withdisabilities in all Pennsylvania charter schools.

145

Chapter ElevenPerformance Accountability: An Analysis of

Student Achievement in Charter Schools

At the heart of the charter concept lies a �bargain�: schools will receive more autonomy in operationsin exchange for being held more accountable than other public schools for student outcomes. Muchof this report has examined what Pennsylvania charter schools are doing with their enhancedautonomy in the areas of teacher professional development, organization, governance, curriculum,instruction, and assessment. In this chapter we attempt to determine whether charter schools havelived up to their end of the bargain by producing improved student outcomes. Specifically, weaddress the following evaluation questions:

! Does increased flexibility in exchange for increased accountability result in improved pupilresults?

! Is there evidence that, over the term of the charter, student learning has significantly improved?! What effect does budget have on student results?

It is important to note at the outset that the findings presented in this chapter are suggestive but inno way definitive. Indeed, readers will find that we devote as much space to methodological caveatsas to findings. While perhaps frustrating, the cause of educational reform is best served by carefuland honest consideration of the evidence; this, in turn, requires stakeholders to weigh the strengthsand weaknesses of that evidence. Given that this evaluation is largely formative in nature, the goalsof this chapter are merely to identify strengths and weaknesses that might guide policymakers andstakeholders in program modifications in the coming years. A more summative (thumbs-up,thumbs-down) judgment of the program�s merits will be more appropriate for the year 5 evaluationmandated by Act 22.

Our mandate in this evaluation was to evaluate the charter school program, not necessarilyindividual schools. Clearly, any assessment of an educational reform or program must look atindividual schools. When there are many schools, evaluators can often couch their findings ingeneric terms (schools with certain characteristics also have certain other characteristics) withoutrevealing the identity of any single school. With so few charter schools, however, suchgeneralizations are often elusive. Thus, we were often faced with the dilemma of either reportinggeneralities that fail to do justice to particular schools or to �delve into the details,� which requiresfocusing on the results from specific schools.

When reporting on data that we collected from charter schools through interviews or surveys, wehave avoided any reference to the identity of individual schools. The student achievement data that

1 http://www.pde.psu.edu. Unless otherwise noted, all subsequent data referred to as �PDE raw data files� weredownloaded from this web site.

2 School participation rates were calculated from raw data files of writing scores provided by PDE.

146

we have analyzed and reported on in this chapter, however, are readily available on PDE�s web site.1Therefore, we decided to attempt to provide the most complete picture of individual schools wecould rather than attempt to preserve the anonymity of the charter schools. Generally speaking, thischapter reports on aggregate patterns, while school-by-school analysis is left to the appendices.

Section 11.1 provides an overview of the data and methods used in this chapter. Section 11.2provides an extensive account of the many limitations to this analysis. Section 11.3 discusses thefindings from each analysis undertaken by the evaluation team. Section 11.4 provides a number ofrecommendations that would enhance the capacity of future evaluations to provide more rigorousassessments of charter schools� impact on student achievement. Finally, Section 11.5 provides abrief summary of the chapter.

11.1 Data and Methods

In this section we discuss three issues: (a) measurement of student achievement, (b) assessment ofcharter schools� impact on student achievement, and (c) data sources.

The PSSA and Other Measures of Student Achievement

�Student achievement� is a contested concept; there is a legitimate and good faith debate about whatit means and how best to measure it. Different assessments, moreover, tap into different parts of theachievement construct. Thus, evaluations involving student achievement are sensitive to the choiceof tests used to measure achievement. The Request for Proposals (RFP) explicitly asked for ananalysis of student achievement using scores from the Pennsylvania System of School Assessment(PSSA [see Appendix One for a complete list of the evaluation question included in the RFP]).Accordingly, most of the analysis in this chapter uses that test as a measure of student achievement.We also sought, where possible, to analyze scores from a number of commercial standardizedachievement tests. The commercial tests included the Iowa Test of Basic Skills, the New StandardsReference Exam, the Metropolitan Achievement Test, Terra Nova, and the Stanford AchievementTest. The evaluation team also requested data from nonstandardized assessments, but received nonefrom charter schools. We did learn, however, that a number of schools are currently in the processof developing such assessments. Future evaluations would do well to consider these data.

Since the 1995/96 academic year, the PSSA has assessed 5th, 8th, and 11th grade students in bothmathematics and reading. A writing assessment for 6th and 9th graders has been phased in gradually,with approximately 66 percent of schools participating in 1995/96, 75 percent in 1996/97, 75 percentin 1997/98, and 92 percent in 1998/99.2

3 The first quartile consists of all students scoring between the 75th and 99th percentiles: the second quartileconsists of all students scoring between the 50th and 74th percentiles; the third quartile consists of all students scoringbetween the 25th and 49th percentiles; and the fourth quartile consists of all students scoring between the 1st and 24th

percentiles. 4 For all analyses we focused on scaled scores rather than quartile percentages because mean scaled scores can

simultaneously capture change in all four quartiles. Indeed, it can be shown that the percentage of students falling in onequartile can change without changing the percentage in all of the remaining quartiles. Thus, analyses that focus on asingle quartile (e.g., those that establish a �cut point� such as the bottom quartile as the primary measure) risk missingsignificant changes in student achievement.

5 Equating also has the statistically perverse consequence that after 1995/96 there can be more or fewer than25 percent of students in any one of the four �quartiles.�

6 The PSSA, however, does not provide a developmental metric. That is, one cannot say, for instance, thata 1310 on the 8th grade math examination represents 10 points worth of gain over a 1300 on the 5th grade mathexamination. Strictly speaking, all the scores can tell us is where a given student stands in relation to his or her peerswho took the examination in 1995/96.

147

The math and reading portions of the PSSA are norm-referenced tests; that is, students� scores reflecttheir position relative to other students in the state. PDE documents and data files report PSSAscores in two formats. First, they report the percentage of students in a given school or districtscoring in each of the four quartiles.3 Second, PDE files report a scaled score that ranges from 1000to 1600.4 While each student receives such a score, public data files include only mean values foreach school.

During the first year of its administration (1995/96), scaled scores were constructed so that the meanwas 1300 and the standard deviation 100. All subsequent versions of the test have been statisticallyequated to the 1995/96 version. This means that the scale is �anchored� in the 1995/96 test and anyincreases or decreases in scaled scores reflect actual changes in student achievement, not justchanges in the distribution of scores across the Commonwealth. Operationally, a 1300 in anysubsequent year represents the mean for the 1995/96 academic year.5 PSSA scores, then, allowevaluators to compare students in one school with students from another school. They also allowevaluators to assess growth or decline over time, subject to the limitations imposed by the need torely on school-level mean scores.6

Isolating Charter School Impacts

Simply knowing a charter school�s mean score on the PSSA, however, does not by itself allow usto assess charter schools� impact on student achievement. Assessing impacts requires us to try toestimate, in addition, what student outcomes might have been like in the absence of charter schools.This task is severely complicated by the fact that scores on achievement tests reflect two sets ofinfluences. First, the fact that students and families tend to sort themselves into school districtsaccording to income and other factors means that achievement scores reflect students� economic andsocial endowments. Indeed, more than a generation of scholarly research, beginning with theColeman Report (Coleman, 1966), has shown that a large proportion in the student-to-student and

148

school-to-school variation in test scores can be explained by student and family socioeconomicstatus. This fact usually makes it difficult to isolate the second major influence on achievementscores�the educational value added by schools� inputs and processes. The value-added componentof achievement test score, therefore, is that part of the variation among scores that can be explainedby what the school is doing. The job of a charter school evaluation is to isolate the value added bycharter schools from the various economic, social, and cultural resources that students bring withthem to the school.

The RFP provided specific direction on this score. Specifically, the RFP called for

! comparisons of charter schools with �district schools�! analysis of achievement trends over time! comparisons with �like� noncharter schools.

We interpreted the term �district schools� in two ways. First, we interpreted it to encompass allnoncharter school schools in the Commonwealth. Hence, the first set of analyses presented in thischapter compare charter schools as a group with all noncharter schools as a group. We alsointerpreted district schools to encompass only those schools that sponsor charter schools. For thetemporal comparisons, we compared consecutive cohorts of students in grade levels assessed by thePSSA over time. The temporal comparisons, in effect, allow each school to serve as its owncomparison group.

We operationalized the directive to compare charter schools� achievement scores with those of �like�noncharter schools using a statistical regression model. This technique generates achievementscores adjusted for students� demographic characteristics by comparing a school�s actualachievement score with a predicted score generated by a model that includes student demographicfactors. The difference�or �residual��indicates how a given charter school did when compared withother schools with the same student demographics. Further discussion of this technique may befound in Appendix D.

Unfortunately, serious data limitations precluded a reliable answer to the evaluation question, �Whateffect does budget have on student results?� There are few issues more contentious or technicallydifficult to resolve than the effect of expenditures on student outcomes. Only recently have rigorousempirical studies begun to find consistent evidence that expenditures affect educational outputs (see,e.g., Grissmer et al., 2000; Wenglinksy, 1998). These studies had to employ large data sets andrelatively sophisticated statistical techniques in order to separate the impact of expenditures from themyriad of other forces that influence student outcomes. The evaluation team explored a number oftechniques for assessing the impact of charter school expenditures on student outcomes. In everyinstance, however, the quality of the findings was insufficient for inclusion even in a chapterpresenting only �suggestive� findings. The most serious limitations stem from the large number ofcases required to investigate the kind of multicausal relationships involving expenditures uncoveredby previous studies.

7 Once again, data for 1998/99 came from the PDE web site. Data for earlier years came from the databaseconstructed by the Education, Policy, and Issues Center in Pittsburgh.

8 The OEI annual reports also included a number of variables not found in the School Profiles data files.

9 The PEPS database was constructed by Professor Emeritus William Cooley of the University of Pittsburgh.The project, among other things, developed district-level indicators based on census data. The U.S. Census Bureau plansto produce these estimates themselves after the 2000 census.

149

As Pennsylvania�s charter school movement grows, the increasing number of cases will open up newpossibilities for future evaluators. Even then, however, it is likely that answers to the expenditurequestion will remain somewhat (if not highly) uncertain.

Data Sources

The evaluation team constructed a relational database that includes test scores and a number of otherschool and district attributes. PSSA scores for the 1998/99 school year came from raw data filesposted on the PDE web site. Scores for previous years came from a relational database constructedby the Education, Policy, and Issues Center in Pittsburgh from PDE raw data files . Unfortunately,scores for the 1999/2000 were not available in time for inclusion in the study. Data on otherattributes of charter and noncharter schools (e.g., concentration of low income students) came fromraw data files associated with the Pennsylvania School Profiles.7 In some instances, the SchoolProfiles data files contained missing values for charter schools. In these cases we relied upon annualreports submitted by the charter schools to PDE�s Office of Educational Initiatives (OEI).8 Wherethere were disagreements between the School Profiles Data and data from the annual reports, weused data from the annual reports because the data in the OEI annual reports were, in most cases,more recent than the values contained in the School Profiles files. Other data (e.g., single-parentstatus) came from the Pennsylvania Educational Policy Studies (PEPS) database, which draws uponthe 1990 U.S. census.9

11.2 Limitations to the Analysis

Given the directives in the RFP and the inherent limitations in the data available to the evaluationteam, there were important limits in the extent to which we could draw solid inferences about charterschool impacts from the analyses presented in this chapter. We begin by discussing a number ofgeneral limitations that apply to all of the aforementioned comparisons. We then considerlimitations that apply to specific comparisons.

Generic Limitations

Several limitations are related to this key question: What proportion of the observed changes inoutcomes is due to the program itself and how much is due to the characteristics (both positive andnegative) that students bring to the school? Put another way, what proportion of the observedoutcomes is due to selection and what proportion is due to value added by the program?

150

Table 11:1 Number of Charter Schools Participating in the PSSA by Subject/Grade

Portion ofthe PSSA

Number of Schools Tested

1997/98 1998/99

Math 5 1 10

Reading 5 1 10

Writing 6 2 9

Math 8 3 8

Reading 8 3 7

Writing 9 1 7

Math 11 2 5

Reading 11 2 5Source: Calculations based on PDE raw data files.

The first such limitation is that the PSSA system provides no way to track the performance ofindividual students as they move from one school to the next. This left the evaluation team with noway of observing students� precharter school achievement levels. Instead, the team had to settle forobserving charter school students at the end of the 1997/98 and 1998/99 academic years�both afterthe administration of the charter school �treatment.� This severely limits the evaluation�s ability todistinguish the value added by charter schools from the influence of other factors. Indeed, it isunlikely that any evaluation could provide strong results without such information.

Second, Pennsylvania�s experiment with charter schools is still quite young, and there are manyreasons to expect that it might take more time for the effects of the reform to show up in measuresof student achievement. Indeed, all of the schools for which we have test score data were in theirfirst or second year of operation. New schools, like most organizations, often face a number of start-up issues that can hinder their ability to produce their intended results (see Chapter 4). Moreover,recent experimental evidence from the Tennessee class size experiment suggests that the effects ofpoor schooling in early years can persist for many years to come (see, e.g., Grissmer, et al., 2000).Hence, even the most effective schools might find it difficult to counteract the effects of years ofschooling in ineffective schools.

A third limitation is the small number of charter schools for which the evaluation team had studentachievement (and other related) data. Of the 46 charter schools in operation during the 1999/2000academic year, the team had reliable student achievement data for only 21. In large part, this isbecause data for the 1999/2000 administration of the PSSA were not available in time for inclusionin this report. Moreover, a small number ofschools either did not enroll students in gradesassessed by the PSSA, did not report scores, orboth. Even fewer schools reported achievementdata for any given grade level or subject area.Table 11:1 shows the number of schools reportingscores for each subject and grade of the PSSAduring the 1997/98 and 1998/99 academic years.The table shows that the 5th grade reading andmath examinations were administered mostfrequently, with 10 schools reporting scores in1998/99. The 6th grade writing examination wasthe next most commonly administered test (9schools). The 11th grade reading and math testswere least frequently administered. Thesepatterns, of course, reflect enrollment patterns;more Pennsylvania charter schools enrolled lowergrade students than high school level students in1998/99. Moreover, some charter schools eitherdid not enroll any students at the grade levelstested or did not report scores for students enrolled

151

in appropriate grades. This accounts for the fact that the analysis includes data for only 21 of the 31charter schools open during the 1998/99 school year. We were unable to obtain reliable estimatesof the number of students tested in each school. However, we know that in several schools fewerthan 10 students took a given assessment.

Finally, the practice of reporting only school means and quartile percentages presents a number ofless severe problems. The mean is a measure of central tendency and provides no information aboutwithin-school variations among students. Knowledge of within-school distributions is especiallyimportant when there are a small number of high- or low-scoring students, since such �outliers� canvery easily skew school means. The problem of outliers and skewness is especially troubling when,as is the case with many charter schools, there are few students in a school. In such situations thepresence of just one or a few outlying students can have a profound impact on school means. Hence,some of the changes we observed in charter schools could be due to high or low scores from a verysmall number of students. The quartile percentages reported provide rough assessments of within-school variations. However, these assessments are rudimentary, at best. Indeed, a large proportionof a school�s students� scores in the top and/or bottom quartiles might suggest the presence ofoutliers. However, it provides no way to judge just how serious the problem is. Readers should bearthis in mind when interpreting any of the PSSA results provided in this chapter.

Next, we turn to specific limitations associated with each analysis outlined in Section 11.1.

Limitations to Comparisons With All Noncharter Schools

Comparing charter schools as a group with all noncharter schools provides an assessment of howwell or poorly charter school students are doing in absolute terms. This might be instructive because,demographic and other differences notwithstanding, charter school students eventually compete inthe same labor markets as noncharter students. By contrast, such comparisons have little value forassessments of charter schools� impact on student achievement. Indeed, it is likely that most of thedifferences between the two groups simply reflects differences in the types of students recruited byand attracted to charter schools. This seems plausible given chapter 4's finding that the likelihoodthat a given district will sponsor a charter school increases with increases in the concentration ofpoor and nonwhite students and chapter 6�s finding that charter schools tend to enroll higherconcentrations of poor students than the Commonwealth�s public schools as a whole. In short, suchcomparisons do little or nothing to allow stakeholders to distinguish the value added by charterschools from a host of other influences on student achievement.

Limitations to Comparisons With Host Districts

Host districts are a potentially attractive comparison group because students and families tend to sortthemselves according to geographic locale. Hence, other things being equal, one would expect thathost districts�normally geographically proximate to charter schools�would have students withsimilar characteristics. Pennsylvania charter schools, however, may recruit students from outsidetheir host districts. One school, for instance, enrolls students from some 70 school districts across

10 This problem is more severe for schools with small student populations, since the addition of just a fewhigh or low scoring students in any given year can substantially alter aggregate school scores. To illustrate therelationship between year-to-year variance and school size, we calculated standard deviations of two-year gain scores(1997/98 - 1998/99) for all Pennsylvania schools with 0-250 students, 250-500 students, and more than 500 students.The standard deviations show a steady decline as one moves from groups of smaller to larger schools: 62, 49, and42 respectively. In probabilistic terms, the likelihood that a small school will have a large two-year change score(either positive or negative) by chance alone is substantially higher than for a larger school (source: calculations basedon raw data files).

11 We received no usable evidence from nonstandardized testing systems. However, we know from our sitevisits that several schools are planning to use them.

152

several counties. Moreover, some charter schools explicitly target special needs populations. Thus,in many cases the host districts� students are likely to be substantially different in educationallyrelevant respects from charter schools� students. As a consequence, many observed differencesbetween charter schools and their host districts are likely the result of student selection effects aswell as charter school impacts. Given that it was well beyond the scope of this study to collect pointof origin data on individual students, the evaluation team had no reliable way of partialling out theindependent impact of the charter schools. In sum, while the host district comparisons are farsuperior to the comparisons with all noncharter schools, they still face important limitations.

Limitations to Analysis of Temporal Trends

The analysis of temporal change in student achievement scores was limited by several factors. First,it is difficult to estimate temporal trends from only two years of data with any confidence. Previousresearch suggests that test scores exhibit the same sorts of cyclical trends as economic data. Hence,a gain from one year to the next might be offset by a decline in subsequent years. Second, two-yearchange score data were available for only four charter schools. Indeed, while there were six schoolsin operation during the 1998/99 academic year, two did not enroll students in grades assessed by thePSSA. This makes any attempt to make generalizations to the Pennsylvania charter schoolmovement as a whole tenuous at best. Third, the PSSA tests students at a limited number of gradelevels. While this has the advantage of minimizing the burdens of testing on schools and students,it creates serious problems for evaluating school effectiveness by ruling out the possibility of genuinelongitudinal analysis. Instead of following a single cohort of students from one grade to the next,the evaluation team was restricted to comparing the 1997/98 cohort of 5th, 6th, 9th, and 11th graderswith the 1998/99 cohort of students from the same grades. For instance, if today�s 5th graders at aschool are different in educationally relevant respects from last year�s 5th graders, then any observedchange in that school�s test scores might simply reflect the change in student body, not the valueadded by the school.10

Examination of other commercially available standardized tests allowed us to avoid some of thedifficulties with the PSSA data, but presented their own limitations as well.11 Many schoolsadminister the Iowa Test of Basic Skills (ITBS) or other tests at the beginning and the end of a single

12 Moreover, inclusion of data from different levels of analysis in a conventional regression model usually leadsto violation of the nonautocorrelation assumption that helps ensure that regression estimates are efficient. Multilevelmodels, which addresses this problem, were not feasible due to the small number of schools in most districts.

153

school year. Hence, unlike the PSSA data, these scores allow us to make reasonably clean pre/postcomparisons on the same group of students (except for cases in which there was high studentattrition during the year). Unfortunately, few schools elected to share such data with the evaluationteam, making for a small and unrepresentative sample of schools. Moreover, data from differenttests rely on different norm groups and are often reported using different types of measures. Thisleads to an �apples and oranges� quality that hinders attempts to combine scores from differentcharter schools into a composite view.

Limitations to Comparison With �Like� Noncharter Schools

The comparisons between charter schools and �like� noncharter schools are limited by two factors.First, student eligibility for free and reduced-price lunch appears to be the only demographic factoron which comparable data exist for both charter and noncharter schools. The evaluation team wasable to locate district-level data on a number of other demographic variables that previous researchhas shown to influence student achievement, including racial composition, community educationlevels, and concentration of single-parent families. These data, however, are quite dated, since theycome from the 1990 census. Moreover, since the estimates are at the district level, any inferencesfrom the data about the characteristics of charter schools would commit the ecologicalfallacy�making inferences about lower units based on aggregate estimates about higher units.12 Theevaluation team also attempted to comply with the RFP�s directive to compare charter andnoncharter schools on PSSA participation rates. Unfortunately, PDE records and data files weremissing participation rates for a large proportion of charter schools.

Second, since the regression models used to make the charter-noncharter comparisons are sensitiveto the exclusion of important variables, the adjusted scores reported in this chapter likely suffer fromsome statistical bias. Fortunately, diagnostics discussed in Appendix D:2 suggest that this bias,while important, is not fatal. Other comparison methods considered would suffer from similarproblems.

11.3 Findings From the Data

In this section we discuss findings from the analyses discussed above. For expository purposes, wecombine the first two evaluation questions (Does increased flexibility in exchange for increasedaccountability result in improved pupil results? Is there evidence that, over the term of the charter,student learning has significantly improved?). As discussed above, data limitations precluded anacceptably accurate answer to the evaluation question, �What impact does budget have on studentachievement?�

13 This estimate is the average of all the difference scores displayed in Table 11:2 weighted by average dailymember (ADM) in the charter schools. ADM is a less than ideal weighting variable since it includes all students inthe charter schools, not just those tested in a given subject and grade. Unfortunately, we were unable to obtainreliable enrollment data for the grade levels tested, which would have provided a more precise weighting variable.

154

Impact on Student Achievement: Comparisons With All Noncharter Schools

Each comparison discussed in Section 11.1 addresses, in various ways, the impact of charter schoolson student achievement. However, as Section 11.2 pointed out, there are serious limitations toinferences drawn from these data. Accordingly, readers should consider the discussion below assuggestive and offered in the spirit of exploration and improvement rather than definitive. In orderto facilitate clear exposition of the findings, we have placed detailed discussions of the findings andthe methods that generated them in Appendix D.

Table 11:2 Aggregate Comparisons Between Charter Schools and All Pennsylvania Schools,1998/99

Portion of the PSSA

Number of Schools Tested

Mean Scaled Score,Charter Schools

Mean Scaled Score, AllNoncharter Public Schools

Difference (charterschools minus all PA)

Math 5 10 1130 1310 -180

Math 8 8 1130 1300 -170

Math 11 5 1140 1300 -160

Reading 5 10 1160 1310 -150

Reading 8 7 1150 1310 -160

Reading 11 5 1200 1300 -100

Writing 6 9 1180 1240 -60

Writing 9 7 1250 1330 -80

Mean13 1160 1300 -140Source: Analysis of PDE raw data files by The Evaluation CenterNotes: All PSSA scores are averages weighted by average daily membership. We have also adopted PDE�s practice ofrounding PSSA scores to the nearest 10. Missing data on participation rates prevented us, however, from adhering tothe PDE practice of excluding schools with low participation rates. Estimates for �all PA Public Schools� are taken fromthe Pennsylvania School Profiles reports.

Table 11:2 presents evidence from aggregate comparisons between charter schools as a group andall noncharter schools in the Commonwealth. In order to simplify the table somewhat, thispresentation includes data from the 1998/99 academic year only. This decision seemed justifiedgiven that we had 1997/98 achievement data for only four charter schools open at the time. Datafrom both academic years are included in all subsequent analyses.

14As discussed in section 11.1, the scale runs from 1000 to 1600.155

Table 11:3 Summary of Host District Comparisons

Portion of PSSA Difference (CSminus HD)

Math 5 -61

Math 8 -89

Math 11 -136

Reading 5 -25

Reading 8 -68

Reading 11 -88

Writing 6 -15

Writing 9 38

Mean -50Note: Mean is weighted by average

daily membership

Table 11:2 reveals that charter school students were strongest in writing, with an average score of1250 in 9th grade writing and 1180 in 6th grade writing.14 Charter school students were also relativelystrong in 11th grade reading, with an average score of 1200. Overall, charter students� weak pointappears to be mathematics, with average scores of 1130 in both 5th and 8th grade math. In all portionsof the PSSA, aggregate charter school scores fell well below Commonwealth averages, with thegreatest difference coming in 5th grade math (180 points) and the smallest difference in 6th gradewriting (60 points). Averaging all the scores together, charter schools scored 140 points below otherpublic schools on the PSSA.

To say that charter schools on average were outperformed by other public schools is not to implythat all charter schools did worse than all other public schools. For most portions of the PSSA,average charter school scores ranged from a low of 1000 (the lowest value on the PSSA scale) toscores well above the Commonwealth average. Scores in other Commonwealth public schools alsoexhibit great variability. However, a much larger proportion of other Commonwealth schools scorednear the top of the PSSA scale (1600).

Readers should refer to the discussion in Section 11.2 to recall the limitations with this analysis.

Impact on Student Achievement: Host District Comparisons

The comparisons of charter schools and their host districts are summarized in Table 11:3. Morecomplete tables are provided in Appendix D:3. The numbers given in the tables report the charterschool score minus the equivalent host district score. Thus, positive values indicate that charterschools outperformed their host districts while negative valuesindicate that the host districts outperformed the charterschools. In cases of multiple sponsorship, the table reportscomparisons with composite values that are the average(weighted by average daily membership) of each of theschool�s host districts. The table also averages data for thefour schools with data from two years.

Examination of the table reveals that most of the differencescores are negative. Indeed, the overall average difference forall charter schools and all portions of the PSSA was -50. Thismeans that the typical charter school scored 50 points below itshost district. This deficit, however, is less severe than the 140point difference between charter schools and allCommonwealth schools on raw PSSA scores. Once againthere was considerable variation among PSSA subject areasand charter schools. In every subject area except 11th grademath and reading, several individual charter schoolsoutperformed their host districts, some by as much as 200points or more. The strongest area for charter schools was the9th grade writing examination where, as a group, charter

15 Drawing upon recent experimental work on class size and other educational innovations, one might expectthat younger students would, other things being equal, generate higher achievement scores than older students. Thisfollows from the finding in the Tennessee class size experiment that achievement at any given grade level continues tobe impacted by school inputs at all earlier stages of the education process. From this it should follow that charter schoolsenrolling younger students will be more successful in improving achievement than charter schools that enroll olderstudents. This does not appear to hold for Pennsylvania charter schools. However, given that the change score analysisincludes data from only four schools, this issue bears watching as the reform matures.

156

Table 11:4 Two-Year Change Scores on the PSSA, 1997/98 to 1998/99

Portion of the PSSA

Numberof

Schools

CharterSchool2-YearChangeScore

Charter SchoolChange Score

Compared withHost District

Change Score Math 5 1 40 40Math 8 3 146 102Math 11 2 99 99Reading 5 1 -120 -150Reading 8 3 222 197Reading 11 2 134 152Writing 6 2 33 30Writing 9 1 50 -10Mean 105 86

Note: Mean is weighted by average dailymembership. Estimates derived from analysisof PDE raw data files by The Evaluation Center

schools outperformed their host districts by 38 points. Some of the deficits shown by individualcharter schools, however, were even larger. Indeed, some schools were outpaced by their hostdistricts by as much 300 points. Eleventh grade math was a particularly weak area for charterschools in the host district comparisons, with the average school scoring 136 points below its hostdistrict. School-by-school comparisons, along with other relevant data, may be found in AppendixD:3. Once again, readers should refer to Section 11.2 to recall the limitations associated with thisanalysis.

Impact on Student Achievement: Analysis of Change Scores

Table 11:4 summarizes the analysis of two-year change scores. The table provides two sets ofnumbers. First, it shows the weighted average two-year change in the charter school�s PSSA scores.Positive values indicate improvements in scores while negative values indicate declines. The tablealso compares the two-year change in the charter school�s PSSA scores with the two-year changein the school�s host district. As before, we have constructed composite comparisons for charterschools with more than one host district. Details on this school-by-school and district-by-districtanalysis, along with more data and analysis, may be found in Appendix D:1.

Of the analyses presented in this chapter, Table11:4 provides the most positive picture ofPennsylvania charter schools. Indeed, there aremostly positive values (indicating improvementover time) for both the two-year change scoresand the comparisons of charter school and hostdistrict change scores. As a group, and averagedacross all portions of the PSSA, charter schoolscores improved 105 points. In doing so, charterschools outgained their host districts by some 86points.15

As discussed in Section 11.2, we encountered anumber of limitations in our attempt to fullyemploy this design. The most serious of theselimitations is the small number of schools andthe absence of consecutive year testing, whichmeans that the two-year comparisons involvetwo different groups of students. Readers shouldalso bear in mind, however, that this analysis is

157

C

S/H

D D

iffer

ence

in P

SS

A C

hang

e

CS/HD Difference in % Low Income-20 0 20 40

-100

0

100

200

300

KEC Math 8 (G)

KEC Read 8 (G)

KEC Math 11 (G)

KEC Read 11 (G)

KEC Math 8 (R)

KEC Read 8 (R)

KEC Math 11 (R)KEC Read 11 (R)

PCA Math 11

PCA Read 11

WC Write 6

WC Math 8

WC Read 8WC Write 9

Figure 11:1 Charter and Host District Change by Percent Low Income Student

based on a small sample. Thus, it is not at all clear that these findings can be generalized to charterschools as a group.

The validity of the comparisons of charter and host district change scores rests on the assumptionthat the two groups of schools are alike in most educationally relevant respects, except for charterschool status. As discussed in chapter 6, however, there is evidence that some charter schools enrollfewer poor and minority students than their host districts. To account for this fact, we comparedchanges in charter and host district PSSA scores with concentrations of low income students. Weattempted to make similar comparisons on race and other demographic factors, but could find nodata that were comparable across charter and noncharter schools.

Detailed results of this analysis may be found in Appendix D:1. Here we present a graph thatsummarizes and illustrates the results. Figure 11.1 shows the relationship between charter schools�relative change scores and their relative concentrations of low income students. The vertical axisrepresents the difference between charter schools� gain scores and those of the relevant host districts.As before, positive values (those above the horizontal �zero� line) represent charter schools whosegain scores outperformed their host districts�. Negative values (those below the horizontal zero line)represent charter schools that were outperformed by their host districts. The horizontal axis of thefigures shows the difference between charter schools and their host districts in the concentration oflow income students. Once again, positive values (those to the right of the vertical zero line)represent charter schools that have a higher percentage of low income students than their hostdistricts. Conversely, negative values (e.g., those to the left of the vertical zero line) represent

158

charter schools that have a lower percentage of low income students than their host districts. Eachpoint on the graph represents a school and PSSA test. Hence, �PCA Read 11" represents the changein 11th grade reading score for the Philadelphia Community Academy Charter School. Moreover,�WC� represents scores for the Philadelphia World Communications Charter School and �KEC� theKeystone Education Center Charter School. The figure includes two sets of comparisons for theKeystone Education Center, one with each of its two host districts (�R� for the Reynolds schooldistrict and �G� for the Greenville school district). The other school included in the change analysis,the Philadelphia Harambee Institute, did not report income data and is therefore excluded from thefigure.

Data points in the upper left quadrant represent tests in which the charter school had fewer lowincome students than their host district and outperformed the host district. Data points in the lowerright quadrant represent tests in which the charter school had a higher concentration of low incomestudents than their host districts and produced lower test scores. Data points in the upper rightquadrant represent tests in which the charter school both had a higher concentration of low incomestudents and outperformed the host district. Data points in the lower left quadrant represent tests inwhich the charter school had a lower concentration of low income students than the host district andwas outperformed by that district.

As the figure shows, most of the data points from the Keystone Education Center Charter School(KEC) fall in the upper-right quadrant. Thus, this particular charter school had a higherconcentration of low income students than the host district; nonetheless, it had change scores thatexceeded those of the host district. The Philadelphia Community Academy Charter School alsoappears in the upper-right quadrant. However, the charter-district difference in the concentration oflow income students is lower, as is the overall difference in change scores. Finally, the PhiladelphiaWorld Communications Charter School appears near the horizontal line to the left of the verticalline. This implies that the school had a lower concentration of low income students than its hostdistrict but very similar change scores. The Philadelphia Harambee Institute was outperformed byits host district in all but one subject/grade. Unfortunately, the school reported no income values.Thus, we are unable to evaluate its change scores in light of its student demographics or include itin the graph. Because of the paucity of data on changes in PSSA scores, we also examined other standardized testscores submitted to us by charter schools for evidence of change over time. During site visits to theschool, field researchers asked charter school Chief Administrative Officers to provide data fromassessments other than the PSSA. The request covered both commercially available standardizedachievement tests and other �home-grown�assessments. Fourteen schools provided such data fromseven different achievement tests. In all cases, however, the data were based on commerciallyavailable standardized achievement tests. We learned from our site visits that a number of schoolsare in the process of developing and using nonstandardized achievement tests. However, wereceived no such data from any of the charter schools. In a few cases, the test data were presentedin such a way that rendered interpretation difficult or impossible. We have excluded these data fromthe analysis. Table 11:5 describes the non-PSSA achievement test data we received from the charterschools.

16 There were inconsistencies in reporting formats even among schools using the same test.

17 As with the PSSA scores, the evaluation team did not have reliable estimates of the number of students testedin each grade and school available to it.

159

Table 11:5 Non-PSSA Achievement Data Provided By Charter Schools

Achievement Test Number ofSchools

ProvidingData

Iowa Test of Basic Skills (ITBS) 4Stanford Achievement Test (SAT) 3Metropolitan Achievement Test (MAT) 2Iowa Test of Educational Development(ITED)

1

New Standards Reference Exam 1TerraNova 1Wide Range Achievement Test (WRAT) 1

Source: Analysis of charter school records

Table 11:6 Non-PSSA Achievement TestData: Grade Equivalents

GradeLevel

School ID #

Percent of tests on whichstudents gained more

than 1 grade level

K - 6 1 0%2 7%3 70%

7 - 9 1 17%3 33%

10 - 12 1 33%

Given that we wished to use these data toevaluate charter schools� impact on studentachievement, we restricted our analysis tothose cases in which schools providedpre/post data. In most cases, this involvedadministering the test to students at thebeginning of the year and again at the end ofthe year. As such, this represents a strongerresearch design than the change score analysisprovided above, since (except for attrition)students given the posttest are the same groupgiven the pretest. Unfortunately, we were notable to obtain comparable data from hostdistricts and were not, therefore, able to add acomparison group to the pre/post design.Nevertheless, most testing companies reportgain scores in ways that facilitate comparisonswith national gain norms.

We received pre/post data from six charter schools. In order to further simplify the analysis, weexamined only those sections of the tests that assess students in reading, writing, or mathematics.Due to the difficulties involved in combining scores from different tests and the variety of reportingformats,16 we grouped the results according to reporting format (e.g., grade equivalents, nationalpercentile rankings, and the like) and school. Given that (unlike the PSSA) these test scores are notpublic information, we have elected to conceal the identities of the schools that provided them.Rather, we assigned each of the six schools a number. We present the results by grade level.

Table 11:6 provides scores expressed as gradeequivalents: the grade level at which the typicalstudent makes a given score. All of the gradeequivalent scores we received were from the IowaTest of Basic Skills. The first column of the tableindicates the grade levels tested. The secondcolumn provides a numeric identifier for the schooland the third column shows the percentage ofsubject areas in which students at the schoolachieved more than a full year�s gain.17 School #1,for instance, administered reading, language, andmath tests to all 6th graders. Students as a groupgained less than a year from pretest to posttest oneach of these subject tests. Hence, the rightmostcolumn reports a zero. This way of reporting test

160

Table 11:7 Non-PSSA Achievement Test Data: Normal Curve Equivalents

GradeLevel

SchoolID #

Percent of tests onwhich school gainedmore than 1 grade level

K - 6 4 40%

Table 11:8 Non-PSSA Achievement Test Data: Gain in Percent Achieving At Grade Level

GradeLevel

SchoolID #

Increase in percentstudents scoring atgrade level

K - 6 6 24%

Table 11:9 Non-PSSA Achievement Test Data: Percent Students Whose Scores Improved From Pre- to Posttest

GradeLevel

SchoolID #

Percent students whosescores rose from pre- to

posttest

7-9 6 72%

10-12 6 52%

results is admittedly crude, since it masks interesting variations among students. However, schools(understandably) did not share individual test score data with us.

Overall, Table 11:6 shows that students in the three schools reporting grade equivalent scores postedmore than a grade level�s gain on a substantial percentage of subject tests. In school #1, for instance,students� gains exceed a year in the junior high and high schools grades, though not in the elementarygrades. At school #2 student gains exceeded a year in just 7 percent of the subject areas, whilestudents at school #3 posted such gains on 33 percent of the subject areas tested.

Table 11:7 provides data on one school thatreported normal curve equivalents (NCEs),which indicate where in a (usually national)distribution of scores a given student orschool falls. The scores shown here comefrom the Metropolitan Achievement Test(MAT). In this school (#4) students postedmore than a grade level�s gain on 40 percentof the subject tests administered.

One school provided data from the Iowa Test of Educational Development, expressed as thepercentage of students scoring at or above grade level. Though similar, the data in Table 11:8 aredifferent from the data in the tables above, which present the percentage of students who gainedmore than one grade level. This table, instead, shows growth in the percentage of student scoringat grade level from the beginning to the end of the academic year. Hence, in this school 24 percentmore students were achieving at grade level at the time of the posttest than at the time of the pretest.

Finally, one school reported the percentage ofstudents tested whose scores rose from thepretest to the posttest. These findings arereported in Table 11:9.

From the commercial test data presentedabove, it is clear that some charter schoolstudents are making progress relative tonational norm groups. However, in theabsence of comparison groups given thesame test, it is difficult to draw evaluativeinferences from the non-PSSA data. Indeed,similar noncharter schools might have postedsimilar gains, weakening the case that thecharter schools caused these gains.Moreover, variations in the way the testscores were reported make it difficult to drawgeneralizations from the data. Given that wewere unable to get such scores for all charterschools (or draw a representative sample), itis doubtful that these results are

161

Me

an

PS

SA

Sco

re

% Students Eligible for Free/Reduced Lunch

0 25 50 75 100

1000

1200

1400

1600

Manchester

Mosaica

Centre

Nittany

Chester CS

Village

EMDHLaboratory

Figure 11:2 Predicted and Observed Math 5 Scores, 1998/99

representative of all charter schools. We must wait for more historical data on charter school studentachievement in order to provide a more certain assessment of their impact. Nevertheless, thecommercial test data provide some preliminary signs of improved student learning in charterschools.

Impact on Student Achievement: Comparisons With �Like� Schools

As discussed in Section 11.1, we used a regression model to generate �predicted� PSSA scores foreach level of income. By comparing these predictions with each charter school�s actual PSSAscores, we were able to provide readily interpretable comparisons between each charter school andall other Commonwealth schools in its demographic cohort. What we have called �income-adjusted�PSSA scores are simply the school�s actual score minus the score predicted by the model. Thus,positive values indicate schools that have outperformed their demographic cohort, while negativevalues indicate schools that have performed worse than their demographic cohort. As discussed inSection 11.2, this technique is subject to a number of limitations. Readers may find furtherdiscussion of the limitations to this approach, more detailed analysis, and other relevant data inAppendix D:2.

Figure 11:2 illustrates how we derived these income-adjusted PSSA scores using 5th grade math asan example. The vertical axis of the graph represents mean PSSA math scores (aggregated at theschool level). The horizontal axis represents the percentage of low income students in a given

162

Table 11:10 Income-Adjusted PSSAScores For Charter Schools(Regression Residuals)

Portion of PSSA Income-AdjustedScore (Residual)

Math 5 -88

Math 8 -69

Math 11 -4

Reading 5 -63

Reading 8 -69

Reading 11 28

Writing 6 -10

Writing 9 18

Mean (weighted) -46

school. Each of the small points represents a singlepublic school in Pennsylvania. The line (running fromupper left to lower right) shows the �regressionestimates��the set of PSSA scores predicted for eachconcentration of low income. Thus, to find thepredicted PSSA score for a given concentration of lowincome students, one simply picks the desired lowincome percentage on the horizontal axis and draws avertical line up from it. The point on the vertical axisat which that line intersects the diagonal line is thepredicted PSSA score for that income level. Thedownward slope of the line confirms the findings inthe literature that increases in the concentration of lowincome students are associated with decreases instudent achievement as measured on standardizedtests.

The graph also shows where the charter schools lierelative to other schools. Schools that lie well abovethe diagonal regression line posted scores better thanone would expect given their concentration of low income students. Put another way, these schoolsare doing better than other schools in their demographic group. Schools that lie well below thediagonal line are doing worse than one would expect relative to their demographic group.

Table 11:10 shows the average residuals�or the difference between the actual and predicted PSSAscores�for each portion of the PSSA for the 1997/98 and 1998/1999 academic years. A positiveresidual (above the regression line) indicates that the charter school outperformed the mean for itsdemographic group, while a negative residual (below the regression line) indicates that the schoolunderperformed relative to the mean of its demographic group. Generally, we may consider theresiduals as �income-adjusted PSSA scores.�

Taking all of the tests into account, charter schools scored an average of 46 points below otherschools with similar demographic characteristics. This suggests that charter schools generallyperform below schools with similar demographic characteristics. However, charter schools as agroup outperformed their peer schools in 11th grade reading and 9th grade writing. Indeed, on everyportion of the PSSA except 5th grade reading and 6th grade writing, there was at least one charterschool whose scores exceeded the mean of its demographic group. Charter schools fell just short,moreover, of the predicted values for 11th grade math. More detailed analysis, data, and graphs maybe found in Appendix D:2.

Figures 11:3 through 11:9 graph the regression residuals for the remaining portions of the PSSA.Given the small number of schools reporting data for the 1997/98 academic year, the graphs presentfindings from the 1998/99 academic year only.

163

M

ea

n P

SS

A S

core

% Students Eligible for Free/Reduced Lunch

0 25 50 75 100

1000

1200

1400

1600

Manchester

Mosaica

Centre

Nittany

Chester CS

Village

EMDH

Laboratory

Figure 11:3 Predicted and Observed Reading 5 Scores, 1998/99

Mean P

SS

A S

core

% Students Eligible for Free/Reduced Lunch

0 25 50 75 100

1000

1200

1400

1600 Nittany

Village

La Academia

Keystone

Phil. Comm.

World Comm.Ridgeview

Figure 11:4 Predicted and Observed Math 8 Scores, 1998/99

164

M

ean P

SS

A S

core

% Students Eligible for Free/Reduced Lunch

0 25 50 75 100

1000

1200

1400

1600

Nittany

Village

La Academia

Keystone

Phil. Comm.

World Comm.Ridgeview

Figure 11:5 Predicted and Observed Reading 8 Scores, 1998/99

Mean P

SS

A S

core

% Students Eligible for Free/Reduced Lunch

0 25 50 75 100

1000

1200

1400

1600

NortheastKeystone

SUSQ-CYBER

Phil. Comm.

Ridgeview

Figure 11:6 Predicted and Observed Math 11 Scores, 1998/99

165

M

ean P

SS

A S

core

% Students Eligible for Free/Reduced Lunch

0 25 50 75 100

1000

1200

1400

1600

Northeast

KeystoneSUSQ-CYBER

Phil. Comm.

Ridgeview

Figure 11:7 Predicted and Observed Reading 11 Scores, 1998/99

Mean P

SS

A S

core

% Students Eligible for Free/Reduced Lunch

0 25 50 75 100

1000

1200

1400

1600

Mosaica

Centre

VillageLa Academia EMDH

World Comm.

Figure 11:8 Predicted and Observed Writing 6 Scores, 1998/99

166

M

ean P

SS

A S

core

% Students Eligible for Free/Reduced Lunch

0 25 50 75 100

1000

1200

1400

1600

VillageSUSQ-CYBEREcon & LawImhotep CS

Multi-culturalPreparatory

World Comm.

Figure 11:9 Predicted and Observed Writing 9 Scores, 1998/99

11.4 Recommendations for Improving the Evaluability of Charter Schools� Impact on Student Achievement

Throughout this chapter, we have stressed that the findings from the analysis of achievement testsare suggestive but in no way conclusive. In a real sense, evaluation findings are rarely, if ever, freefrom uncertainties. Nonetheless, the nature and quality of the data available to the evaluation teamseverely limits the validity�and ultimately, the utility�of the findings. Pursuant to this evaluation�sformative orientation, we close the chapter by offering some preliminary suggestions for improvingdata collection practices that would improve the validity and utility of future evaluations ofPennsylvania�s charter school program. We emphasize that these suggestions are preliminary.Before undertaking any significant changes in data collection and reporting systems, policymakersshould seek the advice from stakeholders and other experts in educational evaluation.

Before continuing, it is appropriate to note the efforts and accomplishments to date of thePennsylvania Department of Education, and its Office of Educational Initiatives, in the developmentof its accountability plan for charter schools and in the collection of school-based data from charterand noncharter public schools alike. Pennsylvania�s practices in this area are far superior to thoseof other charter school policies this evaluation team has studied. Like the charter schoolsthemselves, the Office of Educational Initiatives (OEI) has faced the considerable challenge ofdesigning accountability systems �from the ground up.� We appreciate these difficulties. Indeed,many of the problems encountered in this evaluation are rooted in the properties of current data

167

collection and reporting procedures that were in place before the charter school reform was startedin the state. Nonetheless, it is clear that more efforts are required if future evaluations are to providereasonably certain assessments of charter schools� effectiveness.

Fortunately, some of the problems identified in this chapter (e.g., small sample sizes, the fact thatall schools are still in the early phases of their development) will be self-ameliorating as moreschools are established and as the schools accumulate more years of data. Other problems, such asthe inability to provide assessments of student achievement levels prior to enrollment in a charterand the inability to conduct truly longitudinal analyses, will require proactive approaches.Policymakers might consider two general approaches. The first involves revisions to the PSSA andwould no doubt be costly and time-consuming. The second, perhaps less costly, approach involvesthe use of commercial standardized tests.

States at the forefront of accountability and evaluation (e.g, Tennessee) test every student every year.Such a system allows evaluators to conduct genuine longitudinal analyses that follow the samestudents from one grade to the next. As noted in this chapter, the Pennsylvania system, by contrast,allows evaluators to compare only this year�s students from the 5th grade (and a few other grades)with last year�s 5th graders. Accountability systems in leading states also track individual studentsby assigning to each a confidential numerical identification number. This would allow charterschool evaluators to examine the achievement level of each student prior to enrolling in a charterschool. Analyses based on individual student gains are more effective in identifying the value addedby the school. Individual student data by itself would vastly improve the quality of assessments ofcharter school impact on student achievement. Such a system would also allow evaluators to takebetter account of student attrition, thus improving the quality of longitudinal analyses.

Unfortunately, developing such systems is extremely costly and time-consuming (see, e.g., Millman,1997). It is unlikely that any such changes would be in place in time for the legislatively-mandated5 year charter school evaluation. One relatively inexpensive practice PDE could develop is to reportmore distributional properties of PSSA school scaled scores. Measures of dispersion (such asstandard deviations) and skewness would allow evaluators to more responsibly interpret mean scaledscores.

Another less expensive approach involves commercial standardized tests. Most charter schools usethe ITBS, the SAT or other standardized tests. Many do so in order to fulfill local districtaccountability requirements. Moreover, many schools employ pre/post testing designs that can bevery useful in assessing charter school impact. As noted in this chapter, however, the evaluationteam found it difficult to get such data from a wide range of schools. When it did receive the data,they came in a myriad of formats that made it difficult to construct reliable summary indicators ofstudent progress. In order to maximize the value of such information for evaluations, OEI mightconsider developing guidelines for the reporting and formatting of commercial achievement test data.Even if OEI does not wish to mandate a particular test, schools usually have a wide of range ofchoices about how scores are reported back to them. It is possible that charter school officials wouldbe willing and able to choose formats that accommodate the needs of state-level accountability actorsand evaluators. As always, any such policy must balance the demands of accountability against thevalue of school autonomy.

18 Readers should note that the table excludes the comparisons with all noncharter schools presented in section11.3 of the chapter. Such comparisons are useful in gauging the absolute performance levels of charter school students.Since they do not account for differences in student background characteristics, however, they are of little or no valuein assessing charter school impact. The table also excludes data from commercial tests, since the sample of casesavailable to the evaluation team precluded the construction of quantitative summary indicators appropriate to the table.

19 The PSSA scale ranges from 1000 to 1600. 168

11.5 Summary and Conclusions

This chapter of the evaluation has addressed questions related to charter schools� impact on studentachievement. In particular, it has sought to assess whether the design and implementation of theautonomy-accountability �bargain� explored in other chapters has led to improvements inachievement and whether charter schools have shown improvements in achievement over the courseof their existence. Data limitations precluded answers to an evaluation question about the impactof expenditure levels of student outcomes.

As discussed at great length in the chapter, the findings presented are subject to serious limitations.Thus, any and all findings should be regarded as suggestive but in no way conclusive. The mostimportant limitations included the following:

! Pennsylvania�s charter school policy is still quite young; moreover, data collection limitationsconstrained the evaluation team to examine data that are in many cases over a year old. Thisleft charter schools with little time to demonstrate their ability to improve student achievement.

! Data on only a small number of charter schools were available to the evaluation team. Thus,generalizations to the broader movement are tenuous.

! No data existed on such important characteristics as students� precharter achievement rates,which seriously diminished the evaluation�s ability to provide valid assessments of charterschool impact.

Keeping these caveats in mind, the chapter offered a number of findings, which are summarized inTable 11:11.18 The table�s first column presents an aggregate summary of charter schools� 2-yearchange scores; the second column presents comparisons of charter school and host district changescores. By far, these comparisons provide the most positive picture of charter school impacts onstudent achievement. As a group and averaged across subject areas, PSSA scores for the 4 schoolsfor which data were available grew by more than 100 points from the 1997/98 to 1998/99 academicyears�an increase of 86 points more than the host districts.19 Data from commercial achievementtests also show evidence of student gains. Unfortunately, these analyses are subject to seriouslimitations. The PSSA change score analyses were based on only 4 schools; moreover, the data didnot permit true longitudinal analysis, leaving open the possibility that the observed gains were dueto changes in schools� study profiles. The commercial test data permitted relatively clean pre/postcomparisons but represented a small and self-selected sample of charter schools. Findings from theremaining analyses�comparisons with host districts and other �like� charter schools�were based ona large number of schools. The results of these analyses were far less positive than the change score

169

analyses. The third column of Table 11:11 shows that, as a group and averaged across subject areas,charter schools were outperformed by their host districts by a factor of 50 points. Similarly,comparisons with �like� noncharter schools shows that charter schools were outperformed by 46points.

Table 11:11 Summary of PSSA Findings

Portion ofPSSA

Comparison

2-Year Change(charter only)

2-Year ChangeCompared to Host

District

Comparisons toHost Districts

Comparisons With�Like� Noncharter

SchoolsMath 5 40 40 -6 -95Math 8 146 102 -89 -69Math 11 99 99 -136 -4Reading 5 -120 -150 -25 -63Reading 8 222 197 -68 -69Reading 11 134 152 -88 28Writing 6 33 30 -15 -10Writing 9 50 -10 38 18# Charter schools 4 4 21 21Mean 105 86 -50 -46Source: Analysis of PDE raw data files by The Evaluation CenterNote: Means are weighted by average daily membership.

It is important to note that in each of the comparisons there were both strong and weak schools.Figures 11:2 to 11:9, for instance, show that for almost every subject section of the PSSA there wereschools that outperformed the mean of a set of demographically similar schools. Similarly, severalcharter schools outperformed their host districts, sometimes by wide margins.

In the end, the evidence for charter schools� impact on student achievement is mixed, with differentcomparisons showing different results. Other things being equal, most evaluators would put morestock in the change score analysis than the others. This is because each school serves as its owncomparison group. However, the fact that the change score analysis relied on data from only 4schools casts doubt on the generalizeability of these findings. Only time and more data will showwhich set of findings is more correct.

Looking to the future, policymakers should consider improvements in data collection and reportingpractices relevant to charter schools. Section 11.4 of this chapter provided a set of preliminaryobservations on this score. Such improvements would enhance the accuracy and utility of the year5 summative evaluation. Policymakers might wish to add a systematic examination of variations incharter school student achievement to future evaluation activities. This chapter has identified anumber of particularly strong and weak schools. A small number of in-depth case studies might

20 Such case studies are a hallmark of the �effective schools� tradition of school effects research. See Teddlieand Stringfield (1993) for a recent example of this work.

170

provide more thorough explanations for these differences.20 Nor would this be simply an academicexercise. Findings from such case studies might identify constraining and enabling factors that couldinform program revisions. Moreover, such findings might identify a set of best practices and providea medium for the diffusion of innovations among charter schools and between charter and noncharterschools.

Finally, student achievement is not the only relevant student outcome for charter schools. In the nextchapter we examine a number of nonachievement outcomes, including market accountability (theextent to which educational �consumers� �vote with their feet� for charter schools), attendance, andvarious elements of charter school climate.

171

Chapter TwelveOther Indicators of School Quality

We noted in chapter 11 that there is legitimate debate about precisely what kinds of student outcomescharter schools should be held accountable for. In that chapter we suggested that, in spite of thevariety of positions in the debate, most stakeholders would agree that student achievement is animportant (if not the only) outcome. In this chapter we attempt to assess charter schools on a numberof alternative criteria. Thus, as with the previous chapter, the evaluation question addressed in thischapter is: Does increased flexibility in exchange for increased autonomy result in improved pupilresults? The difference between the two chapters comes in the definition of “pupil results.” Whereaschapter 11 examined student achievement, this chapter assesses other student outcomes.

The chapter begins by supplementing the view of student achievement provided by standardized testswith the assessments of charter school students, parents, and teachers. Next, the chapter turns to the“market accountability” viewpoint by examining the extent to which educational “consumers” have“voted with their feet” for or against charter schools. We do this by examining patterns of entry intoand exit out of charter schools and of overall satisfaction with charter schools. Indeed, on the marketaccountability view, charter schools should be judged not only on their ability to realize certainpublicly pronounced standards and goals, but also on their ability to satisfy consumer preferences.Finally, the chapter examines a number of intermediate or “nonachievement” outcomes, includingattendance, discipline, and various elements of school culture and climate. These latter outcomes aresignificant in two respects. First, some stakeholders consider them ends in themselves. Others,however, view them as necessary prerequisites to the ultimate goal of increasing student achievement.On this view, even if a charter school has not demonstrated gains in student achievement, its abilityto improve student learning habits and school climate might augur well for future improvements inachievement. In short, such outcomes might serve as “leading education indicators” of potentialeducational performance, in much the same way that the Index of Leading Economic Indicatorsserves as a way to predict future economic performance.

12.1 Data and Methods

The data used in this chapter come primarily from surveys of charter school students,parents/guardians, and faculty/staff. Respondents from each group were sampled randomly (seechapter 2 for details). In order to facilitate comparisons among charter and noncharter schools, weused data from the National Association of Secondary School Principals’ School Climate Survey(SCS). The SCS provides national norms on a number of constructs related to school culture andclimate. Since the SCS was not designed specifically for charter schools, we devote most of thechapter to the other surveys, which were designed explicitly for charter schools. Complete survey

172

Student Performance at Previous and Charter School

0%

20%

40%

60%

80%

100%

Excel-lent

Good Average Poor Unsatis-factory

Previous school Charter school

Figure 12:1 Performance at Previous and Charter Schools

results may be found in Appendix B. Additional data come from raw data files provided by PDE andfrom archives of charter school documents generated and maintained by the evaluation team. Theseinclude annual reports submitted by the charter schools to PDE. Data on charter schools in thePhiladelphia school district come form the district policy and planning office.

These data, however, are subject to a number of limitations that readers should bear in mind whenmaking decisions based on them. First, much of the information presented in this chapter is based onrespondents’ subjective judgments of their schools. In most cases, the evaluation team could notverify these assessments. The second limitation involves comparison groups. Many of the surveyitems invite respondents to compare their charter school experiences with experiences at otherschools. Thus, the surveys attempt to simulate pre/post comparisons on a single cohort. However,such comparisons might be tainted by faulty recollections. Moreover, the fact that the sampleincluded only participants in charter schools makes it impossible to gauge whether reported changesin opinion are unique to charter schools. Finally, given that the evaluation took place during the earlystages of the Pennsylvania’s charter school movement means that the findings are based on arelatively small number of schools. Of these schools, the surveys were administered only to schoolsenrolling grades 5 and higher due to ethical and technical issues surrounding gathering data fromsmall children. Thus, readers should be cautious in generalizing these findings to the movement asa whole.

12.2 Subjective Ratings of Student Performance

The student survey instrument included two items that asked students to rate their own performanceat the charter school and to compare it with their performance at their previous school. Of the 1,106students surveyed during the 1999/2000 academic year, just under 65 percent believed that theirperformance was either “excellent”or “good.” This representsvirtually no change from theprevious year. The survey alsoasked students to compare theirperformance in the charter schoolwith their performance at theirprevious school on a scale thatincluded “excellent,” “good,”“ a v e r a ge , ” “ p o o r , ” a n d“unsatisfactory.” Results from thisitem are presented in Figure 12:1.Whereas 12.8 percent of surveyedstudents judged their performanceas “poor” or “unsatisfactory” attheir previous school, only 8percent placed themselves intothese categories at their charter

1 A Wilcoxon Sign Test yielded a p-value of 0.001, indicating that is it is very unlikely that this observeddifference is the result of sampling error.

173

schools, a difference of just under 5 percent. Moreover, while 61.3 percent of students judgedthemselves as “excellent” or “good” at their previous school, 64.7 percent placed themselves intothese categories at their charter schools. In response to open-ended survey questions, students oftenoffered comments such as, “I can have help whenever I need it. My grades improved from myprevious school.” The magnitude of these changes is modest, to be sure. Nonetheless, they representstatistically discernible changes.1 The surveys revealed similar findings for the 1998/99 academicyear. We also tested for differences in perceptions between students who had previously attendedpublic and nonpublic schools but found no discernible differences.

The teacher survey also included an item that asked teachers to assess the extent to which “theachievement levels of students are improving.” The question asked teachers to provide both theirinitial expectation and their current experience. In 1999/2000, 55 percent of those teachers surveyedthought that student achievement was, indeed, improving at their charter school–up slightly from 53.9percent in 1998/99. However, the surveys also revealed that a higher percentage of teachers hadexpected students to show gains in achievement (70 percent in 1998/99 and 74 percent in 1999/2000).We cannot determine whether the gap between expectations and reality is due to inflated expectationsor genuine problems with student achievement.

As the findings in Chapter 10 highlight, many parents report enrolling their children in charter schoolsbecause they address student needs that could not be addressed at other local schools. The charterschools get especially high marks from parents who said that their students’ special needs were notwell served in other schools which suggests that parents with students with special needs are generallysatisfied with the progress their children are making in charter schools. Readers are referred tochapter 10 for more details.

12.3 Market Accountability

Market accountability holds that charter school quality is best measured by whether its consumersare satisfied with the “product” offered by the school. Economists and others who study marketbehavior typically measure consumer satisfaction through their “revealed preferences”; that is,through the choices they actually make. Thus, the decision to seek enrollment in a charter schoolmight, other things equal, be an indicator of school quality. Conversely, the decision to leave acharter school might be, among other things, an indicator that the school is not performing well.

Waiting lists provide an indication of whether local educational “consumers” consider charter schoolsof sufficiently high quality to enroll their children. With the exceptions of siblings and the childrenadults involved in the founding of a school, charter schools must select students randomly from theirwaiting lists in the event that they are oversubscribed (sec. 1723-A). Unfortunately, we could obtaininformation on waiting lists from only 19 charter schools. Table 12:1 shows that the average charter

2 PDE data files include measures of enrollment stability for grades 5, 8, and 9. Unfortunately, the mostrecent available data come from the 1998/99 academic year. Since most of the charter schools we studied openedduring that year, these data were of little use in this analysis.

3 One school has a goal of returning students to noncharter public schools as soon as possible, which mightresult in a high turnover rate. However, we did not have turnover data on this school and did not include it in theestimates discussed in this chapter.

174

school had 125 students on its waiting list as of Summer 2000. Since the number of students on aschool’s waiting list must be understood in terms of the overall number of slots, we also present theaverage number of students on waiting lists as a percentage of current enrollment. Table 12:1 showsthat the average waiting list is 45 percent of the total number of slots currently in the school.However, there was considerable variation around these values. Indeed, some schools report nostudents on their waiting lists, while others report waiting lists that exceed total current enrollment.

Table 12:1 Charter School Waiting Lists, Summer 2000

Mean Median Minimum Maximum

# Students on Waiting List 125 70 0 800

# Students as Percentage ofTotal Enrollment 45 35 0 132

Source: Annual reports submitted by charter schools to PDE’s Office of Educational Initiatives.

In order to assess student turnover, we solicited student rosters from charter schools (collected aspart of the student sampling protocol) for both the 1998/99 and 1999/2000 academic years. Byexamining changes in these rosters we were able to provide rough estimates of student turnover fromone year to the next.2 Unfortunately, we were able to use the rosters for this purpose for only 8 ofthe 30 schools we visited. While we cannot assess whether these schools constitute a representativesample of the others, they provide at least a preliminary view of student turnover. As a group, the8 charter schools examined had a return rate of 62 percent, indicating that 38 percent of student whoenrolled in the charter schools as of Spring 1999 were not enrolled in the same school as of Spring2000. There was, however, tremendous variation among the charter schools, with 79 percent ofstudents returning in 1 school and 16 percent in another.3

As an additional piece of evidence, we obtained records from the Philadelphia Public Schoolsindicating the number of charter school students returning to the system’s public schools during the1999/2000 academic year. We chose not to include these in the estimates discussed in the previousparagraph because the district had no information on students leaving charter schools to attendnonpublic schools. The Philadelphia records indicate that a total of 1,834 students dropped out ofthe district’s charter schools to return to the district’s other public schools. Figure 12:2 shows thatmost of the returning students moved during the first few months of the academic year.

175

253

575

243154 6795105 122 110110

0

200

400

600

800

Sept

OctNov Dec Ja

nFeb

March

April

MayJu

ne

Num

ber o

f Stu

dent

s

Figure 12:2 Number of Students Returning from Charter Schoolsto Philadelphia Public Schools, 1999/2000

Would you recommend this school to a friend?

0%

20%

40%

60%

80%

100%

No Yes Not sureFigure 12:3 Students Responses Regarding Whether or Not

They Would Recommend Their Schools To Others

Another way to assesscustomer satisfaction issimply to ask importantstakeholders for theirsubjective assessments.Thus, we turn next to thesubjective evaluations ofcharter schools� primarycustomers�parents andstudents. Parentsindicated their level ofoverall satisfaction withtheir charter schools inthree main areas:instruction, curriculum,a n d s t u d e n tp e r f o r m a n c e .Approximately 95percent of parentsreported that the qualityof instruction in theircharter school is high.Similarly, some 90percent said that theirchild received sufficientattention. In order tog a u g e s t u d e n tperceptions of overallschool quality, thes t u d e n t s u r v e yins t ru m e n t a skedstudents whether theywould recommend theircharter school to a friend. Approximately 50 percent of students surveyed indicated that wouldrecommend the school to a friend, while 26 percent said they would not and another 23 percent werenot sure (see Figure 12:3). Answers to the same question in 1998/99 were virtually identical.

12.4 Attendance and Discipline

The remainder of this chapter considers a number of elements of the culture and climate in charterschools. Since stakeholders are particularly interested in attendance and discipline issues, we treatthese in a separate section. We discuss a number of other school climate issues in the next section.

4 The average attendance rate for all Commonwealth schools has remained similarly stable at approximately93 percent from 1995/96 to 1998/99.

5 Of those schools reporting rates in both years, the average change was a growth of 1 percentage point,which would confirm the finding of no substantial change in the medians.

176

In order to investigate attendance, we calculated average charter school attendance rates for the1998/99 and 1999/2000 academic years. Table 12:2 reports both the mean (weighted by enrollment)and median values, as well as minimum and maximum values. The table reveals that there wasconsiderable variation in attendance among charter schools, especially during the 1998/99 academicyear. In both years, the school with the best attendance rate reported a 98 percent average, whereasthe school with the lowest rate reported averages of 32 percent in 1998/99 and 60 percent in1999/2000. Given that relatively few of the schools reported such low averages, we recommend thatreaders rely on the median score, which is less sensitive than the mean to outliers. Taking this intoaccount, the table shows that the average attendance rate remained stable across the two years, witha median of 90 percent in both years.4 Unfortunately, only 10 schools reported attendance rates forboth years. Given, therefore, that estimates for the two years are based on different cohorts ofschools, we must exercise caution in speculating about any trends in attendance rates.5

Table 12:2 Attendance Rates in Charter and Noncharter Schools Compared, 1998/99-1999/2000

N Mean Median Min. Max.

Charter Schools, 1998/99 20 89 90 32 98

Charter Schools, 1999/2000 21 84 90 60 98

All Other Public Schools, 1998/99 3,049 93 94 51 100Source: Estimates for the entire Commonwealth and 1998/99 charter schools were derived from PDE raw data files.Estimates for 1999/2000 were derived from annual reports charter school submit to PDE’s Office of EducationalInitiatives. Note: Means are weighted by average daily membership for the 1998/99 charter school and all Commonwealthestimates. Means for charter schools in 1999/2000 are weighted by attendance, due to our inability to get estimatesfor average daily membership. Due to missing cases, the estimates for charter schools are not based on the samecohorts of schools. Readers should exercise caution in interpreting two-year trends. Because the charter schools arelargely urban school, a fairer comparison would be made with only urban schools and not the average for theCommonwealth. Unfortunately, we could not make such a comparison with the data available.

We were unable to locate reliable data on student discipline incidents in charter schools. AvailablePDE data files on such incidents include only a few charter schools, and then only for the 1997/98academic year. Thus, we rely on the subjective assessments of students, teachers, and parents.

In order to elicit students’ general appraisal of school safety, the survey instrument asked them toexpress the extent to which they agree with the statement, “Students feel safe at this school.” Ofthose students surveyed in Spring 2000, approximately half (52 percent) either agreed or strongly

6 A t-test assuming unequal variances yielded a p-value of 0.13. All t-tests are constructed using the meansof the distributions of rating scores.

177

Students Feel Safe at This School

52

80 84

0

20

40

60

80

100

Students Parents Teachers

Perc

ent A

gree

ing

Figure 12:4 Discrepancy Between Students’, Parents’, and Teachers’, Perceptions of School Safety

agreed. This represents amodest increase from theprevious year. Thepercentage of students whoagreed that studentsrespect one another andtheir property was muchlower, and declined from29 percent in 1998/99 to24 percent in 1999/2000.This two-year decline didnot reach conventionallevels of statisticalsignificance, however.6

Thus, it appears thatstudent perceptions ofsafety in charter schools have remained relatively stable, with perhaps a small decline from one yearto the next.

However, there was substantial variation in student perceptions of safety. At one school, as manyas 78 percent of students agreed that their school was safe. In other schools, the percentage was aslow as 15 percent. Somewhat surprisingly, at least some of the schools with high student safetyratings are in the city of Philadelphia.

Interestingly, a much higher percentage of teachers and parents believe that students feel safe at theircharter schools than do the students. Indeed, some 84 percent of teachers and 80 percent of parentsagreed with the statement that “students feel safe at school.” There was virtually no change from the1998/99 to the 1999/2000 academic years. Figure 12:4 illustrates the discrepancy in perceptions ofschool safety.

12.5 Other Attributes of the School Environment

The surveys asked respondents for their opinions on a wide range of school culture and climate issuesbeyond school safety. One set of questions dealt with whether charter school students are held tohigh expectations. Approximately 77 percent of parents surveyed in 1999/2000 agreed with thestatement that “This school has high standards and expectations for students.” A similar percentageof teachers (72 percent) also agreed with this statement. The surveys also attempted to assessstudents’ perceptions of expectations in less direct ways. Some 69 percent of students surveyed in1999/2000 agreed with the statement that “My teachers encourage me to think about my future,” a

178

Students at This School Are More Interested in Learning Than They Were At My Previous School

31%38%

0%

20%

40%

60%

80%

100%

Public Nonpublic

Perc

ent A

gree

ing

Figure 12:5 Agreement with “Students are More Interestedin Learning” by School of Origin

slight increase from theprevious year. Fewerstudents (approximately 50percent) said that they havemore homework at theircharter school than at theirprevious school. Thisrepresents a statisticallydiscernible decline from 45percent in the previous year. Students affirmed thesefindings in their responses toopen-ended questions. Onewrote: “You are stronglymotivated and every teachersupports [and] helps you.”Another student suggested, “The thing I like most about this school is the dedication of the teachersand the extra push they give students.” Yet another student wrote that “The teachers care about youand your future; they don't give up on you unless you give up on your self.”

The surveys also sought to assess important aspects of charter schools’ learning climates and studentattitudes about learning. Approximately 37 percent of students sampled in 1999/2000 agreed withthe statement “Students at this school are more interested in learning than they were at my lastschool.” Interestingly, 38 percent of students from public schools agreed with the statement,compared with 31 percent of students from nonpublic schools (see Figure 12:5). A statisticalsignificance test (t-test) on the score means revealed that this difference was statistically discernible(p=0.008). Asked specifically about their own attitudes toward learning, even more students (44percent) agreed that they were more interested in learning at their charter school (see Figure 12:6).One student cited the “Ability to have fun and learn. The school has lots of strengths, [and an] abilityto teach students well . . .” There was no difference between former public and nonpublic studentsin answers to this question. Finally, there was a substantial decline (37 to 29 percent) from 1998/99to 1999/2000 in the percentage of students who thought that students would continue their studieseven after the teacher left the room. This difference was statistically discernible at conventional levels(p=0.02). The parent and teacher surveys also asked for assessments of student attitudes towardlearning. Approximately 75 percent of parents said that their child is “motivated to learn.” Teachers,however, were much less sanguine, with only approximately 30 percent agreeing with the statement.

Next, the surveys sought to assess whether charter schools exhibit a community atmosphere. Over57 percent of students sampled in 1999/2000 agreed with the statement that “Teachers want me tobe in school and ask me why I wasn’t there when I have been absent,” up slightly from 54 percentin the previous year. Even more students (approximately 78 percent) indicated that teachers andadministrators know them by name. Moreover, some 71 percent agreed with the statement “My

179

Interest in School Work Compared to Previous School

43.8%37.4%

18.7%0%

20%

40%

60%

80%

100%

Moreinterested

About the same

Lessinterested

Figure 12:6 Students’ Self-Rated Levels of Interest Compared to Previous School, 1999-00

teacher is available to talk tome about my classroomperformance.” Parents weresimilarly positive about thecommunity atmosphere incharter schools, withapproximately 70 percentagreeing with the statement,“This school reflects acommunity atmosphere.”

Finally, the surveys queriedrespondents about charterschools’ physical facilities.Approximately 41 percent ofstudents agreed with thestatement, “The schoolbuilding is clean and well maintained,” down slightly from 44 percent in the previous year (thedifference was not statistically discernible). Similarly, some 44 percent of parents agreed with thestatement, “This school has good physical facilities,” down from 50.2 the previous year. Thisdifference was statistically discernible at the 0.09 level. Many students had additional commentsconcerning the state of their building and environment. One student, for instance, stated that “Thebuilding is small and we don't have many extra-curricular activities like other high schools do.”

12.6 Comparisons with National Norms

The data presented above help to illuminate whether charter schools are successful in pursuingnonachievement outcomes. Without benchmarks or comparison groups, however, much of thedata–especially those on school climate–are difficult to interpret. Unfortunately, administering thesurveys to noncharter students, parents, and teachers was well beyond the scope of this evaluation.We were able to partially address this limitation by administering the National Association ofSecondary School Principals’ School Climate Survey (SCS) during the 1999/2000 academic year.The SCS is unique in that it provides national norms on a number of constructs relating to schoolclimate. Since many of the items on the SCS are similar to those on the surveys already discussed,we will not present a detailed review of findings from the SCS. Instead, we report summary indicesof 10 subscales measured by the SCS, along with national norms on these same indices. We have notreported parents’ responses to the SCS due to low response rates. Appendix B contains results fromthe teachers/staff and students on the School Climate Survey.

Findings from the administration of the SCS, along with the national norms, are summarized in twographs. Figure 12:7 provides charter school scores and norms for students; Figure 12:8 provides thesame for teachers and staff. The ten subscales are identified on the horizontal axis, while theaggregate scores for all Pennsylvania charter schools are shown on the vertical axis. For each

180

subscale, the national norm is 50. Hence, points above the horizontal line at 50 indicate subscales onwhich sampled charter schools exceeded national norms, while points below the horizontal lineindicate subscales on which sampled charter schools performed below national norms. Table 12:3provides brief descriptions of each of the ten subscales.

Figures 12:7 and 12:8 show that Pennsylvania charter schools scored above national norms inapproximately half of the subscales (4 for students and 5 for teachers/staff). Charter schools areparticularly strong in teacher-student relationships, administration, and student academic orientation.However, charter schools in the sample scored below national norms on parent- and community-school relationships. This is somewhat surprising given that the norm group includes many largeschools that are unlikely to have the strong connections with the community that many charterschools appear to enjoy. There is substantial agreement between teachers and students on thesesubscales.

Table 12:3 Descriptions of Subscales on the School Climate Survey

Subscale DescriptionTeacher-student relationships The quality of the interpersonal and professional relationships between

teachers and students

Security and maintenance The quality of maintenance and the degree of security people feel at theschool

Administration (principal, asst.principal, etc.)

The degree to which school administrators are effective incommunicating with different role groups and in setting highperformance expectations for teachers and students

Student academic orientation Student attention to tasks and concern for achievement at school

Student behavioral values Student self-discipline and tolerance for others

Guidance The quality of academic and career guidance and academic counselingservices available to students

Student-peer relationships Students’ care and respect for one another and their mutual cooperation

Parent and community-schoolrelationships

The amount and quality of involvement of parents and other communitymembers in the school

Instructional management The efficiency and effectiveness of teacher classroom organization anduse of classroom time

Student activities Opportunities for and actual participation of students in school-sponsored activities

Source: Examiner’s Manual, School Climate Survey

There is significant disagreement in teachers’ and students’ perceptions in three areas. First, students’perceptions of school safety and maintenance were well below national norms, while those of teacherswere only slightly below those norms. This reflects a similar difference revealed by the charter schoolsurveys discussed above. Second, teachers give lower ratings to student behavioral values thanstudents do. Such a difference is perhaps to be expected; indeed, one might expect this

g y

30

40

50

60

70

Teacher-Student

Relationships

Security andMaintenance

Administration(Principal,

Assist.Principal, etc.)

StudentAcademicOrientation

StudentBehavioral

Values

Guidance Student-PeerRelationships

Parent andCommunity-

SchoolRelationships

InstructionalManagement

Student Activities

Stan

dard

Sco

res

National norms designated by the 50th percentile

Figure 12:7 Students’ Results on the Nationally-Normed School Climate Survey, 1999-2000

g y

30

40

50

60

70

Teacher-Student

Relationships

Security andMaintenance

Administration(Principal,

Assist.Principal, etc.)

StudentAcademicOrientation

StudentBehavioral

Values

Guidance Student-PeerRelationships

Parent andCommunity-

SchoolRelationships

InstructionalManagement

Student Activities

Stan

dard

Sco

res

National norms designated by the 50th percentile

Figure 12:8 Teachers’ and Staff Results on the Nationally-Normed School Climate Survey

183

to be the case at most schools, including those in the norm group. However, teachers’ ratings arelower even compared against national norms, suggesting a genuine difference between charter andnoncharter schools. Finally, students rate the availability of student activities higher relative tonational norms than their teachers.

12.7 SummaryThere is legitimate debate about precisely what types of student outcomes charter schools should beheld accountable for. While most stakeholders seem to agree that student achievement is animportant (if not the only) goal of charter schools, others argue that schools should also be judgedon their ability to satisfy their customers. Hence, this chapter provides a preliminary analysis of self-rated performance.

Survey evidence suggests that as a group charter school students believe that their level of academicperformance has improved since they moved to a charter school. Teachers are perhaps a little lesssanguine; most teachers expected student achievement to improve more than it has. Still, more thanhalf of teachers believe that student achievement is on the rise at their school. Moreover, parents andteachers believe that charter schools are serving needs not well served by other schools.

The chapter also examined various indicators of market accountability, or the extent to which parentsand students vote with their feet for or against charter schools. A nonrandom sample of charterschools indicates, for instance, that the average school has a waiting list of 125 students, or 45percent of total current enrollment. However, just as there are students waiting to get into charterschools, a number have left charter schools to return to other public and nonpublic schools. Anonrandom sample of student rosters indicates that the average charter school lost 38 percent of itsstudents from Spring 1999 to Spring 2000. Moreover, data from the Philadelphia school districtindicates that some 1,800 students left Philadelphia charter schools to return to district schools. Wemust emphasize, however, that available evidence on charter school entry and exit is not based onrepresentative samples. Moreover, we have no way of knowing whether decisions to leave charterschools are motivated by concerns about school quality or other reasons. In spite the significantenrollment instability in some charter schools, more than 90 percent of parents surveyed believe thatthe quality of instruction in their charter school is high and that their child receive adequate attention.Approximately half of the students surveyed said that they would recommend their charter school toa friend.

Finally, the chapter examined a number of indicators designed to capture charter schools’ climatesand cultures. Many of these factors might indicate the potential for high achievement levels in thefuture. For instance, charter schools had an estimated attendance rate of 90 percent for both 1998/99and 1999/2000. Moreover, nearly three-fourths of parents and teachers reported that their charterschools had high expectations for students. A similar percentage of students said that their teachersencourage them to think about their future. However, a much smaller percentage (one-third) saidthat students at their charter school were more interested in learning that students in their previousschool. Fewer than one-half of respondents report being fully satisfied with school facilities.

184

Chapter ThirteenSummary of Findings and Issues

for Further Consideration

This report has provided a largely formative evaluation of Pennsylvania’s charter school policies.We have sought to identify strengths and weaknesses with an eye toward providing a factual basisupon which policymakers and stakeholders might improve the program. As such, the report takesno position on whether there should be a charter law in Pennsylvania. This issue will be taken upin a more summative evaluation due in the Fall of 2002. Thus, another goal of the current evaluationwas to lay the foundation for the summative evaluation, the purpose of which will be to assistpolicymakers in determining whether the program should be continued in its current form, continuedwith revisions, or terminated. This concluding chapter summarizes the report’s main findings andidentifies a number of issues and implications for further consideration.

13.1 Summary of Main Findings

At the heart of Act 22 lies a bargain: charter schools will receive more autonomy in operations inexchange for being held more accountable for results. In many respects, this evaluation is an earlyand formative attempt to assess the implementation of this autonomy-accountability bargain.

Patterns of Growth in Pennsylvania Charter Schools

One of the key evaluation questions is whether the charter school law is providing students, parents,and teachers with new alternatives within the public school system. Indeed, choice policies will notbe effective without a sufficient “supply” of schools. As of the 2000/01 academic year, 66 charterschools are in operation in Pennsylvania. A 67th school was closed after its first year of operation.Taken as a whole, these schools have enrolled more than 20,000 students, or just over 1 percent ofall public school students in Pennsylvania. Throughout the life of the charter school law, there hasbeen relatively steady growth both in the number of charter schools and charter school students.Most of these schools enroll far fewer students than the typical Pennsylvania public school but morethan the typical charter school nationwide. There is, moreover, evidence of a trend toward largerschools.

Given the size of Pennsylvania and its student population, it would be unreasonable to expect thatcharter schools would provide meaningful choices to all students after just a few years of operation.It is clear that charter schools are concentrated in certain parts of the Commonwealth, particularly

185

Philadelphia. Indeed, while Philadelphia enrolls approximately 11 percent of the public schoolstudents in the state, it has 51 percent of the charter schools and 69 percent of the charter schoolstudents.

Charter schools appear to target students of a reasonably wide variety of grade levels. There is,however, a tendency for charter schools to seek students in the lower grade levels of the elementary,middle, and high school grades. We speculate that many of these schools are planning to “grow fromthe bottom,” which might well even out the distribution of grade levels covered. Similarly, charterschools’ mission statements indicate that they intend to serve a wide variety of educational interestsand goals. A large number of the Pennsylvania charter schools seek to serve at-risk students.

Charter School Start-Up

The range of charter school alternatives depends, in the first instance, on founders’ commitment toand skill in gaining charters and opening schools. Under Act 22 virtually any individual or groupmay apply for a charter, with the exception of for-profit and sectarian groups. Act 22 is somewhatmore restrictive, however, in defining who may grant charters. The Act gives primary authority toconsider and approve (or deny) charters to local districts (LEAs). Since July 1999, however, theCharter Appeals Board (CAB) has heard appeals from denied applicants, those who believe theircharter has been unjustly denied, and those who believe that an LEA has wrongly chosen not torenew a charter.

Not surprisingly, charter schools appear to be born of dissatisfaction with noncharter public schools,as evidenced by low PSSA scores. Lower performing districts, in turn, tend to have higherconcentrations of poor and nonwhite students. The founding coalitions behind charter schools ofteninclude public school teachers and administrators, academics, and members of the businesscommunity. However, there is little evidence thus far that parents have played a significant role infounding charter schools (though they do become active in the operation of charter schools).Founding coalitions’ goals have included providing a choice for poor children, creating a venue inwhich to operationalize ideas and practices hindered by district practices, promoting change innoncharter schools, and inculcating a particular ethnic or cultural perspective. Private conversioncharter schools, moreover, are often founded with an eye toward scaling up practices alreadyemployed in private schools and attracting a broader base of students.

Finally, there is evidence that in order to be successful, founding coalitions usually need to be ableto muster considerable political resources. In some cases this has involved tapping into support forcharter schools among influential community leaders and the more general dissatisfaction withpublic school bureaucracies. In some instances, however, it appears that host districts have seenadvantages in sponsoring charter schools, making such political tactics unnecessary. The moststriking finding, however, is the extent to which successful founding coalitions appear to have reliedon ties with preexisting schools, community development groups, ethnic/cultural groups, and othernongovernmental organizations. Some 78 percent of the charter schools in operation as of the1998/99 academic year had such an organizational base.

186

Charter School Finance

Like other school choice policies, Act 22 mandates that funding follows students. Thus, charterschools have an incentive to satisfy students and their parents who can “vote with their feet” for andagainst the school. Under the terms of the Act, this funding is funneled through host districts andother LEAs sending students to a given charter school. The size of the LEA subsidy is based on theLEA’s per-pupil expenditure on its own students. The mean per pupil subsidy that LEAs pay tocharter schools is $5,849. Subsidies for special education is based on a formula that reflects theLEAs spending patterns for special education and is typically at least twice the amount for a nonspecial education student.

Analysis of charter schools’ financial reports indicates that the median charter school receivedapproximately 81 percent of its total revenues from LEAs during the 1998/99 school year. There is,however, a large amount of variation among charter schools, with some charter schools receivingas little as 17 percent and others as much as 89 percent of their revenues from other LEAs. Muchof this variation, of course, is explained by the fact that some schools spend much more per pupilthan others and that the size of the LEA varies with district expenditures. Beyond that, schools varyin their reliance on non-LEA revenue sources. Next to LEA transfers, the largest revenue source forcharter schools is the federal government, mostly through Title I monies and charter school start-upgrants. While there is considerable school-by-school variation, the median charter school receivedapproximately 7 percent of its total revenues from the federal government. The remainder of charterschool revenues came from state grants and “local” sources, including earnings on investments,charitable donations, and revenues from student activities (e.g., candy sales, car washes, and so on).In addition, a few schools relied on proceeds from extended term financing during the 1998/99school year.

The report also examines charter schools’ expenditure patterns. Analysis of charter schools’financial data indicates that the median charter school spent approximately the same amount perpupil as its host district during 1998/99. Of that total amount, charter schools typically spent asmaller percentage (59 percent) on instructional items than their host districts (66 percent) and alarger percentage on support services (which include renting and maintaining facilities) and on othernoninstructional items. The reasons for these differences in expenditure patterns apparently lie morein the exigencies of starting new schools (e.g., acquiring and maintaining physical facilities) than inany inherent inefficiency in charter schools. Moreover, the fact that teacher salaries (a large portionof instructional expenditures) at charter schools are substantially lower than at noncharter schools(even after controlling for differences in education and experience) might account for some of thedifference. However, definitive answers to these questions lie beyond the scope of this study.

Last, the report examines charter schools’ fiscal viability using two indicators. First, analysis offinancial reports indicates that charter schools appear to be relatively conservative in budgeting,taking in more than expected on the revenue side and spending less than expected on the expenditureside of the ledger. Moreover, there is some evidence that schools benefit from experience, as secondyear schools had slightly lower expenditure variances than first year schools (there was no

187

discernible relationship on the revenue side). Second, analysis of end-of-year balances indicates thatmost schools ran surpluses for the 1998/99 school year. Seven charter schools (23 percent),however, showed negative balances (deficits), the largest of which constituted 10.7 percent of totalexpenditures.

Student and Parent Characteristics

Another important set of inputs to charter schools are the students who attend charter schools, andthe parent who send their children to these schools and volunteer at the schools. On the whole,charter schools enroll roughly equal proportions of males and females. The proportion of lowincome students enrolled in charter schools, moreover, is roughly comparable to host districts; inboth groups approximately two-thirds of students are eligible for free or reduced-price lunch.Charter schools, however, enroll a significantly higher proportion of nonwhite students (80 percent)than their host districts (57 percent) and charter schools nationwide (52 percent). Finally, moststudents (80 percent) previously attended a public school, 17 percent a private/parochial school, and3 percent who attended another charter school, or were homeschooled. The proportion of studentscoming from private schools roughly corresponds to the overall proportion of private school studentsin the Commonwealth (15 percent). However, there is significant variation among charter schools.

The charter school concept assumes that parents and students will choose charter schools on the basisof educationally relevant criteria and that they will be aware of and agree with the school’s mission.Evidence from surveys suggests that parents choose charter schools mainly because of the perceivedquality of instruction, the school’s academic reputation, dissatisfaction with their child’s previousschool, and because they agree with the school’s educational philosophy. Indeed, some 88 percentof parents sampled indicated that they are aware of their school’s mission. The most common reasoncited by students was that their parents believe the school is good for them. Beyond that, studentscite school safety, teacher quality, small classes, and dissatisfaction with their previous schools asthe most important factors driving their choice. Approximately 60 percent of students report thatthey are aware of their school’s mission. Some proponents of charter schools argue that allowing parents and students to select schools whoseeducational philosophies match their own will lead them to invest more time and energy in theschool. Only one-quarter of surveyed parents, however, indicated that they volunteer more than 3hours per month at their charter school. This is in spite of the fact that approximately half report thattheir school requires such involvement.

Teacher Characteristics

Teachers and staff represent another important human capital input for charter schools. The majorityof teachers and staff are female (71 percent), while the gender split for administrators and directorsis close to equal. This generally matches the gender distribution in noncharter public schools.Charter school teachers, however, are generally younger than their counterparts in noncharter publicschools, with approximately 50 percent under the age of 30 compared with 11 percent in publicschools across the nation. Approximately 40 percent of teachers are nonwhite, with African

188

Americans comprising the largest group, followed by Hispanics, Asian/Pacific Islanders, and NativeAmericans. Nearly half of sampled charter school staff indicated that they were classroom teachers(52 percent), 11.6 percent teaching assistants, and 7.3 percent specialists. Approximately 8.2 percentindicated that they were Chief Administrative Officers, principals, or school directors, and 21 percentindicated that they had some other title or position at the school.

Some 75 percent of charter school teachers surveyed in 1999/2000 reported that they are currentlycertified to teach in Pennsylvania. This represents a decrease from 82 percent in the previous year.On the other hand, the percentage of teachers certified to teach in other states rose from 2 percentto 4 percent over the same period, as did the percentage of teachers working to obtain certification(12 to 17 percent). The vast majority (75 percent) of charter school teachers with university degreeshad attained a BA as their highest level of education, while 22 percent had attained a MA. Amongnoncharter school public teachers in Pennsylvania, 55.6 percent of the teachers have BAs and 43.2percent have MAs. The average experience level of the teachers surveyed was just under five years.In short, charter school teachers are younger, less experienced, and possess less formal educationthan their counterparts in other public schools.

As with students and parents, the charter school concept assumes that teachers choose to work atschools with educational philosophies similar to their own. Teachers who agree on mission andphilosophy are, in this view, more likely to work effectively as a team. Indeed, the most commonlycited reason for joining the faculty of a charter school was an interest in being involved in a schoolreform effort, followed by the opportunity to work with like-minded educators. Other popularresponses included small class sizes, safety, and the school’s academic reputation. Some 97 percentof teachers, moreover, indicated that they are familiar with their school’s mission. Of these teachers,72 percent believe that their school’s mission is being followed “well” or “very well.”

Teacher attrition rates appear to be quite high in Pennsylvania charter schools. Nearly 40 percentof the charter school teachers left during or between the 1998-99 and 1999-00 school years. Duringthis same time period slightly less than 10 percent of the CAOs left or were replaced. Interviewsand analysis of salary data suggest that these high attrition rates among teachers might be linked toteacher dissatisfaction with salaries, which are substantially lower than in comparable noncharterpublic schools.

Professional Opportunities for Teachers

One of the goals of Act 22 is to provide enhanced autonomy and professional developmentopportunities for teachers. Many teachers have come to charter schools seeking autonomy increating and implementing curriculum. We found evidence at a number of charter schools that therewas a conscious effort to involve teachers in developing curricula. Teachers indicated that they hadautonomy in curriculum decisions and freedom to utilize creative approaches. Indeed, many teachersreport that they have considerable flexibility and opportunities for creativity in their day-to-dayactivities. We found clear evidence in slightly more than half of the charter schools that the teachers

189

in the schools work collaboratively. They are encouraged in their collaborative work efforts throughprograms of team teaching, mentoring, and staff members creating presentations.

Evidence from teacher surveys indicates that approximately one-quarter to one-third of teachers aresatisfied with their school’s physical facilities, while one-third to one-half were satisfied withresources available for instruction and other educational functions. The average charter school paidteachers an annual salary of approximately $30,000, compared with the state average ofapproximately $48,000 and an average of $40,000 for schools with similar levels of teachereducation and experience and similar per pupil expenditures. Only 30 percent of teachers report thatthey are satisfied or very satisfied with their salaries.

There was a measurable difference between initial teacher expectations and current experience onmany topics. The largest differences between initial expectations and current experience came onsuch topics as the effectiveness of leadership and administration, parents’ ability to influence thedirection and activity of the school, availability of support services to students, the extent to whichstudents receive sufficient individual attention, and good communication between parents/guardiansand the school. There was also a large gap between expectations and experience in terms of teacherempowerment and the degree to which they are able to influence the steering and direction of theschool. Other areas that teachers felt less positive about were class size, and emphasis on academics.One must note that many teachers were hired before the school opened, some teachers simplyexpected too much, and many were young teachers. Despite these figures, many teachers reportedthat they are still satisfied with their teaching environment and about 75 percent planned to returnedto the school the following year.

Charter schools devote considerably more time to teacher professional development activities thannoncharter schools, with the average charter school allotting 7 professional days and noncharterschools 5 days. A strong emphasis on graduate study was frequently reported, with some programsrequiring it. The format for professional development opportunities included workshops,conferences, in-service training, and graduate courses. The content of professional developmentopportunities included methods of teaching, technology, student assessment, and classroommanagement.

Innovations in Governance, Curriculum, and Instruction

One of the stated goals of Act 22 and other charter school statutes is to encourage the developmentand diffusion of innovative practices. In this sense, charter schools are to be public education’s“R&D.” This report examines innovations in leadership, organization, discipline, curriculum, andinstruction.

Teachers and parents generally expressed approval of their school’s leadership, with 67 percent ofparents and 53 percent of teachers indicating that the statement “This school has good administrativeleadership” was true. Moreover, approximately 70 percent of teachers agreed that their school’s

190

leaders set high standards, communicate them effectively to others, and set a good example byworking hard themselves.

Parents, teachers, and students, moreover, often have enhanced opportunities to participate in schooldecision making, including those listed below:

� Involvement in the strategic planning process, including attendance at planning meetings

� Attendance at meetings to plan and evaluate curricula

� Attendance at school board meetings and otherwise providing input to board members

� Input via school surveys and interviews

� Formal appeals process for teacher grievances

While opportunities for influence do not guarantee actual influence, approximately half of teachers(54 percent) said that they are involved in decision making at their school. A similar proportion ofstudents report that administrators listen to their ideas about the school. Approximately 83 percentof parents, moreover, said that it was true or partly true that “I am able to influence the direction andactivities in the school.” Many teachers, however, indicated that their workloads did not leave themenough time to remain involved in school decision making. Parents cited work and familyobligations as barriers to their participation.

Notable organizational and disciplinary practices in the charter schools include extended hours andSaturday sessions, inclusion of both middle and high school grades in order to address commonproblems in making the transition from middle to high school, and preventive disciplinary measuresthat focus on behavior modification techniques.

Curricular innovations in charter schools often come in the form of a focus on a particular ethnic orcultural perspective. Some schools, for instance, offer courses on Spanish and Swahili; in somecases, such instruction begins in kindergarten. Other schools focus on social and vocational skills,with some offering individualized instructional plans for all students. In some cases, charter schoolsemploy packaged programs such as Success for All reading, Everyday Mathematics, or DiscoveryWorks.

Evidence of charter schools’ use of technology was scant. However, the weighted average numberof students to computers in charter schools is 4.5, compared with 7.5 for their host districts. Oneschool maintains student portfolios on-line, and–at the time of the study–at least one school offerson-line courses that students can participate in from home.

Special Education

There is an important conflict between the spirit of charter schools laws, which seek to deregulatecharter schools, and the preemption of special education by federal law. Indeed, it is not clear thatstates have the authority to significantly ease the burden of special education regulations. Act 22

191

may have created some confusion over the matter given its lack of attention to the special educationissue. To its credit, the Commonwealth has made considerable headway since enactment of Act 22in clarifying charter school obligations and providing guidance to charter schools, in proposing newregulations in this area, and also in providing support and technical assistance to help charter schoolsprovide special education services that are in compliance with IDEA.

The average proportion of students with disabilities in the charter schools was 10.5 percent in 1999-00, which is slightly lower than the statewide average of 12.5 percent. Among the schools thatopened during the first two years of the reform, the proportion of students with disabilities washigher (12.5 percent) than for the 17 schools that opened for the 1999/2000 school year (7.9 percent).However, there was great variation among the schools, with some schools reporting as many as 42percent special education students and some reporting no such students.

While the findings reported in this report that not all is well with special education in thePennsylvania charter schools, it is important to recognize that regular public schools are also not ableto fully satisfy parents who have students with special needs. Indeed, a large proportion of parentssurveyed indicated that an important reason for choosing a charter school was the inability of theirprevious school to accommodate the child’s special needs.

Student Achievement in Charter Schools

The data on student achievement in charter schools precludes conclusive statements about charterschools’ impacts on student learning. First, the Commonwealth’s charter school policy is still quiteyoung, leaving charter schools with little time to demonstrate their ability to improve studentachievement. Second, data on only a small number of charter schools were available to theevaluation team. Thus, generalizations to the broader movement are tenuous. Finally, no data existon such important characteristics as students’ precharter school achievement rates, which seriouslydiminishes the evaluation’s ability to provide valid assessments of charter school impact.

Nonetheless, the report presents a number of suggestive, if mixed, findings. Charter schools as agroup produced Pennsylvania System of School Assessment (PSSA) scores that were considerablylower than all noncharter public schools in the Commonwealth. Such comparisons, however,probably measure differences in the types of students who choose to attend charter schools more thanany impact the charters have on their students. Thus, the report presents a number of more precisecomparisons.

First, the report calculates two-year change scores for the four schools for which there are two yearsof PSSA data. Taken as a group and averaged across grades and subject areas, these schools postedgains of more than 100 points (the scale ranges from 1000 to 1600). In doing so, charter schools asa group outgained their host districts as a group by some 86 points over the same period. However,the change score analysis is limited by the fact that it cannot follow the same cohort of students overtime. A number of schools provided single cohort, pre/post comparisons of the same cohort ofstudents based on commercial standardized tests. These data also show that a substantial number

192

of charter schools posted gains in student achievement. However, this analysis is limited by the factthat it too is based on a small sample of charter schools.

Second, the report compares each charter school with its host district(s) and with a set ofdemographically similar noncharter schools. In both cases, charter schools as a group wereoutperformed by approximately 50 points on the PSSA. While based on a larger sample of charterschools, comparisons to host districts are imperfect since many charter schools target special needspopulations. Thus, comparisons with host districts might not fully account for self-selection effects.Comparisons with demographically similar schools were also based on a larger sample of charterschools, but are limited by the fact that the only reliable demographic variable on which charter andnoncharter schools could be compared is the percentage of students eligible for free or reduced-pricelunch. Thus, these comparisons were unable to match schools on special education, race, and otherfactors associated with student achievement.

Other Indicators of School Quality

There is legitimate debate about precisely what types of student outcomes charter schools should beheld accountable for. While most stakeholders seem to agree that student achievement is animportant (if not the only) goal of charter schools, others argue that schools should also be judgedon their ability to satisfy their customers.

Survey evidence suggests that as a group charter school students report that their level of academicperformance has improved since they moved to a charter school. Teachers are perhaps a little lesssanguine; most teachers expected student achievement to improve more than it has. Still, more thanhalf the teachers believe that student achievement is on the rise at their school. Moreover, parentsand teachers reported that charter schools are serving needs not well served by other schools.

The report also examines various indicators of market accountability, or the extent to which parentsand students “vote with their feet” for or against charter schools. A nonrandom sample of charterschools indicates that the average school has a waiting list of 125 students, or 45 percent of totalcurrent enrollment. However, just as there are students waiting to get into charter schools, a numberhave left charter schools to return to other public and nonpublic schools. A nonrandom sample ofstudent rosters indicates that the average charter school lost 38 percent of its students from Spring1999 to Spring 2000. Moreover, data from the Philadelphia school district indicates that some 1,800students left Philadelphia charter schools to return to district schools. We must emphasize, however,that available evidence on charter school entry and exit is not based on representative samples.Moreover, we have no way of knowing whether decisions to leave charter schools are motivated byconcerns about school quality or other reasons. In spite of the significant enrollment instability insome charter schools, more than 90 percent of parents surveyed report that the quality of instructionin their charter school is high and that their child receives adequate attention. Approximately halfof the students surveyed said that they would recommend their charter school to a friend.

193

Finally, the report examines a number of indicators designed to capture charter schools’ climates andcultures. Many of these factors might indicate the potential for high achievement levels in the future.For instance, charter schools had an estimated attendance rate of 90 percent for both 1998/99 and1999/2000. Moreover, nearly three-fourths of parents and teachers reported that their charter schoolshad high expectations for students. A similar percentage of students said that their teachersencourage them to think about their future. However, a much smaller percentage (one-third) saidthat students at their charter school were more interested in learning than students at their previousschool. Fewer than one-half of respondents report being fully satisfied with school facilities.

13.2 Research and Evaluation Issues for Further Consideration

Given that this evaluation report seeks, among other things, to provide the foundation for a 5-yearsummative evaluation, we conclude by identifying a number of research and evaluation issues forfurther consideration. While the current evaluation lacked sufficient data to address them, these areissues that warrant closer examination in any summative evaluation of Pennsylvania’s charter schoollaw. The list is not meant to be exhaustive. Rather, it is intended to initiate discussion on directionsfor further research and evaluation on the subject.

� Charter approval and the accountability process. Charter school application and approval isthe first step in the accountability process. In order to better understand this process, researchersand evaluators should systematically gather data on both successful and unsuccessfulapplications. Such an exercise would provide a clearer picture of the factors that determinesuccess. As discussed in chapter 4, the current evaluation was restricted to observing thecharacteristics of successful applications only. While this identifies some of the commonfeatures of successful applications, it does nothing to assess whether those features arenecessary for success. Such information would clearly be useful to would-be applications.More than that, however, it would also help policymakers better assess how effective theapplication and approval process is as a quality control mechanism that filters out weakapplications and lays the foundation for educationally effective charter schools.

� LEA approval/oversight and charter school quality. Another issue is the role of the variousLEA approval and oversight processes in charter school quality and, ultimately, studentoutcomes. While this issue was beyond the scope of the current evaluation, the evaluation teamnoticed significant variation in the extent to which LEAs provide clear standards for andmeaningful oversight of charter schools, both during the approval process and beyond.Researchers and evaluators might attempt to assess whether certain types of local approval andoversight processes are more likely to produce strong schools and improvements in studentoutcomes. Given that Act 22 essentially delegates much of the responsibility for charter schoolaccountability to LEAs, it is important to study and learn from local accountability practices.

� Charters granted on appeal. The first charter schools approved by the Charter Appeals Boardare just now coming to existence. An important question for future researchers and evaluatorsinvolves the climate in which these schools operate. Experience with similar processes in other

194

states suggests that the schools might face hostility from local districts. Future research andevaluation activities should attempt to assess the extent of any such hostilities and, moreimportantly, their impact on charter school quality and student outcomes. Future research mightalso examine strategies for making such relationships cooperative and productive over time.

� The role of nongovernmental organizations. Chapter 4 reported that most (78 percent) of thefirst 31 charter schools were established upon the foundations of a preexisting nongovernmentalorganization. Future research and evaluation activities should assess whether these trends holdfor more recently established charter schools. Indeed, it is conceivable that such reliance issymptomatic of the early stages of the reform and that it will diminish as the charter schoolcommunity develops more collective experience in charter start-up. Moreover, futureevaluation and research activities should continue to track the role of private educationmanagement organizations (EMOs) in charter development and start-up. As noted in chapter3, EMOs have played a remarkably small role in Pennsylvania charter schools thus far.However, there is evidence that the role of EMOs is on the increase.

� Explaining variations in expenditure patterns. Chapter 5 reports the findings that as a groupcharter schools spend a lower percentage of their total expenditures on instructional items thantheir host districts and noncharter public schools in general. Some of the difference might beexplained by the need to cover rent and other facilities costs and the fact that charter schoolteachers receive lower salaries than teachers in comparable noncharter public schools. Futureresearch and evaluation activities should seek to test these and other explanations, since theissue will shed light on how efficiently charter schools are spending their money.

� Value for money. More generally, future research and evaluations should seek to assess whethercharter schools provide more educational value for the money than noncharter public schools.Answers to questions about efficiency, however, require evaluators to examine the relationshipbetween fiscal inputs and student (and perhaps other) outcomes. Unfortunately, the dataavailable for this report were insufficient to address such questions. With the passage of timeand the opening of new schools, future evaluators might be able to provide at least preliminaryanswers to this important question.

� Explaining variations in student achievement. Chapter 11 reports the considerable variation instudent achievement among charter schools. Future research and evaluation activities shouldconduct in-depth case studies that attempt to identify whether there are common characteristicsthat distinguish charter schools producing high (or higher than expected) achievement scores.Nor would this be simply an academic exercise. Findings from such case studies might identifyconstraining and enabling factors that could inform program revisions that would make allcharter schools more effective. Moreover, such findings might identify a set of best practicesand provide a medium for the diffusion of innovations among charter schools and betweencharter and noncharter schools.

195

� Market accountability. Chapter 12 reported that the average charter school lost 38 percent ofits students from Spring 1999 to Spring 2000. Such turnover might imply dissatisfaction withthe charter schools. However, there is no way to know for sure without more information onwhy students and parents choose to leave the schools. Future research and evaluation activitiesshould include an examination of the motivations for leaving charter schools. This will providemore precise information on the schools’ market accountability.

� Data collection on student achievement. Chapter 11 emphasizes that the findings on studentachievement are suggestive but in no way conclusive. This is because of important limitationsin the PSSA data available to researchers and evaluators. Among the most serious of theselimitations is the inability to follow a single cohort of students over time. Revisions to thePSSA would be costly and time-consuming. Fortunately, many schools provide pre/post, singlecohort comparisons based on commercial standardized tests. However, the evaluation teamfound it difficult to get such results for a large number of schools. Moreover, available datacame in a variety of formats that made analysis difficult. Thus, a more feasible approach toimproving student achievement data would be to develop guidelines for reporting andformatting commercial achievement test data. Even if PDE does not wish to mandate aparticular test, schools usually have a wide of range of choices about how scores are reportedback to them. Charter school officials might be willing and able to choose formats thataccommodate the needs of state-level accountability actors and evaluators. As always, any suchpolicy must balance the demands of accountability against the value of school autonomy. Suchimprovements in data collection would vastly improve the quality of inferences about charterschool achievement and thus provide a much sounder foundation for the 5-year summativeevaluation yet to come.

13.3 Policy Issues for Further Consideration

In addition to the research and evaluation questions listed above, the report raises a number of policyissues for further consideration. Some of these issues involve minor changes in oversight. Othersmight require more fundamental debate over the goals and policy tools of Act 22. Once again, thislist is not exhaustive.

� Reliance on nongovernmental organizations. The finding that a large proportion of the first 31charter schools rely on nongovernmental organizations for support and finance raises a numberof questions. If the primary goal of the Act is to serve as public education’s “R&D” by creatinginnovative schools that can help leverage change in noncharter public schools, then suchreliance might be good. Indeed, the need to garner the support of such organizations mightprovide a de facto quality control check on charter school development, weeding out weakapplications and ensuring that approved schools have adequate organizational and financialsupport. If, however, the primary goal of Act 22 is to provide public school choice to a largeproportion of the student population, such dependence on external groups casts some doubt onthe scalability of the charter school reform in Pennsylvania and its ability to provide choice to

1 It is notable that Act 22, unlike many other charter school statutes, provides no caps on thetotal number of charter schools in the Commonwealth.

196

a large proportion of students.1 Indeed, it is likely that there are only so many organizationswilling and able to undertake the burden of supporting charter schools. Thus, such dependencemight indicate that charter schools will require more support from the Commonwealth and otherofficial actors. Resolution of these issues is not merely an empirical and technical issue.Rather, it requires a clear statement of Act 22's many goals and, most importantly, aspecification of the relative weights given to each of these goals.

� The role of competition between charter and noncharter schools. One of the ostensible goalsof Act 22 is to foster head-to-head competition between charter and noncharter schools.However, the apparent tendency for LEAs to approve charters that target special needspopulations might subvert this goal. From one perspective, this pattern of charter schoolsspecialization in at-risk student is desirable inasmuch as this segment of the student populationhas perhaps not been well served by noncharter public schools. Head-to-head competition,however, might require that the schools compete for the same groups of students. Furtherresearch is required to determine whether this pattern holds for more recently approved schools.Even if the patterns appears to be robust, however, its policy implications depend upon howpolicymakers and stakeholders weigh the importance of competition against other goals in Act22. In short, it is as much a question of values as an empirical question.

� Innovation in charter schools. Whether a given practice is innovative depends, in part, on thepurposes at hand. Charter school proponents variously trumpet a number of purposes orultimate goals for charter schools, each of which has different implications for innovation. Thisstatement of charter schools’ goals is consistent with a fairly ambitious view of innovation–thecreation of truly unique practices that can be shared and perhaps emulated by a large group ofschools. A more modest view of innovation is that charter schools exist primarily to providechoice. If this is the case, then innovations must simply provide new options for students andparents in a particular geographic area. Hence, the bar is high on the former view and somewhatlower on the latter view. The Charter Appeals Board (CAB) articulated a view of innovationthat corresponds more closely with the second, narrower view of innovation. Policymakersmight wish to reconsider this definition to ensure that it is in accordance with Act 22's balanceof choice versus other goals.

� Market vs. other forms of accountability. This report provides evidence of charter schoolaccountability based on student achievement data and on various measures of “marketaccountability,” including “customer” satisfaction and the extent to which students and parents“vote with their feet” for or against charter schools. The report finds that, on balance, there ismore evidence of accountability defined as customer satisfaction than of accountability definedin terms of student achievement. However, it is unclear from Act 22 how evaluators andstakeholders should weigh these alternative measures of success. Thus, policymakers might

197

wish to consider the relative importance of these criteria in advance of the 5-year summativeevaluation.

In closing, we return to the evaluation’s core question: “Does increased flexibility in exchange forincreased accountability result in improved pupil results?” As Chapter 11 points out, evidence aboutstudent achievement in charter schools is far from conclusive, making a definitive answer to thisquestion impossible at this juncture. Student achievement notwithstanding, this report indicates thatPennsylvania charter schools are well on their way to fulfilling their promise in some areas but areexperiencing difficulties in others. Fortunately, the passage of more time and the introduction ofnew charter schools will enable future evaluation reports to provide more definitive answers to theseimportant questions. In the meantime, this report is intended to help policymakers and stakeholdersimprove the charter law and the schools founded under its auspices.

198

References

Act 22. (1997). Pennsylvania Charter School Law. Prepared and approved by the GeneralAssembly of the Commonwealth of Pennsylvania. Available from the web site of thePennsylvania Department of Education. <http://www.pde.psu.edu/charter.html>

Ahearn, E. M. (1999). Charter schools and special education: A report on state policies.Alexandria, VA: National Association of State Directors of Special Education.

Bourantas, D., & Papalexandrif, N. (1992). Variables affecting organizational commitment: Privatevs. public owned organization Greece. Journal of Managerial Psychology, 7(1), 3-10.

Choy, S. P., Chen, X., & Ross, M. (1998). Toward better teaching: Professional development in1993-1994, NCES 98-230. Washington, DC: Department of Education, National Center forEducation Statistics.

Chubb, J.E. & Moe, T. (1990). Politics, markets, and America’s schools. Washington, D.C.:Brookings Institute

Cobb, C.D. & Glass, G.V. (1999). Ethnic segregation in Arizona charter schools. Education PolicyAnalysis Archives, [Online]. Available: http://epaa.asu.edu/epaa/v7n1/.

Coleman, J.S., et al. (1966). Equality of educational opportunity. Washington, D.C.: U.S.Government Printing Office.

Dale, R. E. (1999). The politics of special education policy in charter school legislation: Lessonsfrom Pennsylvania. Pennsylvania State University

Driscoll, M. E. (1993). Choice, achievement, and school community. In E. Rasell & R. Rothstein(Eds.), School choice examining the evidence (pp. 147-172). Washington, DC: EconomicPolicy Institute.

Duquense Charter Schools Project. (2000). Pennsylvania’s charter appeal board: A first look at itsimpact on charter approval. Pittsburgh, PA: Duquesne University.

Elmore, R. E. (1988). Choice in public education. In W. L. Boyd & C. T. Kerchner (Eds.), Thepolitics of excellence and choice in education (pp. 79-98). New York: Falmer.

Grissmer, D. et al. (2000). Improving student achievement: What NAEP state test scores tell us,[Online] Available: http://www.rand.org/publications/MR/MR924/.

Hanushek, E. (1997). Assesing the effects of school resources on student performance: An update.Educational Evaluation and Policy Analysis, 19(2), 141-164.

Hill, P. et al. (1997). Reinventing Public education: how contracting can transform America’sschools. Chicago: University of Chicago Press.

199

Hirschman, A. O. (1970). Exit, voice, and loyalty: Responses to decline in firms, organizations, andstates. Cambridge, MA: Harvard University Press.

Horn, J., & Miron, G. (1999). Evaluation of the Michigan public school academy initiative finalreport. Kalamazoo, MI: Evaluation Center, Western Michigan University.

Horn, J., & Miron, G. (2000). An evaluation of the Michigan charter school initiative:Performance, accountability, and impact. Kalamazoo, MI: The Evaluation Center, WesternMichigan University.

Hubley, N. (2000). Comments of proposed regulations for charter schools services and programsfor children with disabilities. Available: http://www.elc-pa.org/Charters3.htm.

ISSP. (2000). Inclusive practices. Harrisburg: Instructional Support System of PennsylvaniaAvailable: (http://www.cisc.k12.pa.us/svs/inclusive/default.htm).

Lange, C. M. (1997). Charter schools and special education: A handbook. Alexandria, VA: National Association of State Directors of Special Education.

Lowery et al. (1992). The politics of dissatisfaction: citizens, services, and urban institutions. Armonk, NY: M.E. Sharpe.

Meister, G., & Schuh, A. (2000). A report on Philadelphia charter schools: Facing challenges,forging solutions. Philadelphia, PA: Drexel University/FOUNDATIONS.

Meyer, J. (1970). The charter: Conditions of diffuse socialization in schools. In W. Richard Scott(Ed.), Social processes and social structure (pp. 564-578). New York: Holt, Rinehart &Winston.

Meyer, J. (1972). The effects of the institutionalization of colleges in society. In K. Feldman (Ed.),College and student selected readings in the social psychology of higher education (pp. 109-126). New York: Pergamon Press.

Mintrom, M. (2000). Policy entrepreneurs and school choice. Washington, D.C.: GeorgetownUniversity Press.

Miron, G. (2000). The initial study of Pennsylvania charter schools: First annual report.Kalamazoo, MI: Evaluation Center, Western Michigan University.

Millman, J. ed. (1997). Grading teachers, grading schools: is student achievement a validevaluation measure? Thousand Oaks, CA: Corwin Press.

National Center for Education Statistics (2000). State profiles of public elementary and secondarye d u c a t i o n , 1 9 9 6 - 1 9 9 7 : U . S . p r o f i l e , [ O n l i n e ] . A v a i l a b l e :http://nces.edu.gov/pubs2000/stateprofiles/us_profile.html.

Perrow, C. (1986). Complex organizations: a critical essay. New York: McGraw-Hill.

200

PDE. (1997). Preliminary guidance for charter schools regarding special education PennsylvaniaD e p a r t m e n t o f E d u c a t i o n , B u r e a u o f S p e c i a l E d u c a t i o n .(http://www.pde.psu.edu/charse.html)

PDE. (2000). Proposed rulemaking Pennsylvania Bulletin, 30(28). 3463-3467. Harrisburg:Pennsylvania Department of Education (8 July, 2000).

PDE, Bureau of Information Systems (2000). Classroom teachers by sex, type, race, level ofeducation, and local education agency 1999-2000. Harrisburg, PA: PennsylvaniaDepartment of Education Available: www.pde.psu.edu/esstats.html..

PDE, Bureau of Special Education (2000). Proposed 22 Pa. Code, ch. 14, revised 8/16/00.Harrisburg: Pennsylvania Department of Education. Available:http:www.pde.psa.edu/regs/chap14prop.pdf.

Philadelphia Public Schools (2000). Briefing paper on charter schools. Philadelphia, PA: Author.

Rofes, E. (1998). How are school districts responding to charter laws and charter schools?Berkeley, CA: Policy Analysis for California Education, University of California.

RPP International. (1998). A national study of charter schools, Second-year report. Washington,DC: U.S. Department of Education.

RPP International. (2000). The state of charter schools, national study of charter schools, fourth-year report. Washington, DC: U.S. Department of Education.

Schneider, M., & Teske, P. (1992, Sept.). Toward a theory of the political entrepreneur: Evidencefrom local government. The American Political Science Review, 86(3), 737-747.

Shaffer, G. (May, 2000). Charter trouble. Philadelphia City Paper.

Teddlie, C. & Stringfield, S. (1993). Schools make a difference: lessons learned from a 10-yearstudy of school effects. New York: Teachers College Press.

Turnbull, H. R., & Turnbull, A. P. (1998) Free appropriate public education (5th ed.). Denver, CO:Love Publishing Company.

Wells, A.S., et al. (1999). Underlying policy assumptions of charter school reform: The multiplemeanings of a movement. Teachers College Record, 100(3), 513-535.

Wenglinsky, H. (1997). When money matters: how educational expenditures improve studentperformance and how they don’t. Princeton, NJ: Educational Testing Service.

Wilson, J. (1989). Bureaucracy: What government agencies do and why they do it. New York:Basic Books.

A-1

Appendix A

Complete List of Evaluation Questions1. Does increased flexibility in exchange for increased accountability result in

improved pupil results?

2. What effect does budget have on student results, nonacademic services, andschool facilities?

3. Are opportunities offered to charter school teachers, parents, and students toinfluence classroom and school policy significantly different from those offered attraditional public schools?

4. Are the opportunities (i.e., professional growth, salaries, benefits, employeerights) for teachers and other employees significantly different at a charter schoolthan at a traditional public school?

5. What is the impact of charter schools as related to district reform efforts?

6. Is there evidence that, over the term of the charter, student learning hassignificantly improved?

7. What are promising practices in charter schools that could be included in districtsystemic reform?

Specific areas to be addressed in the evaluation are listed as follows:

1. Staffing

a. Staff profile indicating degree of constancy and/or stability of teaching and classifiedstaff

b. Percentage of credentialed professional employees holding a valid Pennsylvaniacertificate

c. Qualifications of noncertified professional employees

d. Pupil-teacher ratio for special education needs

e. Staff development and implementation

f. Other relevant measures that address the issue of staffing of charter schools

A-2

2. Educational Achievement

a. Percentage of students who have achieved instructional benchmarks identified in charter proposal

b. Evidence that charter students are more/less successful at next level of education (i.e., elementary to middle school) than noncharter students

c. Evaluate transition of students from charter schools to traditional public and privateschools, and vice versa, in terms of academic achievement

d. Evidence that limited English proficient and special education students have access tocore curriculum

e. Comparison of PSSA scores to district schools and to "like" noncharter schools and percentage of student body tested

f. Determine quality of evaluation instruments wherein students demonstrate achievementand validate the program objective (other than the PSSA)

g. Evaluate appropriate implementation of individualized educational programs (IEP) foridentified special education students

h. Effectiveness of services offered to at-risk youngsters

i. Quality of technology program

j. Other relevant measures that address the educational achievements of students in charter schools

3. Student Attendance

a. Comparison of student attendance to district schools and to "like" noncharter schools

b. Other relevant measures that address attendance issues of students in charter schools

4. Student Discipline

a. Comparison of student discipline to district schools and to "like" noncharter schools

b. Other relevant measures that address discipline issues of students in charter schools

A-3

5. Governance Process

a. Effectiveness of process as detailed in charter proposal

b. Assess vertical and horizontal decision-making process detailed in school's charter

c. Terms of authority of charter school principal/director

d. Procedure used to elect/select teachers, parents, and community members to various governance committees stated in school's charter

e. Other relevant measures that address governance issues of children in charter schools

6. Community Involvement

a. Process of reporting student progress to parents

b. Effectiveness of parent outreach programs/advisory boards

c. Survey indicating level of parent satisfaction

d. Racial/ethnic balance among pupils compared with district student population

e. Other relevant measures that address the issue of community involvement and charter schools

7. Budget, Accounting, and Business Practices

a. Compliance with guidelines for specific federal, state, and local grants/revenue sources

b. Documentation on file at school to support attendance accounting

c. Flexibility of purchasing procedures available to charter schools, e.g., outside vendors, bidding, timely purchases

d. Determine whether revenues exceeded expenses or if expenses exceeded revenues andthe source of such revenues

e. Percentage of variation from budget to actual expenditures

f. Determine cash balance at the end of fiscal year. Did it increase or decrease each year?

WMU Charter School SurveyInformant Group: Teachers/Staff (N=447) 1999 Descriptive statistics

1. What is your role at this school?

TeacherTeachingassistant

SpecialistStudentteacher

Principal/director

OtherTotal Missing

N 216 55 23 1 20 131 446 1

% 48.4% 12.3% 5.2% 0.2% 4.5% 29.4% 100.0%

2. What is your current teaching certification status? (Teachers only n=214)

Total

N 214

% 100.0%

3. Are you teaching in a subject area in which you are certified to teach?(Classroom teachers only, n=214)

Yes NoNot

applicable Total

N 163 29 22 214

% 76.2% 13.6% 10.3% 100.0%

4. With which grade do you mostly work?Grade Level

K 1st 2nd 3td 4th 5th 6th 7th 8th 9th 10th 11th 12th Total Missing

N 37 40 29 21 20 12 24 12 8 36 22 9 19 151 440 7

% 8.4% 9.1% 6.6% 4.8% 4.5% 2.7% 5.5% 2.7% 1.8% 8.2% 5.0% 2.0% 4.3% 34.3% 100.0%

5. What is your age? (all staff)Youngerthan 20

20-29 30-39 40-4950 orolder

Total Missing

N 6 166 116 101 53 442 5

% 1.4% 37.6% 26.2% 22.9% 12.0% 100.0%

6. What is your race/ethnicity?

White Black HispanicAsian/Pac.

IslanderNative

AmericanTotal Missing

N 224 160 34 4 2 424 23

% 52.8% 37.7% 8.0% 0.9% 0.5% 100.0%

7. What is your gender?

Female Male Total Missing

N 298 131 429 18

% 69.5% 30.5% 100.0%

Totals for All Pennsylvania Charter School Teachers and Staff, 1998-1999

Notapplicable

Not certified and notworking to obtain

certification

81.8% 2.3% 12.1% 3.7%

175 5 26 8

Currently certified toteach in this state

Currently certified toteach in another state

Workingto obtain

certification

Role in school

0%

20%

40%

60%

80%

100%

Teacher Teachingassistant

Specialist Studentteacher

Principal/director

Other

Distribution by Race/Ethnicity

0%

20%

40%

60%

80%

100%

White Black Hispanic Asian/Pac.Islander

NativeAmerican

Gender

Female

Male0%

20%

40%

60%

80%

100%

Teaching a subject in which you are certified

Notapplicable

Yes

No

0%

20%

40%

60%

80%

100%

Current certification status

0%

20%

40%

60%

80%

100%

Currently certifiedto teach in this

state

Currently certifiedto teach in

another state

Working to obtaincertification

Not certified & notworking to obtain

one

Age

0%

20%

40%

60%

80%

100%

Younger than20

20-29 30-39 40-49 50 orolder

Appendix_B1_CSS_Tch-Staff_98-99 1 WMU Charter School Survey

8. How many years of experience have you had 9. Years atin each of these types of schools? current school?

Privateschool

Parochialschool

Charterschool

Publicschool

Other TotalTotal

(excluding"other")

Years atcurrentschool

Mean 1.07 0.41 1.09 2.18 0.65 5.40 4.75 1.23

STD 2.68 1.85 0.51 4.20 1.62 5.44 5.21 0.90

10. How much formal education have you had?Did not

completehigh school

Completedhigh

school

Less than4 years

of college

CollegegraduateBA/BS

Graduatecourses,

no degree

Graduate/professional

degreeTotal

Mis-sing

N 4 38 57 143 76 111 429 18

% 0.9% 8.9% 13.3% 33.3% 17.7% 25.9% 100.0%

11. What is the highest college degree you hold?

Bachelors Masters5-6- yearCertificate

Doctorate Total Missing

N 216 92 9 14 331 116

% 65.3% 27.8% 2.7% 4.2% 100.0%

12a. Are you working toward another degree at this time?No Yes Total Missing

N 289 133 422 25

% 68.5% 31.5% 100.0%

12b. If yes, what degree?

Bachelors Masters5-6- yearCertificate

Doctorate Total Missing

N 23 86 11 14 134 313

% 17.2% 64.2% 8.2% 10.4% 100.0%

13a. Are you aware of the school's mission?No Yes Total Missing

N 15 419 434 13

% 3.5% 96.5% 100.0%

13b. If yes, to what extent is the missionbeing followed by the school?

Not verywell

Fair WellVerywell

Total Missing

1 2 3 4

N 16 109 176 121 422 25

% 3.8% 25.8% 41.7% 28.7% 100.0%

14. Do you plan (hope) to be teaching at this school next year?No Yes Total Missing

N 99 302 401 46

% 24.7% 75.3% 100.0%

Average years of experience by school type

0

3

6

9

12

Privateschool

Parochialschool

Charterschool

Publicschool

Other

Yea

rs

Level of formal education

0%

20%

40%

60%

80%

100%

Did notcomplete

high school

Completedhigh

school

Less than4 years

of college

CollegegraduateBA/BS

Graduatecourses,

no degree

Graduate/professional

degree

Highest college degree

Bachelors

MastersDoctorate

5-6- yearCertificate0%

20%

40%

60%

80%

100%

Are you working toward another degree?

NoYes

0%

20%

40%

60%

80%

100%

Degree you are working toward

Bachelors

Masters

5-6- yearCertificate Doctorate0%

20%

40%

60%

80%

100%

Are you aware of the school's mission?

No

Yes

0%

20%

40%

60%

80%

100%

To what extent is the school mission being followed?

Not very wellFair

Well Verywell

0%

20%

40%

60%

80%

100%

Do you plan (hope) to be teaching at this school next year?

No

Yes

0%

20%

40%

60%

80%

100%

Appendix_B1_CSS_Tch-Staff_98-99 2 WMU Charter School Survey

15. Rate the importance of the following factors in your decision to seekemployment at this school.

Percentages

Mean STD N Missing

1 2 3 4 5

Convenient location 24.8% 14.4% 22.5% 15.8% 22.5% 2.97 1.48 436 11

More emphasis on academics as opposed toextracurricular activities

9.1% 7.9% 32.0% 27.6% 23.4% 3.48 1.19 428 19

My interest in being involved in an educational reformeffort

4.0% 4.2% 13.4% 29.3% 49.1% 4.15 1.06 426 21

Promises made by charter school's spokespersons 7.4% 7.2% 26.7% 24.4% 34.3% 3.71 1.22 431 16

Academic reputation (high standards) of this school 9.6% 4.9% 26.3% 25.6% 33.6% 3.69 1.25 426 21

Parents are committed 11.5% 9.4% 24.6% 21.1% 33.5% 3.56 1.34 427 20

Safety at school 6.3% 6.1% 20.7% 23.5% 43.4% 3.92 1.20 429 18

Difficulty to find other positions 38.6% 13.0% 21.6% 12.3% 14.5% 2.51 1.46 422 25

Opportunity to work with like-minded educators 4.0% 3.0% 18.2% 30.1% 44.8% 4.09 1.05 429 18

This school has small class sizes 5.4% 6.1% 16.9% 30.6% 40.9% 3.96 1.15 425 22

Notimportant

Veryimportant

Reasons for Seeking Employment at Your Charter School,Rated by Mean Scores

1

2

3

4

5

Conven-ient

location

More emphasison academics

thanextracurricular

activities

My interestin an

educationalreform effort

Promisesmade bycharterschool's

spokespersons

Academicreputation

of this school

Parentsare

committed

Safetyat

school

Difficultyto findother

positions

Opportunityto work withlike-mindededucators

This schoolhas small

class sizes

Veryimportant

Notimportant

Appendix_B1_CSS_Tch-Staff_98-99 3 WMU Charter School Survey

16. Rate each of the following statements as to what you expected when you first began workingat this school (initial expectation)and how you would rate it today (current experience).

Initial Expectation Current Experience

False PartlyTrue

True Mean STD Don'tknow

Mis-sing

False Partlytrue

True Mean STD Don'tknow

Mis-sing

1 2 3 1 2 3

Students willbe/are eagerandmotivated tolearn

8.3% 32.2% 49.9% 2.46 0.66 26 17 9.2% 53.0% 30.9% 2.37 0.69 9 22

The quality ofinstructionwill be/is high

2.2% 18.6% 70.2% 2.75 0.49 18 22 4.3% 35.3% 53.0% 2.45 0.72 11 22

Students willreceive/receivesufficientindividualattention

1.3% 17.2% 71.8% 2.78 0.45 21 22 7.4% 36.9% 47.9% 2.39 0.71 9 26

Parents willbe/are able toinfluence thedirection andactivities atthe school

12.1% 25.5% 46.5% 2.41 0.73 49 22 22.1% 39.8% 23.7% 2.62 0.61 37 27

There willbe/is goodcommunica-tion betweenthe schooland parents/guardians

4.3% 17.4% 67.6% 2.71 0.55 26 22 11.4% 35.3% 45.0% 2.54 0.60 13 24

Students will be/are eager and motivated to learn

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

The quality of instruction will be/is high

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Students will receive/receive sufficient individual attention

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Parents will be/are able to influence the school's direction andactivities

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

There will be/is good communication between the school andparents

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Appendix_B1_CSS_Tch-Staff_98-99 4 WMU Charter School Survey

Rate each of the following statements as to what you expected when you first began workingat this school (initial expectation)and how you would rate it today (current experience).

Initial Expectation Current Experience

False PartlyTrue

True Mean STD Don'tknow

Mis-sing

False Partlytrue

True Mean STD Don'tknow

Mis-sing

1 2 3 1 2 3

Students willhave/haveaccess tocomputersand other newtechnologies

5.6% 13.9% 71.1% 2.72 0.57 22 20 12.3% 26.6% 54.8% 2.59 0.54 6 22

The schoolwill have/haseffectiveleadershipand admin-istration

0.7% 14.1% 75.6% 2.83 0.40 20 23 11.9% 32.2% 47.7% 2.25 0.77 10 27

The schoolwill have/hassmall classsizes

2.9% 12.5% 74.9% 2.80 0.48 16 27 6.0% 22.6% 63.5% 2.52 0.63 7 28

Schoolpersonnel willbe/areaccountablefor theachievement/performanceof students

2.2% 20.6% 65.8% 2.72 0.50 33 18 5.1% 32.2% 54.1% 2.66 0.55 14 24

Theachievementlevels ofstudents willimprove/areimproving

0.4% 17.4% 70.0% 2.79 0.42 35 19 2.2% 31.5% 53.9% 2.25 0.73 34 21

The school will have/has effective leadership and administration

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

The school will have/has small class sizes

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

School personnel will be/are accountable for theachievement/performance of students

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

The achievement levels of students will improve/are improving

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Students will have/have access to computers and other newtechnologies

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Appendix_B1_CSS_Tch-Staff_98-99 5 WMU Charter School Survey

Rate each of the following statements as to what you expected when you first began workingat this school (initial expectation)and how you would rate it today (current experience).

Initial Expectation Current Experience

False PartlyTrue

True Mean STD Don'tknow

Mis-sing

False Partlytrue

True Mean STD Don'tknow

Mis-sing

1 2 3 1 2 3

Supportservices (i.e.,counseling,health care,etc.) willbe/areavailable tostudents

3.4% 20.1% 63.5% 2.69 0.54 36 22 18.1% 31.5% 40.9% 2.60 0.61 18 24

The schoolwill support/issupportinginnovativepractices

0.4% 16.3% 70.0% 2.80 0.41 40 19 6.7% 29.5% 52.6% 2.21 0.74 27 23

Teachers willbe able toinfluence thesteering anddirection ofthe school

3.6% 22.1% 60.6% 2.23 0.61 38 23 15.0% 36.7% 36.7% 2.85 0.37 27 25

There willbe/are newprofessionalopportunitiesfor teachers

4.9% 20.8% 51.0% 2.53 0.58 85 19 15.2% 33.3% 32.2% 2.65 0.54 61 25

Teachers willbe/arecommitted tothe mission ofthe school

0.2% 13.4% 76.3% 2.44 0.64 27 18 2.7% 27.1% 62.4% 2.76 0.46 14 21

Teachers willbe/areautonomousand creativein theirclassrooms

1.3% 18.6% 70.2% 2.02 0.73 26 18 2.2% 27.1% 61.7% 2.65 0.53 18 22

The school will support/is supporting innovative practices

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Teachers will be able to influence the steering and direction of theschool

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

There will be/are new professional opportunities for teachers

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Teachers will be/are committed to the mission of the school

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Support services will be/are available to students

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Teachers will be/are autonomous and creative in their classrooms

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Appendix_B1_CSS_Tch-Staff_98-99 6 WMU Charter School Survey

17. Rate your level of satisfaction with the following aspects or features of your school.Percentages

Mean STD NDon'tknow Missing

1 2 3 4 5

Salary level 11.2% 19.5% 34.4% 22.3% 12.6% 3.06 1.17 430 11 6

Fringe benefits 8.9% 16.3% 28.9% 26.7% 19.3% 3.31 1.21 405 32 10

Relations with the community at large 3.2% 10.4% 37.3% 29.9% 19.3% 3.52 1.02 405 30 12

School mission statement 1.4% 3.6% 18.7% 38.8% 37.6% 4.07 0.91 418 17 12

Ability of the school to fulfill its stated mission 5.2% 9.2% 26.3% 35.8% 23.5% 3.63 1.10 422 15 10

Evaluation or assessment of your performance 7.0% 7.7% 22.4% 33.0% 29.9% 3.71 1.17 388 46 13

Resources available for instruction 11.8% 16.4% 25.6% 28.5% 17.6% 3.24 1.26 414 19 14

School buildings and facilities 14.8% 19.6% 26.1% 21.9% 17.6% 3.08 1.31 433 2 12

Availability of computers and other technology 11.1% 11.3% 17.3% 27.7% 32.6% 3.59 1.34 433 3 11

School governance 8.9% 13.2% 28.8% 28.8% 20.4% 3.39 1.20 417 16 14

Administrative leadership of school 6.7% 12.0% 22.9% 25.2% 33.1% 3.66 1.24 432 7 8

Not verysatisfied

Verysatisfied

Level of Satisfaction with Aspects or Features of Your School,Rated by Mean Scores

1

2

3

4

5

Salarylevel

Fringebenefits

Relationswith the

communityat large

Schoolmission

statement

Ability ofschool tofulfill its

stated mission

Evaluation orassessment

of yourperformance

Resourcesavailable forinstruction

Schoolbuildings

and facilities

Availabilityof computers

and othertechnology

Schoolgovernance

Administrativeleadershipof school

Verysatisfied

Verydis-

satisfied

Appendix_B1_CSS_Tch-Staff_98-99 7 WMU Charter School Survey

18. To what extent do you agree or disagree with the following statements about your school?

Mean STD NDon'tknow

Mis-sing

1 2 3 4 5

This school ismeetingstudents'needs thatcould not beaddressed atother localschools

3.7% 7.2% 23.5% 28.5% 37.1% 3.88 1.10 404 24 19

Students feelsafe at thisschool

1.9% 2.8% 11.4% 29.4% 54.5% 4.32 0.92 422 8 17

Class sizes aretoo large tomeet theindividualstudent'sneeds

48.5% 24.1% 15.4% 7.3% 4.7% 1.96 1.17 423 6 18

Teachers aredisenchantedwith whatcan beaccomplishedat this school

20.7% 21.0% 33.2% 15.1% 10.1% 2.73 1.23 377 39 31

Teachers areinvolved indecisionmaking at thisschool

9.7% 12.3% 30.4% 25.8% 21.7% 3.38 1.23 391 30 26

Stronglydisagree

Stronglyagree

Students feel safe at this school

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Class sizes are too large to meet the individual student's needs

Strongly disagree

Disagree Neither agree ordisagree

Agree Strongly agree

0%

20%

40%

60%

80%

100%

Teachers are disenchanted with whatcan be accomplished at this school

Strongly disagree Disagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

Teachers are involved in decision making at this school

Strongly disagree Disagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

This school is meeting students' needs that could not be addressed atother local schools

Strongly disagreeDisagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B1_CSS_Tch-Staff_98-99 8 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD NDon'tknow

Mis-sing

1 2 3 4 5

The school hassufficientfinancialresources

12.4% 12.1% 35.3% 25.9% 14.4% 3.18 1.19 348 77 22

I am satisfiedwith theschool'scurriculum

5.0% 11.6% 28.5% 35.5% 19.4% 3.53 1.08 397 23 27

Parents aresatisfied withthe instruction

0.6% 5.2% 29.5% 43.5% 21.3% 3.80 0.85 329 91 27

Teachers arechallenged tobe effective

2.3% 7.0% 21.9% 33.2% 35.7% 3.93 1.03 398 19 30

This schoolhas been wellreceived by thecommunity

0.8% 5.8% 23.9% 32.6% 36.8% 3.99 0.96 380 48 19

Stronglydisagree

Stronglyagree

I am satisfied with the school's curriculum

Strongly disagreeDisagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Parents are satisfied with the instruction

Strongly disagreeDisagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Teachers are challenged to be effective

Strongly disagreeDisagree

Neither agree ordisagree

Agree Strongly agree

0%

20%

40%

60%

80%

100%

This school has been well received by the community

Strongly disagreeDisagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

The school has sufficient financial resources

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B1_CSS_Tch-Staff_98-99 9 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD NDon'tknow

Mis-sing

1 2 3 4 5

I think thisschool has abright future

1.7% 1.9% 15.5% 30.4% 50.5% 4.26 0.91 414 12 21

Too manychanges areoccurring at theschool

18.8% 23.2% 27.6% 18.0% 12.4% 2.82 1.28 410 15 22

This schoolreflects acommunityatmosphere

3.8% 7.2% 26.9% 32.5% 29.7% 3.77 1.07 391 31 25

Extracurricularactivities are notemphasized atthe expense ofacademics

2.9% 5.2% 17.5% 23.8% 50.7% 4.14 1.06 383 38 26

This school hashigh standardsandexpectations forstudents

3.6% 4.5% 16.2% 33.1% 42.6% 4.07 1.04 420 5 22

This school hasgood physicalfacilities

18.5% 22.2% 26.7% 20.0% 12.6% 2.86 1.28 405 7 35

Stronglydisagree

Stronglyagree

Too many changes are occurring at the school

Strongly disagreeDisagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

This school reflects a community atmosphere

Strongly disagreeDisagree

Neither agree ordisagree

Agree Strongly agree

0%

20%

40%

60%

80%

100%

Extracurricular activities are not emphasized at the expense of academics

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

This school has high standards and expectations for students

Strongly disagree Disagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

I think this school has a bright future

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

This school has good physical facilities

Strongly disagree Disagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

Appendix_B1_CSS_Tch-Staff_98-99 10 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD NDon'tknow

Mis-sing

1 2 3 4 5

Parents areinvolved andcan influenceinstruction andschool activities

16.5% 17.0% 32.2% 21.1% 13.2% 2.97 1.25 394 34 19

Teachers andschoolleadership areaccountable forstudentachievement/performance

2.5% 4.0% 24.3% 36.1% 33.2% 3.94 0.98 404 23 20

Students aresatisfied withthe instruction

1.6% 7.1% 24.2% 49.7% 17.4% 3.74 0.88 368 48 31

Lack of studentdisciplinehinders myability to teachand theopportunity forother studentsto learn

22.3% 14.9% 20.3% 20.0% 22.5% 3.06 1.46 395 29 23

Teachers areinsecure abouttheir future atthis school

21.0% 17.4% 25.1% 22.9% 13.5% 2.91 1.33 362 64 21

Teachers havemanynoninstructionalduties

15.3% 14.7% 23.6% 22.3% 24.1% 3.25 1.37 373 53 21

Stronglydisagree

Stronglyagree

Teachers and school leadership are accountable for student achievement/performance

Strongly disagree Disagree

Neither agree ordisagree

Agree Strongly agree

0%

20%

40%

60%

80%

100%

Students are satisfied with the instruction

Strongly disagreeDisagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Lack of student discipline hinders my ability to teach and the opportunity forother students to learn

Strongly disagreeDisagree

Neither agree ordisagree Agree Strongly agree

0%

20%

40%

60%

80%

100%

Teachers are insecure about their future at this school

Strongly disagree Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Parents are involved and can influence instruction and school activities

Strongly disagree Disagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

Teachers have many noninstructional duties

Strongly disagree Disagree

Neither agree ordisagree Agree Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B1_CSS_Tch-Staff_98-99 11 WMU Charter School Survey

2000 WMU Charter School SurveyInformant Group: Teachers/Staff (N=537) Descriptive statistics

1. What is your role at this school?

TeacherTeachingassistant

SpecialistStudentteacher

CAO/Administrator

OtherTotal Missing

N 280 62 39 0 44 110 535 2

% 52.3% 11.6% 7.3% 0.0% 8.2% 20.6% 100.0%

2. What is your current teaching certification status (teachers only)?

Total

N 210 12 50 8 280

% 100.0%

3. Are you teaching in a subject area in which you are certified to teach?

Yes NoNot

applicable Total(teachers only)

N 208 31 38 277

% 75.1% 11.2% 13.7% 100.0%

4. With which grade do you mostly work?Grade Level

K 1st 2nd 3td 4th 5th 6th 7th 8th 9th 10th 11th 12th Total Missing

N 38 45 28 33 21 28 25 38 19 36 30 14 32 139 526 11

% 7.2% 8.6% 5.3% 6.3% 4.0% 5.3% 4.8% 7.2% 3.6% 6.8% 5.7% 2.7% 6.1% 26.4% 100.0%

5. What is your age?Youngerthan 20

20-29 30-39 40-4950 orolder

Total Missing

N 6 211 154 108 51 530 7

% 1.1% 39.8% 29.1% 20.4% 9.6% 100.0%

6. What is your race/ethnicity?

White Black HispanicAsian/Pac.

IslanderNative

AmericanTotal Missing

N 284 179 35 11 9 518 19

% 54.8% 34.6% 6.8% 2.1% 1.7% 100.0%

7. What is your gender?

Female Male Total Missing

N 352 158 510 27

% 69.0% 31.0% 100.0%

Note: Questions 2 and 3 include the responses from only those staff who indicated that they were teachers

Totals for All Pennsylvania Charter School Teachers and Staff, 1999-2000

Currently certified toteach in this state

Currently certified toteach in another state

Workingto obtain

certification

Notapplicable

Not certified and notworking to obtain

certification

75.0% 4.3% 17.9% 2.9%

Role in school

0%

20%

40%

60%

80%

100%

Teacher Specialist CAO/Administrator

Distribution by Race/Ethnicity

0%

20%

40%

60%

80%

100%

White Black Hispanic Asian/Pac.Islander

NativeAmerican

Gender

Female

Male

0%

20%

40%

60%

80%

100%

Teaching a subject in which you are certified

Notapplicable

Yes

No

0%

20%

40%

60%

80%

100%

Current certification status

0%

20%

40%

60%

80%

100%

Currently certified toteach in this state

Currently certified toteach in another

state

Working to obtaincertification

Not certified & notworking to obtain

one

Age

0%

20%

40%

60%

80%

100%

Younger than20

20-29 30-39 40-49 50 orolder

Appendix_B2_CSS_Tch-Staff_99-00 1 WMU Charter School Survey

8. How many years of experience have you had in each 9. Years atof these types of schools (teachers only) current school?

Privateschool

Parochialschool

Charterschool

Publicschool

Other TotalTotal

(excluding"other")

Years atcurrentschool

Mean 0.95 0.31 1.69 1.84 0.43 5.23 4.80 1.72

STD 2.54 1.37 0.82 3.28 1.51 4.50 4.36 0.94

10. How much formal education have you had (teachers only)Did not

completehigh school

Completedhigh

school

Less than4 years

of college

CollegegraduateBA/BS

Graduatecourses,

no degree

Graduate/professional

degreeTotal

Mis-sing

N 0 0 3 123 86 67 279 258

% 0.0% 0.0% 1.1% 44.1% 30.8% 24.0% 100.0%

11. What is the highest college degree you hold? (teachers only)

Bachelors Masters5-6- yearCertificate

Doctorate Total Missing

N 207 57 4 7 275 262

% 75.3% 20.7% 1.5% 2.5% 100.0%

12a. Are you working toward another degree at this time?No Yes Total Missing

N 337 171 508 29

% 66.3% 33.7% 100.0%

12b. If yes, what degree?

Bachelors Masters5-6- yearCertificate

Doctorate Total Missing

N 22 121 12 17 172 365

% 12.8% 70.3% 7.0% 9.9% 100.0%

13a. Are you aware of the school's mission?No Yes Total Missing

N 16 502 518 19

% 3.1% 96.9% 100.0%

13b. If yes, to what extent is the missionbeing followed by the school?

Not verywell

Fair WellVerywell

Total Missing

1 2 3 4

N 22 116 224 138 500 37

% 4.4% 23.2% 44.8% 27.6% 100.0%

14. Do you plan (hope) to be working at this school next year?No Yes Total Missing

N 111 380 491 46

% 22.6% 77.4% 100.0%

Note: Questions 8, 9, 10 and 11 include the responses from only those staff who indicated that they were teachers

Average years of experience by school type

0

2

4

6

8

10

Privateschool

Parochialschool

Charterschool

Publicschool

Other

Yea

rs

Level of formal education

0%

20%

40%

60%

80%

100%

Did notcomplete

high school

Completedhigh

school

Less than4 years

of college

CollegegraduateBA/BS

Graduatecourses,

no degree

Graduate/professional

degree

Highest college degree

Bachelors

MastersDoctorate

5-6- yearCertificate0%

20%

40%

60%

80%

100%

Are you working toward another degree?

NoYes

0%

20%

40%

60%

80%

100%

Degree you are working toward

Bachelors

Masters

5-6- yearCertificate Doctorate0%

20%

40%

60%

80%

100%

Are you aware of the school's mission?

No

Yes

0%

20%

40%

60%

80%

100%

To what extent is the school mission being followed?

Not very wellFair

Well Verywell

0%

20%

40%

60%

80%

100%

Do you plan (hope) to be teaching at this school next year?

No

Yes

0%

20%

40%

60%

80%

100%

Appendix_B2_CSS_Tch-Staff_99-00 2 WMU Charter School Survey

15. Rate the importance of the following factors in your decision to seekemployment at this school.

Percentages

Mean STD Median N Missing

1 2 3 4 5

Convenient location 18.9% 13.0% 28.0% 18.7% 21.4% 3.11 1.39 3 529 8

More emphasis on academics as opposed toextracurricular activities

5.8% 10.4% 33.4% 29.9% 20.5% 3.49 1.10 4 521 16

My interest in being involved in an educational reformeffort

3.1% 3.5% 20.7% 31.5% 41.3% 4.04 1.02 4 518 19

Promises made by charter school's spokespersons 10.4% 9.8% 26.5% 27.9% 25.4% 3.48 1.26 4 520 17

Academic reputation (high standards) of this school 7.8% 7.6% 28.3% 28.9% 27.3% 3.60 1.19 4 512 25

Parents are committed 13.4% 10.3% 25.5% 24.1% 26.6% 3.40 1.34 4 522 15

Safety at school 6.3% 6.3% 19.4% 24.4% 43.5% 3.92 1.20 4 520 17

Difficulty to find other positions 36.8% 15.0% 23.7% 13.0% 11.5% 2.47 1.39 2 506 31

Opportunity to work with like-minded educators 2.9% 3.5% 20.4% 33.1% 40.1% 4.04 1.00 4 514 23

This school has small class sizes 3.5% 3.7% 21.0% 31.9% 40.0% 4.01 1.04 4 518 19

Notimportant

Veryimportant

Reasons for Seeking Employment at Your Charter School,Rated by Mean Scores

1

2

3

4

5

Conven-ient

location

More emphasison academics

thanextracurricular

activities

My interestin an

educationalreform effort

Promisesmade bycharterschool's

spokespersons

Academicreputation

of this school

Parentsare

committed

Safetyat

school

Difficultyto findother

positions

Opportunityto work withlike-mindededucators

This schoolhas small

class sizes

Veryimportant

Notimportant

Appendix_B2_CSS_Tch-Staff_99-00 3 WMU Charter School Survey

16. Rate each of the following statements as to what you expected when you first began workingat this school (initial expectation)and how you would rate it today (current experience).

Initial Expectation Current Experience

False PartlyTrue

True Mean STD Don'tknow

Mis-sing

False Partlytrue

True Mean STD Don'tknow

Mis-sing

1 2 3 1 2 3

Students willbe/are eagerandmotivated tolearn

9.6% 40.7% 49.8% 2.40 0.66 21 24 10.6% 60.1% 29.3% 2.19 0.60 6 40

The quality ofinstructionwill be/is high

2.2% 23.7% 74.1% 2.72 0.50 21 26 4.1% 43.2% 52.7% 2.49 0.58 10 39

Students willreceive/receivesufficientindividualattention

2.8% 23.2% 74.0% 2.71 0.51 17 24 11.4% 39.8% 48.8% 2.37 0.68 9 36

Parents willbe/are able toinfluence thedirection andactivities atthe school

15.4% 33.8% 50.8% 2.35 0.73 52 30 27.3% 45.9% 26.8% 2.00 0.74 33 38

There willbe/is goodcommunica-tion betweenthe schooland parents/guardians

6.8% 24.4% 68.9% 2.62 0.61 25 24 11.8% 41.6% 46.5% 2.35 0.68 10 37

Students will be/are eager and motivated to learn

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

The quality of instruction will be/is high

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Students will receive/receive sufficient individual attention

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Parents will be/are able to influence the school's direction andactivities

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

There will be/is good communication between the school andparents

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Appendix_B2_CSS_Tch-Staff_99-00 4 WMU Charter School Survey

Rate each of the following statements as to what you expected when you first began workingat this school (initial expectation)and how you would rate it today (current experience).

Initial Expectation Current Experience

False PartlyTrue

True Mean STD Don'tknow

Mis-sing

False Partlytrue

True Mean STD Don'tknow

Mis-sing

1 2 3 1 2 3

Students willhave/haveaccess tocomputersand other newtechnologies

6.8% 18.5% 74.7% 2.68 0.60 25 26 7.6% 31.5% 60.9% 2.53 0.63 5 33

The schoolwill have/haseffectiveleadershipand admin-istration

1.2% 15.9% 82.9% 2.82 0.42 24 29 10.7% 36.1% 53.2% 2.43 0.68 6 35

The schoolwill have/hassmall classsizes

2.1% 17.6% 80.4% 2.78 0.46 23 30 9.7% 29.4% 60.9% 2.51 0.67 6 37

Schoolpersonnel willbe/areaccountablefor theachievement/performanceof students

2.8% 28.0% 69.3% 2.67 0.53 38 27 3.2% 39.6% 57.3% 2.54 0.56 25 37

Theachievementlevels ofstudents willimprove/areimproving

1.5% 24.9% 73.5% 2.72 0.48 46 26 3.3% 41.5% 55.1% 2.52 0.56 23 35

The school will have/has effective leadership and administration

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

The school will have/has small class sizes

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

School personnel will be/are accountable for theachievement/performance of students

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

The achievement levels of students will improve/are improving

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Students will have/have access to computers and other newtechnologies

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Appendix_B2_CSS_Tch-Staff_99-00 5 WMU Charter School Survey

Rate each of the following statements as to what you expected when you first began workingat this school (initial expectation)and how you would rate it today (current experience).

Initial Expectation Current Experience

False PartlyTrue

True Mean STD Don'tknow

Mis-sing

False Partlytrue

True Mean STD Don'tknow

Mis-sing

1 2 3 1 2 3

Supportservices (i.e.,counseling,health care,etc.) willbe/areavailable tostudents

3.5% 28.6% 67.8% 2.64 0.55 60 26 14.5% 40.4% 45.1% 2.31 0.71 17 30

The schoolwill support/issupportinginnovativepractices

1.1% 25.2% 73.8% 2.73 0.47 43 29 6.8% 34.2% 59.0% 2.52 0.62 22 42

Teachers willbe able toinfluence thesteering anddirection ofthe school

4.2% 29.2% 66.6% 2.62 0.56 57 25 9.3% 42.4% 48.3% 2.39 0.65 20 33

There willbe/are newprofessionalopportunitiesfor teachers

5.1% 32.4% 62.5% 2.57 0.59 97 32 14.0% 39.5% 46.4% 2.32 0.71 59 43

Teachers willbe/arecommitted tothe mission ofthe school

0.8% 16.6% 82.6% 2.82 0.41 35 26 2.7% 34.6% 62.8% 2.60 0.54 16 32

Teachers willbe/areautonomousand creativein theirclassrooms

1.7% 19.9% 78.4% 2.77 0.46 39 25 3.6% 31.5% 64.8% 2.61 0.56 12 30

The school will support/is supporting innovative practices

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Teachers will be able to influence the steering and direction of theschool

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

There will be/are new professional opportunities for teachers

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Teachers will be/are committed to the mission of the school

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Support services will be/are available to students

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Teachers will be/are autonomous and creative in their classrooms

0%

20%

40%

60%

80%

100%

False Partly true True

Initial expectation Current experience

Appendix_B2_CSS_Tch-Staff_99-00 6 WMU Charter School Survey

17. Rate your level of satisfaction with the following aspects or features of your school.Percentages

Mean STD Median NDon'tknow Missing

1 2 3 4 5

Salary level 12.8% 19.2% 38.6% 22.7% 6.8% 2.91 1.09 3.00 516 8 13

Fringe benefits 7.0% 13.8% 30.9% 31.1% 17.1% 3.38 1.13 3.00 485 34 18

Relations with the community at large 2.7% 11.5% 38.7% 30.9% 16.4% 3.47 0.98 3.00 489 30 18

School mission statement 2.8% 3.2% 23.9% 37.7% 32.5% 3.94 0.97 4.00 507 11 19

Ability of the school to fulfill its stated mission 6.5% 7.7% 29.4% 37.7% 18.7% 3.54 1.08 4.00 504 15 18

Evaluation or assessment of your performance 5.7% 9.5% 22.5% 37.0% 25.4% 3.67 1.12 4.00 476 41 20

Resources available for instruction 7.5% 17.2% 28.5% 26.9% 19.8% 3.34 1.19 3.00 494 22 21

School buildings and facilities 13.9% 23.4% 31.3% 20.7% 10.8% 2.91 1.19 3.00 518 2 17

Availability of computers and other technology 7.9% 14.5% 20.5% 24.8% 32.3% 3.59 1.29 4.00 517 7 13

School governance 9.9% 10.8% 32.5% 28.4% 18.5% 3.35 1.19 3.00 493 22 22

Administrative leadership of school 7.0% 11.4% 24.0% 28.1% 29.5% 3.62 1.21 4.00 516 5 16

Not verysatisfied

Verysatisfied

Level of Satisfaction with Aspects or Features of Your School,Rated by Mean Scores

1

2

3

4

5

Salarylevel

Fringebenefits

Relationswith the

communityat large

Schoolmission

statement

Ability ofschool tofulfill its

stated mission

Evaluation orassessment

of yourperformance

Resourcesavailable forinstruction

Schoolbuildings

and facilities

Availabilityof computers

and othertechnology

Schoolgovernance

Administrativeleadershipof school

Verysatisfied

Verydis-

satisfied

Appendix_B2_CSS_Tch-Staff_99-00 7 WMU Charter School Survey

18. To what extent do you agree or disagree with the following statements about your school?

Mean STD NDon'tknow

Mis-sing

1 2 3 4 5

This school ismeetingstudents'needs thatcould not beaddressed atother localschools

4.9% 9.0% 21.8% 31.0% 33.3% 3.79 1.15 487 32 18

Students feelsafe at thisschool

0.8% 2.8% 12.6% 28.7% 55.1% 4.35 0.86 506 11 20

Class sizes aretoo large tomeet theindividualstudent'sneeds

41.2% 26.2% 17.1% 8.5% 7.0% 2.14 1.24 515 5 17

Teachers aredisenchantedwith whatcan beaccomplishedat this school

17.4% 24.6% 34.8% 14.7% 8.5% 2.72 1.16 448 60 29

Teachers areinvolved indecisionmaking at thisschool

5.6% 9.3% 30.9% 32.8% 21.4% 3.55 1.10 482 30 25

Stronglydisagree

Stronglyagree

Students feel safe at this school

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Class sizes are too large to meet the individual student's needs

Strongly disagree

Disagree Neither agree ordisagree

Agree Strongly agree

0%

20%

40%

60%

80%

100%

Teachers are disenchanted with whatcan be accomplished at this school

Strongly disagreeDisagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

Teachers are involved in decision making at this school

Strongly disagreeDisagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

This school is meeting students' needs that could not be addressed atother local schools

Strongly disagreeDisagree

Neither agree ordisagree

Agree Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B2_CSS_Tch-Staff_99-00 8 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD NDon'tknow

Mis-sing

1 2 3 4 5

The school hassufficientfinancialresources

12.3% 21.1% 29.7% 23.3% 13.5% 3.05 1.22 407 104 26

I am satisfiedwith theschool'scurriculum

5.0% 10.4% 30.6% 33.5% 20.6% 3.54 1.08 481 35 21

Parents aresatisfied withthe instruction

2.7% 4.9% 31.0% 44.1% 17.3% 3.69 0.91 410 106 21

Teachers arechallenged tobe effective

1.6% 4.4% 18.7% 38.2% 37.0% 4.05 0.94 497 18 22

This schoolhas been wellreceived by thecommunity

1.1% 4.3% 25.2% 35.3% 34.2% 3.97 0.93 465 55 17

Stronglydisagree

Stronglyagree

I am satisfied with the school's curriculum

Strongly disagreeDisagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Parents are satisfied with the instruction

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Teachers are challenged to be effective

Strongly disagree Disagree

Neither agree ordisagree

Agree Strongly agree

0%

20%

40%

60%

80%

100%

This school has been well received by the community

Strongly disagreeDisagree

Neither agree ordisagree

Agree Strongly agree

0%

20%

40%

60%

80%

100%

The school has sufficient financial resources

Strongly disagree

Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B2_CSS_Tch-Staff_99-00 9 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD NDon'tknow

Mis-sing

1 2 3 4 5

I think thisschool has abright future

2.6% 2.6% 19.7% 30.7% 44.3% 4.11 0.99 492 21 24

Too manychanges areoccurring at theschool

18.0% 23.4% 30.3% 16.0% 12.3% 2.81 1.25 495 18 24

This schoolreflects acommunityatmosphere

2.4% 9.3% 29.9% 33.7% 24.6% 3.69 1.02 492 21 24

Extracurricularactivities are notemphasized atthe expense ofacademics

5.5% 6.0% 18.7% 22.3% 47.4% 4.00 1.18 470 43 24

This school hashigh standardsandexpectations forstudents

1.0% 4.3% 21.9% 30.6% 42.1% 4.08 0.95 506 10 21

This school hasgood physicalfacilities

22.9% 22.1% 28.8% 16.6% 9.7% 2.68 1.26 507 7 23

Stronglydisagree

Stronglyagree

Too many changes are occurring at the school

Strongly disagreeDisagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

This school reflects a community atmosphere

Strongly disagreeDisagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Extracurricular activities are not emphasized at the expense of academics

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

This school has high standards and expectations for students

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

I think this school has a bright future

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

This school has good physical facilities

Strongly disagree Disagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

Appendix_B2_CSS_Tch-Staff_99-00 10 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD NDon'tknow

Mis-sing

1 2 3 4 5

Parents areinvolved andcan influenceinstruction andschool activities

13.5% 21.9% 34.1% 18.5% 12.0% 2.94 1.19 475 37 25

Teachers andschoolleadership areaccountable forstudentachievement/performance

1.4% 6.7% 30.0% 36.1% 25.8% 3.78 0.95 496 19 22

Students aresatisfied withthe instruction

1.5% 7.4% 35.6% 39.7% 15.7% 3.61 0.89 458 53 26

Lack of studentdisciplinehinders myability to teachand theopportunity forother studentsto learn

21.4% 17.0% 21.7% 18.3% 21.7% 3.02 1.44 471 37 29

Teachers areinsecure abouttheir future atthis school

19.5% 23.2% 22.3% 19.8% 15.2% 2.88 1.34 435 83 19

Teachers havemanynoninstructionalduties

14.5% 19.0% 24.8% 20.5% 21.2% 3.15 1.34 468 48 21

Stronglydisagree

Stronglyagree

Teachers and school leadership are accountable for student achievement/performance

Strongly disagreeDisagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Students are satisfied with the instruction

Strongly disagreeDisagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Lack of student discipline hinders my ability to teach and the opportunity forother students to learn

Strongly disagreeDisagree

Neither agree ordisagree Agree Strongly agree

0%

20%

40%

60%

80%

100%

Teachers are insecure about their future at this school

Strongly disagree DisagreeNeither agree or

disagree AgreeStrongly agree

0%

20%

40%

60%

80%

100%

Parents are involved and can influence instruction and school activities

Strongly disagree

Disagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

Teachers have many noninstructional duties

Strongly disagreeDisagree

Neither agree ordisagree

Agree Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B2_CSS_Tch-Staff_99-00 11 WMU Charter School Survey

WMU Charter School SurveyInformant Group: Students (N=923) 1999 Descriptive statistics

1. In what grade are you this year?Grade level

5th 6th 7th 8th 9th 10th 11th 12th Total Missing

N 130 209 85 58 213 129 54 42 920 3

% 14.1% 22.7% 9.2% 6.3% 23.2% 14.0% 5.9% 4.6% 100.0%

2. How old are you? Years

Mean 13.74 9 10 11 12 13 14 15 16 17 18 19 20 Total

STD 2.30 N 1 46 139 163 86 115 141 118 64 28 12 7 920

Missing 3 % 0.1% 5.0% 15.1% 17.7% 9.3% 12.5% 15.3% 12.8% 7.0% 3.0% 1.3% 0.8% 100.0%

3. How many years, including this year, have youattended this school?

Years at current school

<1 1 2 3 4 5 Total Missing

N 303 506 93 12 3 2 919 4

% 33.0% 55.1% 10.1% 1.3% 0.3% 0.2% 100.0%

4. What kind of school did you attend before enrolling in this school?Publicschool

Privateschool

Paro-chial

Homeschooled

Did notattend

Other Total Missing

N 739 58 79 8 1 23 908 15

% 81.4% 6.4% 8.7% 0.9% 0.1% 2.5% 100.0%

5. How many of your brothers and sisters are attending this or another charter school?Number of siblings attending charter schools

Mean 0.45 N 0 1 2 3 4 5 6 7+ Total

STD 0.84 % 632 214 51 16 5 2 0 3 923

Missing 0 68.5% 23.2% 5.5% 1.7% 0.5% 0.2% 0.0% 0.3% 100.0%

6. Would you recommend to a friend that he/she enroll in this school?

No Yes Not sure Total Missing

N 231 463 224 918 5

% 25.2% 50.4% 24.4% 100.0%

7. Do you maintain friendships with students from your old school?

No Yes Total Missing

N 112 801 913 10

% 12.3% 87.7% 100.0%

Totals for All Pennsylvania Students (Grades 5-12), 1998-1999

Distribution by Grade

0%

20%

40%

60%

80%

100%

5th 6th 7th 8th 9th 10th 11th 12th

Years at Current School

0%

20%

40%

60%

80%

100%

<1 1 2 3 4 5

Type of schooling before the charter school

0%

20%

40%

60%

80%

100%

Publicschool

Privateschool

Paro-chial

Homeschooled

Did notattend

Other

Would you recommend this school to a friend?

0%

20%

40%

60%

80%

100%

No Yes Not sure

Do you maintain friendships from your old school?

0%

20%

40%

60%

80%

100%

No Yes

Appendix_B3_CSS_Students_98-99 1 WMU Charter School Survey

8. How did you do in your previous school? (Self-rated)Excel-lent Good Average Poor

Unsatis-factory

Mean 2.37 1 2 3 4 5 Total

STD 1.04 N 194 340 255 87 35 911

Missing 12 % 21.3% 37.3% 28.0% 9.5% 3.8% 100.0%

9. How are you doing so far in this school? (Self-rated)Excel-lent Good Average Poor

Unsatis-factory

Mean 2.24 1 2 3 4 5 Total

STD 0.93 N 201 383 255 53 19 911

Missing 12 % 22.1% 42.0% 28.0% 5.8% 2.1% 100.0%

10. Compared to your previous school, how interestedare you in your school work?

Moreinterested

Aboutthe same

Lessinterested

Mean 1.75 1 2 3 Total

STD 0.75 N 399 341 169 909

Missing 14 % 43.9% 37.5% 18.6% 100.0%

11. What is your gender?Female Male Total Missing

N 440 466 906 17

% 48.6% 51.4% 100.0%

12. What is your race/ethnicity?

White Black HispanicAsian/Pac.

IslanderNative

AmericanTotal Missing

N 233 484 129 27 31 904 19

% 25.8% 53.5% 14.3% 3.0% 3.4% 100.0%

13a. Do you have paid employment outside 13b. If yes, how many hours do you work per week?of the home? No Yes Total Missing 0-8

hours9-15hours

16-25hours

26 or morehours Total Missing

N 759 114 873 50 N 33 24 34 26 117 806% 86.9% 13.1% 100.0% % 28.2% 20.5% 29.1% 22.2% 100.0%

14. Highest level of education you plan to complete?High

school2 yearcollege

4 yearcollege

Graduateschool

Not sureyet

Total Missing

N 35 73 325 272 207 912 11% 3.8% 8.0% 35.6% 29.8% 22.7% 100.0%

Performance at Previous and Charter School

0%

20%

40%

60%

80%

100%

Excel-lent

Good Average Poor Unsatis-factory

Previous school Charter

Distribution by Race/Ethnicity

0%

20%

40%

60%

80%

100%

White Black Hispanic Asian/Pac.Islander

NativeAmerican

Gender

0%

20%

40%

60%

80%

100%

Female Male

Interest in School Work Compared to Previous School

0%

20%

40%

60%

80%

100%

Moreinterested

Aboutthe same

Lessinterested

Highest Level of Education You Plan to Complete

0%

20%

40%

60%

80%

100%

Highschool

2 yearcollege

4 yearcollege

Graduateschool

Not sureyet

Appendix_B3_CSS_Students_98-99 2 WMU Charter School Survey

15. Why did you and your family choose this school?Percentages

Mean STD N Missing

1 2 3 4 5

This school has a convenient location 43.3% 11.7% 20.0% 7.1% 17.9% 2.45 1.53 896 27

My parents thought this school is better for me 13.3% 3.9% 11.2% 14.8% 56.8% 3.98 1.43 900 23

I was not doing very well at the previous school 43.5% 6.9% 14.5% 9.2% 25.9% 2.67 1.68 895 28

This school is smaller 42.5% 7.9% 13.7% 9.7% 26.3% 2.69 1.68 891 32

This school has better computers and other equipment 40.8% 11.2% 14.0% 8.9% 25.1% 2.66 1.65 894 29

This school is safer 29.1% 9.3% 16.2% 10.1% 35.3% 3.13 1.66 890 33

Teachers at the previous school did not help meenough

34.4% 8.3% 17.6% 10.2% 29.4% 2.92 1.65 890 33

We heard that teachers were better in this school 31.4% 10.1% 13.6% 13.2% 31.7% 3.04 1.66 892 31

My friends were attending this school 66.3% 9.2% 10.7% 4.2% 9.7% 1.82 1.33 890 33

This school has small classes 35.4% 8.3% 14.4% 10.5% 31.4% 2.94 1.69 902 21

Notimportant

Veryimportant

Reasons for Choosing Your Charter School, Rated by Mean Scores

1

2

3

4

5

This schoolhas a

convenientlocation

My parentsthought

this schoolis better for

me

I was notdoing verywell at thepreviousschool

This schoolis smaller

This schoolhas bettercomputersand otherequipment

This schoolis safer

Teachersat the

previousschool didnot help

me enough

We heardthat

teacherswere better

in thisschool

My friendswere

attendingthis school

This schoolhas smallclasses

Veryimportant

Notimportant

Appendix_B3_CSS_Students_98-99 3 WMU Student Charter School Survey

16. To what extent do you agree or disagree with the following statements?

Mean STD NDon'tknow

Mis-sing

1 2 3 4 5

I think I deservethe grades Ireceive

8.4% 5.0% 17.6% 15.0% 54.0% 4.01 1.29 859 54 10

I have morehomework atthis school thanI had at myprevious school

30.4% 10.3% 14.4% 8.6% 36.4% 3.10 1.69 876 38 9

I think that I amlearning morehere than at theprevious schoolI attended

19.5% 7.7% 15.5% 13.6% 43.6% 3.54 1.57 880 33 10

Students at thisschool aremore interestedin learning thanthey were at mylast school

30.6% 11.0% 22.1% 12.8% 23.4% 2.87 1.55 819 90 14

My parents areglad that Iattend thisschool

13.7% 6.5% 16.4% 14.0% 49.3% 3.79 1.45 833 66 24

This schoolprovidesenoughextracurricularactivities for me

44.5% 12.6% 14.1% 10.9% 17.9% 2.45 1.56 851 38 34

Stronglydisagree

Stronglyagree

More homework at this school than I had at my previous school

Strongly disagree

Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

I am learning more here than at the previous school I attended

Strongly disagree

Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Students at this school are more interested in learning

Strongly disagree

Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

My parents are glad that I attend this school

Strongly disagreeDisagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

This school provides enough extracurricular activities

Strongly disagree

Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

I think I deserve the grades I receive

Strongly disagreeDisagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B3_CSS_Students_98-99 4 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD NDon'tknow

Mis-sing

1 2 3 4 5

I thought theteachers at thisschool wouldbe better

21.7% 8.0% 22.0% 14.2% 34.1% 3.31 1.54 838 60 25

My parents askme every dayabout whathappened atschool

19.0% 10.0% 15.6% 13.9% 41.5% 3.49 1.56 886 15 22

I wish therewere morecourses I couldchoose from

14.3% 7.6% 15.4% 12.1% 50.5% 3.77 1.48 825 76 22

I have acomputeravailable atschool when Ineed one

27.9% 8.3% 15.7% 13.7% 34.5% 3.19 1.64 879 23 21

My grades aredeterminedalmost totallyby tests

23.5% 14.8% 28.5% 13.1% 20.1% 2.91 1.42 786 111 26

Students feelsafe at thisschool

20.9% 10.1% 20.9% 16.3% 31.6% 3.28 1.51 759 133 31

Stronglydisagree

Stronglyagree

My parents ask me every day about what happened at school

Strongly disagree

Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

I wish there were more courses I could choose from

Strongly disagreeDisagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

I have a computer available at school when I need one

Strongly disagree

Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

My grades are determined almost totally by tests

Strongly disagreeDisagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

Students feel safe at this school

Strongly disagree

Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

I thought the teachers at this school would be better

Strongly disagree

Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B3_CSS_Students_98-99 5 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD NDon'tknow

Mis-sing

1 2 3 4 5

I am aware ofthe mission ofmy school

13.8% 7.7% 19.8% 19.3% 39.4% 3.63 1.42 767 117 39

My teachersencourage meto think aboutmy future

14.3% 5.7% 15.1% 17.6% 47.4% 3.78 1.45 876 25 22

Studentsrespect oneanother andtheir property

37.9% 11.8% 21.1% 14.7% 14.5% 2.56 1.47 863 39 21

The schoolbuilding isclean and wellmaintained

27.2% 10.2% 19.0% 17.0% 26.6% 3.05 1.56 881 17 25

There are rulesat this schoolwe must follow

6.5% 3.6% 9.3% 15.0% 65.6% 4.30 1.18 881 16 26

There arestudents whodon't follow therules

16.1% 5.4% 15.8% 15.4% 47.4% 3.73 1.49 853 28 42

Stronglydisagree

Stronglyagree

My teachers encourage me to think about my future

Strongly disagreeDisagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Students respect one another and their property

Strongly disagree

Disagree

Neither agree ordisagree

Agree Strongly agree

0%

20%

40%

60%

80%

100%

The school building is clean and well maintained

Strongly disagree

Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

There are rules at this school we must follow

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

There are students who don't follow the rules

Strongly disagree

Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

I am aware of the mission of my school

Strongly disagreeDisagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B3_CSS_Students_98-99 6 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD NDon'tknow

Mis-sing

1 2 3 4 5

If the teacherleft the room,most studentswould continueto work on theirassignments

30.7% 11.4% 20.7% 17.6% 19.6% 2.84 1.51 876 27 20

Almost everyassignmentthat I turn in tothe teacher isreturned withcorrections andsuggestions forimprovement

12.8% 9.0% 22.8% 17.3% 38.1% 3.59 1.40 877 28 18

Students takeresponsibilityfor their ownachievement atthis school

12.2% 7.6% 22.1% 23.7% 34.4% 3.61 1.35 819 79 25

Teachers andadministratorsat this schoolknow me by myname

5.3% 3.7% 11.3% 14.6% 65.1% 4.30 1.14 865 33 25

My teacher isavailable to talkabout myclassroomperformance,(i.e., coursework,homework,grades, etc.)

7.9% 5.6% 14.2% 19.8% 52.4% 4.03 1.27 858 35 30

This school is agood choice forme

24.3% 8.8% 16.9% 14.2% 35.8% 3.28 1.60 840 45 38

Stronglydisagree

Stronglyagree

Almost every assignment that I turn in to the teacher is returned withcorrections and suggestions for improvement

Strongly disagreeDisagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Students take responsibility for their own achievement

Strongly disagreeDisagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Teachers and administrators know me by my name

Strongly disagree Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

My teacher is available to talk about my classroom performance

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

This school is a good choice for me

Strongly disagree

Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

If the teacher left the room, most students would continue to work ontheir assignments

Strongly disagree

Disagree

Neither agree ordisagree Agree Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B3_CSS_Students_98-99 7 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD NDon'tknow

Mis-sing

1 2 3 4 5

Teachers wantme to be inschool and askme why I wasn'tthere when Ihave beenabsent

18.5% 8.6% 19.3% 15.9% 37.7% 3.46 1.51 849 41 33

We work ingroups most ofthe time

14.5% 11.3% 32.0% 19.1% 23.0% 3.25 1.32 874 19 30

At this school, amistake isunderstood asa learningexperience

18.6% 10.2% 19.5% 18.6% 33.1% 3.37 1.49 832 59 32

A counselor isavailable for meto talk aboutpersonalproblems*

17.2% 6.6% 12.2% 16.8% 47.2% 3.70 1.52 564 58 301

A counselor isavailable for meto talk aboutacademicmatters*

17.2% 6.6% 16.1% 16.6% 43.5% 3.63 1.51 559 56 308

* The last two items were only to be answered by those middle and high school students who have access to counselors

Stronglydisagree

Stronglyagree

We work in groups most of the time

Strongly disagree Disagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

At this school, a mistake is understood as a learning experience

Strongly disagree

Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

A counselor is available for me to talk about personal problems*

Strongly disagree

Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

A counselor is available for me to talk about academic matters*

Strongly disagree

Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Teachers want me to be in school and ask me why I wasn't therewhen I have been absent

Strongly disagree

Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B3_CSS_Students_98-99 8 WMU Charter School Survey

2000 WMU Charter School SurveyInformant Group: Students (N=1,106) Descriptive statistics

1. In what grade are you this year?Grade level

5th 6th 7th 8th 9th 10th 11th 12th Total Missing

N 137 204 183 79 142 155 111 91 1102 4

% 12.4% 18.5% 16.6% 7.2% 12.9% 14.1% 10.1% 8.3% 100.0%

2. How old are you? Years

Mean 13.77 9 10 11 12 13 14 15 16 17 18 19 20 Total

STD 2.45 N 0 57 191 196 91 133 126 141 94 41 23 11 1104

Missing 2 % 0.0% 5.2% 17.3% 17.8% 8.2% 12.0% 11.4% 12.8% 8.5% 3.7% 2.1% 1.0% 100.0%

3. How many years, including this year, have youattended this school?

Years at current school

<1 1 2 3 4 5 Total Missing

N 283 230 509 63 10 7 1102 4

% 25.7% 20.9% 46.2% 5.7% 0.9% 0.6% 100.0%

4. What kind of school did you attend before enrolling in this school?Publicschool

Privateschool

Paro-chial

Homeschooled

Did notattend

Other Total Missing

N 869 89 90 13 3 28 1092 14

% 79.6% 8.2% 8.2% 1.2% 0.3% 2.6% 100.0%

5. Would you recommend to a friend that he/she enroll in this school?

No Yes Not sure Total Missing

N 288 554 257 1099 7

% 26.2% 50.4% 23.4% 100.0%

6. Do you maintain friendships with students from your old school?

No Yes Total Missing

N 156 935 1091 15

% 14.3% 85.7% 100.0%

Totals for All Pennsylvania Students (Grades 5-12), 1999-2000

Distribution by Grade

0%

20%

40%

60%

80%

100%

5th 6th 7th 8th 9th 10th 11th 12th

Years at Current School

0%

20%

40%

60%

80%

100%

<1 1 2 3 4 5

Type of schooling before the charter school

0%

20%

40%

60%

80%

100%

Publicschool

Privateschool

Paro-chial

Homeschooled

Did notattend

Other

Would you recommend this school to a friend?

0%

20%

40%

60%

80%

100%

No Yes Not sure

Do you maintain friendships from your old school?

0%

20%

40%

60%

80%

100%

No Yes

Appendix_B4_CSS_Students_99-00 1 WMU Charter School Survey

7. How did you do in your previous school? (Self-rated)Excel-lent Good Average Poor

Unsatis-factory

Mean 2.32 1 2 3 4 5 Total

STD 1.06 N 256 411 282 90 49 1088

Missing 18 % 23.5% 37.8% 25.9% 8.3% 4.5% 100.0%

8. How are you doing so far in this school? (Self-rated)Excel-lent Good Average Poor

Unsatis-factory

Mean 2.32 1 2 3 4 5 Total

STD 0.93 N 246 461 298 68 20 1093

Missing 13 % 22.5% 42.2% 27.3% 6.2% 1.8% 100.0%

9. Compared to your previous school, how interestedare you in your school work?

Moreinterested

Aboutthe same

Lessinterested

Mean 1.75 1 2 3 Total

STD 0.75 N 473 404 202 1079

Missing 27 % 43.8% 37.4% 18.7% 100.0%

10. What is your gender?Female Male Total Missing

N 552 524 1076 30

% 51.3% 48.7% 100.0%

11. What is your race/ethnicity?

White Black HispanicAsian/Pac.

IslanderNative

AmericanTotal Missing

N 288 551 169 26 39 1073 33

% 26.8% 51.4% 15.8% 2.4% 3.6% 100.0%

12. Highest level of education you plan to complete?High

school2 yearcollege

4 yearcollege

Graduateschool

Not sureyet

Total Missing

N 74 102 302 414 188 1080 26% 6.9% 9.4% 28.0% 38.3% 17.4% 100.0%

Performance at Previous and Charter School

0%

20%

40%

60%

80%

100%

Excel-lent

Good Average Poor Unsatis-factory

Previous school Charter

Distribution by Race/Ethnicity

0%

20%

40%

60%

80%

100%

White Black Hispanic Asian/Pac.Islander

NativeAmerican

Gender

0%

20%

40%

60%

80%

100%

Female Male

Interest in School Work Compared to Previous School

0%

20%

40%

60%

80%

100%

Moreinterested

Aboutthe same

Lessinterested

Highest Level of Education You Plan to Complete

0%

20%

40%

60%

80%

100%

Highschool

2 yearcollege

4 yearcollege

Graduateschool

Not sureyet

Appendix_B4_CSS_Students_99-00 2 WMU Charter School Survey

Why did you and your family choose this school?Percentages

Mean STD Median N Missing

1 2 3 4 5

This school has a convenient location 35.5% 11.6% 26.0% 11.2% 15.8% 2.60 1.46 3 1097 9

My parents thought this school is better for me 9.3% 6.1% 16.6% 19.4% 48.6% 3.92 1.31 4 1100 6

I was not doing very well at the previous school 40.2% 10.4% 16.9% 10.9% 21.6% 2.63 1.60 2 1093 13

This school is smaller 40.6% 9.7% 15.4% 10.8% 23.5% 2.67 1.63 2 1096 10

This school has better computers and other equipment 37.1% 11.2% 17.3% 10.9% 23.5% 2.72 1.60 3 1094 12

This school is safer 20.9% 9.7% 21.2% 14.9% 33.4% 3.30 1.53 3 1096 10

Teachers at the previous school did not help meenough

30.1% 10.9% 18.3% 16.1% 24.6% 2.94 1.57 3 1090 16

We heard that teachers were better in this school 27.4% 9.8% 20.0% 16.3% 26.5% 3.05 1.55 3 1095 11

My friends were attending this school 57.7% 11.4% 11.9% 7.9% 11.0% 2.03 1.42 1 1097 9

This school has small classes 36.3% 10.9% 17.8% 12.8% 22.2% 2.74 1.58 3 1102 4

Notimportant

Veryimportant

Reasons for Choosing Your Charter School, Rated by Mean Scores

1

2

3

4

5

This schoolhas a

convenientlocation

My parentsthought this

school isbetter for

me

I was notdoing verywell at thepreviousschool

This schoolis smaller

This schoolhas bettercomputersand otherequipment

This schoolis safer

Teachers atthe

previousschool did

not help meenough

We heardthat

teacherswere better

in thisschool

My friendswere

attendingthis school

This schoolhas smallclasses

Veryimportant

Notimportant

Appendix_B4_CSS_Students_99-00 3 WMU Student Charter School Survey

16. To what extent do you agree or disagree with the following statements?

Mean STD Median NDon'tknow

Mis-sing

1 2 3 4 5

I think I deservethe grades Ireceive

8.3% 6.7% 20.2% 18.1% 46.7% 3.88 1.29 4 1031 65 10

I have morehomework atthis school thanI had at myprevious school

23.4% 8.6% 18.5% 10.4% 39.1% 3.33 1.61 3 1075 27 4

I think that I amlearning morehere than at theprevious schoolI attended

16.4% 8.8% 16.4% 16.0% 42.4% 3.59 1.50 4 1059 40 7

Students at thisschool are moreinterested inlearning thanthey were at mylast school

24.8% 12.9% 25.5% 15.1% 21.6% 2.96 1.46 3 975 121 10

My parents areglad that Iattend thisschool

10.4% 6.0% 18.0% 17.2% 48.4% 3.87 1.35 4 1014 79 13

Stronglydisagree

Stronglyagree

More homework at this school than I had at my previous school

Strongly disagree

Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

I am learning more here than at the previous school I attended

Strongly disagree

Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Students at this school are more interested in learning

Strongly disagree

Disagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

My parents are glad that I attend this school

Strongly disagreeDisagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

I think I deserve the grades I receive

Strongly disagree Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B4_CSS_Students_99-00 4 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD Median NDon'tknow

Mis-sing

1 2 3 4 5

This schoolprovidesenoughextracurricularactivities for me

38.1% 14.9% 15.1% 12.6% 19.3% 2.60 1.55 2 1033 61 12

I thought theteachers at thisschool would bebetter

16.4% 11.3% 26.3% 17.0% 29.0% 3.31 1.42 3 1002 85 19

My parents askme every dayabout whathappened atschool

17.5% 9.0% 20.1% 15.7% 37.7% 3.47 1.50 4 1072 19 15

I wish therewere morecourses I couldchoose from

15.7% 7.0% 14.9% 13.7% 48.7% 3.73 1.50 4 1023 62 21

I have acomputer isavailable atschool when Ineed one

24.5% 8.8% 14.5% 13.7% 38.5% 3.33 1.62 4 1061 37 8

Stronglydisagree

Stronglyagree

My parents ask me every day about what happened at school

Strongly disagree

Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

I wish there were more courses I could choose from

Strongly disagree

Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

I have a computer is available at school when I need one

Strongly disagree

Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

I thought the teachers at this school would be better

Strongly disagreeDisagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

This school provides enough extracurricular activities

Strongly disagree

DisagreeNeither agree or

disagree AgreeStrongly agree

0%

20%

40%

60%

80%

100%

Appendix_B4_CSS_Students_99-00 5 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD Median NDon'tknow

Mis-sing

1 2 3 4 5

Students feelsafe at thisschool

16.3% 9.6% 22.0% 21.7% 30.4% 3.40 1.42 4 958 131 17

I am aware ofthe mission ofmy school

14.6% 8.6% 21.2% 21.2% 34.5% 3.52 1.41 3 934 149 23

My teachersencourage meto think aboutmy future

7.7% 5.3% 17.8% 23.6% 45.7% 3.94 1.24 4 1070 28 8

Studentsrespect oneanother andtheir property

33.9% 18.1% 24.4% 15.1% 8.5% 2.46 1.32 4 1048 48 10

The schoolbuilding is cleanand wellmaintained

21.8% 13.1% 24.4% 18.8% 21.9% 3.06 1.44 4 1072 20 14

Stronglydisagree

Stronglyagree

My teachers encourage me to think about my future

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Students respect one another and their property

Strongly disagree

Disagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

The school building is clean and well maintained

Strongly disagree

Disagree

Neither agree ordisagree

AgreeStrongly agree

0%

20%

40%

60%

80%

100%

I am aware of the mission of my school

Strongly disagreeDisagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Students feel safe at this school

Strongly disagreeDisagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B4_CSS_Students_99-00 6 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD Median NDon'tknow

Mis-sing

1 2 3 4 5

There are rulesat this schoolwe must follow

3.0% 3.8% 12.3% 16.4% 64.5% 4.36 1.03 5 1072 20 14

There arestudents whodon't follow therules

10.3% 7.7% 16.9% 18.5% 46.7% 3.84 1.36 4 1050 38 18

If the teacherleft the room,most studentswould continueto work on theirassignments

29.5% 15.8% 25.5% 15.5% 13.6% 2.68 1.39 3 1063 36 7

Almost everyassignment thatI turn in to theteacher isreturned withcorrections andsuggestions forimprovement

9.7% 12.2% 24.2% 20.4% 33.6% 3.56 1.32 4 1060 29 17

Teachers andadministratorsat this schoolknow me by myname.

5.4% 4.4% 12.5% 15.0% 62.7% 4.25 1.16 5 1072 21 13

Stronglydisagree

Stronglyagree

Almost every assignment that I turn in to the teacher is returned withcorrections and suggestions for improvement

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Teachers and administrators know me by my name

Strongly disagree Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

If the teacher left the room, most students would continue to work ontheir assignments

Strongly disagree

Disagree

Neither agree ordisagree

Agree Strongly agree

0%

20%

40%

60%

80%

100%

There are rules at this school we must follow

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

There are students who don't follow the rules

Strongly disagree Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B4_CSS_Students_99-00 7 WMU Charter School Survey

To what extent do you agree or disagree with the following statements?

Mean STD Median NDon'tknow

Mis-sing

1 2 3 4 5

My teacher isavailable to talkabout myclassroomperformance,(i.e., coursework,homework,grades, etc.)

5.9% 5.5% 17.6% 20.8% 50.2% 4.04 1.19 5 1054 43 9

This school is agood choice forme

23.2% 7.9% 18.9% 18.0% 32.0% 3.28 1.55 4 1031 57 18

Teachers wantme to be inschool and askme why I wasn'tthere when Ihave beenabsent

12.9% 9.5% 20.0% 18.4% 39.2% 3.61 1.41 4 1023 71 12

A counselor isavailable for meto talk aboutpersonalproblems

18.9% 7.5% 17.9% 12.6% 43.1% 3.54 1.55 4 610 76 420

A counselor isavailable for meto talk aboutacademicmatters*

15.2% 8.6% 19.2% 16.5% 40.5% 3.59 1.46 4 593 85 428

* The last two items were only to be answered by those middle and high school students who have access to counselors

Stronglydisagree

Stronglyagree

A counselor is available for me to talk about personal problems

Strongly disagree

Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

A counselor is available for me to talk about academic matters*

Strongly disagreeDisagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Teachers want me to be in school and ask me why I wasn't there whenI have been absent

Strongly disagreeDisagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

My teacher is available to talk about my classroom performance

Strongly disagree Disagree

Neither agree ordisagree

Agree

Strongly agree

0%

20%

40%

60%

80%

100%

This school is a good choice for me

Strongly disagree

Disagree

Neither agree ordisagree Agree

Strongly agree

0%

20%

40%

60%

80%

100%

Appendix_B4_CSS_Students_99-00 8 WMU Charter School Survey

WMU Charter School SurveyInformant Group: Parents (N=292) 1999 Descriptive statistics

1. In what grades do you have children enrolled in this charter school?Grade level

K 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 10th 11th 12th Total

First child 40 30 30 25 13 26 39 10 11 24 20 8 4 280Second child 9 10 11 5 13 2 4 3 2 2 3 2 0 66

Third child 1 2 5 4 1 2 1 1 1 2 1 0 1 22Fourth child 0 0 0 2 1 0 0 0 0 1 0 0 0 4

Total number 50 42 46 36 28 30 44 14 14 29 24 10 5 372Total percent 13.4% 11.3% 12.4% 9.7% 7.5% 8.1% 11.8% 3.8% 3.8% 7.8% 6.5% 2.7% 1.3% 100.0%

2a. Do you have other school-age children not attending this charter school? No Yes Total Missing

N 157 131 288 4% 54.5% 45.5% 100.0%

2b. If yes, in what type of school(s) are they enrolled? Publicschool

Privateschool

Parochialschool

Home-schooled

Anothercharter

Other Total Missing

N 92 7 9 2 1 13 124 168% 74.2% 5.6% 7.3% 1.6% 0.8% 10.5% 100.0%

3. Approximately how many miles do you live from this charter school?Miles

Mean 4.34 1 2 3 4 5 6 7 8 9 10 >10 Total

STD 3.47 N 72 45 37 26 15 14 9 10 3 11 35 277Missing 15 % 26.0% 16.2% 13.4% 9.4% 5.4% 5.1% 3.2% 3.6% 1.1% 4.0% 12.6% 100.0%

4. Approximately how many miles do you live from the nearest traditional public school where your childcould be enrolled? Miles

Mean 2.22 1 2 3 4 5 6 7 8 9 10 >10 Total

STD 2.05 N 158 46 31 9 17 4 4 5 0 1 5 280Missing 12 % 56.4% 16.4% 11.1% 3.2% 6.1% 1.4% 1.4% 1.8% 0.0% 0.4% 1.8% 100.0%

5. What is your gender? 6. Which best describes your household?

Female Male Total MissingTwo

parents/guardians

Singleparent/

guardianOther Total Missing

N 249 37 286 6 N 156 127 7 290 2% 87.1% 12.9% 100.0% % 53.8% 43.8% 2.4% 100.0%

7. What is the estimated annual income of your household/family?Less than$10,000

$10,000-$19,999

$20,000-$29,999

$30,000-$39,999

$40,000-$59,999

$60,000-$99,999

$100,000or more Total Missing

N 39 55 51 32 54 45 6 282 10% 13.8% 19.5% 18.1% 11.3% 19.1% 16.0% 2.1% 100.0%

8a. Are you aware of the school's mission? 8b. If yes, to what extent is the missionbeing followed by the school?

No Yes Total MissingNot very

wellFair Well

Verywell Total Missing

N 25 262 287 5 1 2 3 4

% 8.7% 91.3% 100.0% N 8 43 108 102 261 31% 3.1% 16.5% 41.4% 39.1% 100.0%

Totals for All Pennsylvania Charter School Parents, 1998-1999

Appendix_B5_CSS_Parents_98-99 1 WMU Charter School Survey

9. Do you have concerns about your child's safety in this school? No Yes Total Missing

N 231 53 284 8

% 81.3% 18.7% 100.0%

10a. Estimate the total number of hours that you

and other adults in your household have Ohours

1-3hours

4-6hours

7-9hours

10-12hours

More than12 hours Total Missing

served as a volunteer at the school N 113 93 48 4 7 23 288 4

during an average month? % 39.2% 32.3% 16.7% 1.4% 2.4% 8.0% 100.0%

10b. Is voluntary work required by the school? No Yes Total Missing

attending this charter school? N 84 152 236 56% 35.6% 64.4% 100.0%

11. What is your race/ethnicity? White Black HispanicAsian/Pac.

IslanderNative

AmericanTotal Missing

N 118 132 21 4 1 276 16

% 42.8% 47.8% 7.6% 1.4% 0.4% 100.0%

12. How much formal education

Did notcomplete

high school

Completedhigh

school

Less than4 years

of college

CollegegraduateBA/BS

Graduatecourses,

no degree

Graduate/profession-al degree

Total Missing

have you had? N 25 82 81 41 20 31 280 12

% 8.9% 29.3% 28.9% 14.6% 7.1% 11.1% 100.0%

13. What kind of school did your child previouslyPublicschool

Privateschool

Parochialschool

Home-schooled

Anothercharter

Other Total Missing

attend before this charter school? N 197 23 28 3 13 9 273 19

% 72.2% 8.4% 10.3% 1.1% 4.8% 3.3% 100.0%

14. Rate the importance of the following factors in your decision to enroll your child in this school.Percentages

Mean STD NMis-sing

1 2 3 4 5

a. Convenient location 26.8% 13.0% 23.2% 12.3% 24.6% 2.95 1.52 284 8

b. More emphasis on academics than extracurricular activities 2.8% 5.3% 18.7% 27.2% 45.9% 4.08 1.05 283 9

c. My interest in an educational reform effort 9.4% 8.7% 23.0% 26.5% 32.4% 3.64 1.27 287 5

d. Promises made by charter school's spokespersons 5.3% 2.8% 18.8% 28.7% 44.3% 4.04 1.11 282 10

e. Academic reputation (high standards) of this school 3.3% 5.8% 10.9% 24.1% 55.8% 4.23 1.07 274 18

f. Safety for my child 3.1% 3.8% 10.1% 16.4% 66.4% 4.39 1.03 286 6

g. I prefer the emphasis and educational philosophy of this school 3.5% 2.1% 14.2% 26.0% 54.2% 4.25 1.01 288 4

h. My child has special needs that were not met at previous school 24.4% 9.2% 15.5% 17.3% 33.6% 3.27 1.59 271 21

i. Good teachers and high quality of instruction 1.8% 3.2% 7.1% 17.7% 70.3% 4.52 0.89 283 9

j. I prefer a private school but could not afford it 28.7% 10.0% 20.1% 15.1% 26.2% 3.00 1.57 279 13

k. My child wanted to attend this school 16.3% 9.5% 22.3% 22.3% 29.7% 3.40 1.42 283 9

l. My child was performing poorly at previous school 38.8% 6.8% 16.9% 12.6% 24.8% 2.78 1.64 278 14

m. I was unhappy with the curriculum & instruction at previous school 20.8% 7.5% 21.9% 15.4% 34.4% 3.35 1.52 279 13

n. This school has good physical facilities 16.0% 17.1% 29.1% 18.2% 19.6% 3.08 1.33 275 17

o. Recommendations of teacher/official at my child's previous school 44.6% 10.1% 16.3% 9.4% 19.6% 2.49 1.59 276 16

Notimportant

Veryimportant

Appendix_B5_CSS_Parents_98-99 2 WMU Charter School Survey

15. Rate each of the following statements as to what you expected when you first began workingat this school (initial expectation)and how you would rate it today (current experience).

FalsePartlytrue

True Mean STDDon'tknow

Missing FalsePartlytrue

True Mean STDDon'tknow

Missing

a. My child will be/is motivated to learn 1.1% 15.1% 83.9% 2.83 0.41 8 5 4.4% 18.7% 76.9% 2.73 0.54 2 17

b. The quality of instruction will be/is high 0.7% 15.2% 84.1% 2.83 0.39 10 6 4.8% 25.7% 69.5% 2.65 0.57 5 18c. My child will receive/receives sufficient

individual attention1.1% 20.1% 78.9% 2.78 0.44 9 4 6.2% 25.2% 68.6% 2.62 0.60 5 13

d. I will be/am able to influence the directionand activities in the school

7.7% 36.5% 55.8% 2.48 0.64 24 8 10.5% 39.3% 50.2% 2.40 0.67 26 19

e. There will be/is good communicationbetween the school and my household

0.7% 14.4% 84.8% 2.84 0.39 7 8 4.4% 20.6% 75.0% 2.71 0.54 3 17

f. My child will have/has access to computersand other new technologies

2.6% 17.6% 79.9% 2.77 0.48 13 6 4.5% 21.0% 74.5% 2.70 0.55 7 18

g. The school will have/has effectiveleadership and administration

2.2% 16.0% 81.7% 2.79 0.46 13 11 6.5% 21.1% 72.4% 2.66 0.60 11 20

h. The school will have/has small class sizes 3.7% 14.8% 81.5% 2.78 0.50 9 12 7.6% 17.9% 74.4% 2.67 0.61 8 22i. School personnel will be/are accountable

for my child's achievement/performance4.1% 34.0% 61.9% 2.58 0.57 16 8 6.2% 35.3% 58.5% 2.52 0.61 17 17

j. My child's achievement levels of studentswill improve/is improving

1.8% 24.2% 74.0% 2.72 0.49 8 7 8.6% 21.8% 69.5% 2.61 0.64 7 19

k. Support services (i.e., counseling, healthcare, etc.) will be/are available to my child

10.5% 26.9% 62.6% 2.52 0.68 48 6 11.3% 31.2% 57.6% 2.46 0.69 40 21

l. The school will support/is supportinginnovative practices

0.4% 25.2% 74.4% 2.74 0.45 43 11 2.1% 29.9% 67.9% 2.66 0.52 34 24

m. I will be/am able to participate in volunteerwork and other activities

6.3% 17.4% 76.3% 2.70 0.58 25 14 6.5% 24.8% 68.7% 2.62 0.61 25 21

16. To what extent do you agree or disagree with the following statements about your child's school?Percentages

Mean STD NDon'tknow

Missing

1 2 3 4 5a.

4.9% 6.0% 21.4% 22.9% 44.7% 3.97 1.16 266 22 4

b. Students feel safe at this school 2.2% 1.8% 15.0% 24.8% 56.2% 4.31 0.94 274 15 3

c. Class sizes are too large to meet the individual student's needs 52.1% 22.1% 12.9% 5.7% 7.1% 1.94 1.23 280 8 4

d. Teachers are disenchanted with what can be accomplished 53.2% 11.7% 21.2% 9.5% 4.5% 2.00 1.24 222 59 11

e. Teachers are involved in decision making at this school 2.8% 3.2% 16.7% 26.9% 50.5% 4.19 1.01 216 69 7

f. This school has sufficient financial resources 11.7% 10.6% 33.9% 23.3% 20.6% 3.31 1.24 180 99 13

g. I am satisfied with the school's curriculum 2.5% 6.7% 17.3% 30.6% 43.0% 4.05 1.05 284 2 6

h. I am satisfied with the instruction offered 4.6% 4.6% 20.6% 29.1% 41.1% 3.98 1.10 282 2 8

i. Teachers are challenged to be effective 2.4% 4.4% 16.5% 26.2% 50.4% 4.18 1.02 248 39 5

j. This school has been well received by the community 4.6% 8.4% 20.5% 20.5% 46.0% 3.95 1.19 239 47 6

k. I think this school has a bright future 3.3% 3.7% 10.3% 17.9% 64.8% 4.37 1.03 273 9 10

l. Too many changes are occurring at the school 34.1% 22.9% 22.5% 8.8% 11.6% 2.41 1.34 249 35 8

m. This school reflects a community atmosphere 6.5% 3.5% 21.5% 30.8% 37.7% 3.90 1.15 260 22 10

n. Extracurricular activities are not emphasized at the expense of academics 6.2% 6.6% 18.5% 21.6% 47.1% 3.97 1.22 259 25 8

o. This school has high standards and expectation for students 2.5% 3.6% 12.5% 22.1% 59.3% 4.32 0.99 280 7 5

p. This school has small class sizes 6.1% 5.4% 22.7% 18.0% 47.8% 3.96 1.21 278 8 6

q. This school has good physical facilities 12.2% 9.9% 27.8% 23.6% 26.6% 3.43 1.31 263 19 10

r. This school has good administrative leadership 4.1% 6.3% 14.4% 25.1% 50.2% 4.11 1.12 271 9 12

s. Parents are involved and can influence instruction and school activities 4.0% 4.4% 22.5% 29.3% 39.8% 3.96 1.08 249 26 17

t.3.9% 5.1% 20.0% 26.7% 44.3% 4.02 1.10 255 18 19

Teachers and school leadership are accountable for studentachievement/performance

This school is meeting students' needs that could not be addressed atother local schools

Initial Expectation Current Experience

Stronglydisagree

Stronglyagree .

Appendix_B5_CSS_Parents_98-99 3 WMU Charter School Survey

WMU Charter School SurveyInformant Group: Parents (N=364) 2000 Descriptive statistics

1. In what grades do you have children enrolled in this charter school?Grade level

K 1st 2nd 3rd 4th 5th 6th 7th 8th 9th 10th 11th 12th Total

First child 47 25 31 19 21 27 43 36 14 32 32 22 9 358

Second child 6 11 12 8 12 10 8 7 2 5 5 3 1 90

Third child 2 1 1 1 0 0 0 2 0 0 0 0 0 7

Total number 55 37 44 28 33 37 51 45 16 37 37 25 10 455

Total percent 12.1% 8.1% 9.7% 6.2% 7.3% 8.1% 11.2% 9.9% 3.5% 8.1% 8.1% 5.5% 2.2% 100%

2. If you have other children attending anotherPublicschool

Privateschool

Parochialschool

Home-schooled

Anothercharter

Other Total Missing

K-12 school, in what type of school(s) are N 124 9 11 3 9 21 177 187they enrolled % 70.1% 5.1% 6.2% 1.7% 5.1% 11.9% 100%

3. Approximately how many miles do you live from this charter school?Miles

Mean 4.87 1 2 3 4 5 6 7 8 9 10 >10 Total

STD 3.71 N 80 49 44 23 40 12 8 9 2 21 61 349Missing 15 % 22.9% 14.0% 12.6% 6.6% 11.5% 3.4% 2.3% 2.6% 0.6% 6.0% 17.5% 100%

4. Approximately how many miles do you live from the nearest traditional public school where your childcould be enrolled? Miles

Mean 2.26 1 2 3 4 5 6 7 8 9 10 >10 Total

STD 2.29 N 216 49 28 18 19 5 3 5 1 2 11 357Missing 7 % 60.5% 13.7% 7.8% 5.0% 5.3% 1.4% 0.8% 1.4% 0.3% 0.6% 3.1% 100%

5. What is your gender? 6. Which best describes your household?

Female Male Total MissingTwo

parents/guardians

Singleparent/

guardianOther Total Missing

N 309 48 357 7 N 204 150 6 360 4% 86.6% 13.4% 100% % 56.7% 41.7% 1.7% 100%

7. What is the estimated annual income of your household/family?Less than$10,000

$10,000-$19,999

$20,000-$29,999

$30,000-$39,999

$40,000-$59,999

$60,000-$99,999

$100,000or more Total Missing

N 34 75 68 42 78 44 10 351 13% 9.7% 21.4% 19.4% 12.0% 22.2% 12.5% 2.8% 100%

8a. Are you aware of the school's mission? 8b. If yes, to what extent is the missionbeing followed by the school?

No Yes Total MissingNot very

wellFair Well

Verywell Total Missing

N 45 316 361 3 1 2 3 4

% 12.5% 87.5% 100% N 16 58 140 103 317 47% 5.0% 18.3% 44.2% 32.5% 100%

Totals for All Pennsylvania Charter School Parents, 1999-2000

Appendix_B6_CSS_Parents_99-00 1 WMU Charter School Survey

9. Do you have concerns about your child's safety in this school? No Yes Total Missing

N 299 57 356 8

% 84.0% 16.0% 100%

10a. Estimate the total number of hours that you

and other adults in your household have 0hours

1-3hours

4-6hours

7-9hours

10-12hours

More than12 hours Total Missing

served as a volunteer at the school N 172 104 44 17 7 15 359 5

during an average month? % 47.9% 29.0% 12.3% 4.7% 1.9% 4.2% 100%

10b. Is voluntary work required by the school? No Yes Total Missing

attending this charter school? N 148 165 313 51% 47.3% 52.7% 100%

11. What is your race/ethnicity? White Black HispanicAsian/Pac.

IslanderNative

AmericanTotal Missing

N 135 192 21 4 2 354 10

% 38.1% 54.2% 5.9% 1.1% 0.6% 100%

12. How much formal education

Did notcomplete

high school

Completedhigh

school

Less than4 years

of college

CollegegraduateBA/BS

Graduatecourses,

no degree

Graduate/profession-al degree

Total Missing

have you had? N 26 120 115 33 20 44 358 6

% 7.3% 33.5% 32.1% 9.2% 5.6% 12.3% 100%

13. What kind of school did your child previouslyPublicschool

Privateschool

Parochialschool

Home-schooled

Anothercharter

Other Total Missing

attend before this charter school? N 247 35 24 3 25 20 354 10

% 69.8% 9.9% 6.8% 0.8% 7.1% 5.6% 100%

14. Rate the importance of the following factors in your decision to enroll your child in this school.Percentages

Mean STD Median NMis-sing

1 2 3 4 5

a. Convenient location 24.2% 11.5% 25.6% 9.9% 28.7% 3.07 1.53 3 364 9

b. My interest in an educational reform effort 6.8% 9.3% 21.4% 22.0% 40.6% 3.80 1.25 4 364 9

c. Promises made by charter school's spokespersons 6.6% 4.3% 15.9% 25.9% 47.3% 4.03 1.18 4 364 17

d. Academic reputation (high standards) of this school 1.4% 3.7% 13.0% 24.3% 57.6% 4.33 0.93 5 364 10

e. Safety for my child 2.3% 3.1% 10.0% 12.9% 71.7% 4.49 0.96 5 364 14

f. I prefer the emphasis and educational philosophy of this school 2.3% 2.3% 12.7% 30.6% 52.0% 4.28 0.93 5 364 18

g. My child has special needs that were not met at previous school 25.9% 12.5% 16.1% 10.1% 35.4% 3.17 1.63 3 364 28

h. Good teachers and high quality of instruction 1.7% 1.7% 11.0% 16.7% 68.9% 4.49 0.88 5 364 10

I. I prefer a private school but could not afford it 28.7% 7.8% 17.8% 15.5% 30.2% 3.11 1.61 3 364 16

j. My child wanted to attend this school 16.8% 10.3% 25.9% 14.0% 33.0% 3.36 1.45 3 364 13

k. My child was performing poorly at previous school 34.8% 10.7% 18.0% 9.9% 26.7% 2.83 1.63 3 364 19

l. I was unhappy with the curriculum & instruction at previous school 23.5% 12.2% 14.2% 13.9% 36.2% 3.27 1.61 4 364 19

m. Recommendations of teacher/official at my child's previous school 46.1% 11.0% 15.9% 8.1% 18.8% 2.43 1.57 2 364 19

Notimportant

Veryimportant

Appendix_B6_CSS_Parents_99-00 2 WMU Parent Charter School Survey

15. Rate each of the following statements as to what you expected when you first began workingat this school (initial expectation) and how you would rate it today (current experience).

FalsePartlytrue

True Mean STDDon'tknow

Missing FalsePartlytrue

True Mean STDDon'tknow

Mis-sing

a. My child will be/is motivated to learn 1.5% 11.3% 87.2% 2.86 0.39 10 10 5.4% 20.1% 74.6% 2.69 0.57 6 24

b. The quality of instruction will be/is high 0.9% 13.0% 86.1% 2.85 0.38 20 13 5.1% 24.5% 70.4% 2.65 0.57 7 26

c. My child will receive/receives sufficientindividual attention

3.3% 14.5% 82.2% 2.79 0.48 18 14 9.3% 27.5% 63.3% 2.54 0.66 14 26

d. I will be/am able to influence the directionand activities in the school

11.3% 39.2% 49.5% 2.38 0.68 50 13 16.6% 43.4% 40.0% 2.23 0.72 42 27

e. There will be/is good communicationbetween the school and my household 3.0% 14.9% 82.1% 2.79 0.48 16 13 9.8% 22.0% 68.2% 2.58 0.66 3 25

f. My child will have/has access tocomputers and other new technologies 3.1% 11.1% 85.8% 2.83 0.45 26 14 8.5% 19.5% 72.0% 2.63 0.64 12 24

g. The school will have/has effectiveleadership and administration

1.2% 14.9% 83.8% 2.83 0.41 23 13 6.1% 27.0% 67.0% 2.61 0.60 13 21

h. The school will have/has small class sizes 2.8% 12.6% 84.7% 2.82 0.45 21 17 8.4% 19.3% 72.3% 2.64 0.63 17 26

i. School personnel will be/are accountablefor my child's achievement/performance

6.9% 28.1% 65.0% 2.58 0.62 26 18 12.3% 35.2% 52.5% 2.40 0.70 23 23

j. My child's achievement levels of studentswill improve/is improving

2.1% 17.6% 80.2% 2.78 0.46 19 16 5.5% 27.8% 66.7% 2.61 0.59 11 26

k. Support services (i.e., counseling, healthcare, etc.) will be/are available to my child

12.0% 22.0% 66.0% 2.54 0.70 59 14 17.8% 25.3% 56.8% 2.39 0.77 48 24

l. The school will support/is supportinginnovative practices

2.7% 16.7% 80.5% 2.78 0.48 56 15 7.8% 28.0% 64.2% 2.56 0.63 45 23

m. I will be/am able to influence instructionand school activities

13.3% 38.8% 48.0% 2.35 0.70 58 12 21.2% 38.8% 39.9% 2.19 0.76 60 26

16. To what extent do you agree or disagree with the following statements about your child's school?Percentages

Mean STD edian NDon'tknow

Mis-sing

1 2 3 4 5a.

8.5% 5.0% 21.0% 21.6% 43.9% 3.87 1.27 4 319 35 10

b. Students feel safe at this school 2.9% 5.3% 12.1% 24.2% 55.5% 4.24 1.05 5 339 20 5

c. This school has sufficient financial resources 10.7% 13.7% 27.0% 22.3% 26.2% 3.39 1.30 3 233 121 10

d. I am satisfied with the school's curriculum 4.0% 9.1% 18.2% 25.3% 43.5% 3.95 1.16 4 352 7 5

e. I am satisfied with the instruction offered 4.0% 8.2% 14.8% 28.4% 44.6% 4.01 1.13 4 352 5 7

f. This school has been well received by the community 2.4% 7.6% 20.5% 21.5% 47.9% 4.05 1.10 4 288 74 2

g. I think this school has a bright future 4.4% 2.6% 9.3% 21.5% 62.2% 4.35 1.05 5 344 13 7

h. Too many changes are occurring at the school 33.0% 23.3% 22.3% 10.1% 11.3% 2.43 1.34 2 318 41 5

I. This school reflects a community atmosphere 4.7% 5.6% 20.2% 28.6% 41.0% 3.96 1.12 4 322 35 7

j. 9.3% 5.3% 16.9% 27.5% 41.1% 3.86 1.27 4 302 56 6

k. This school has high standards and expectation for students 3.7% 4.0% 15.1% 20.6% 56.6% 4.22 1.08 5 350 6 8

l. This school has small class sizes 7.6% 5.9% 14.4% 23.8% 48.4% 3.99 1.25 4 341 16 7

m. This school has good physical facilities 14.7% 15.6% 25.2% 19.0% 25.5% 3.25 1.38 3 326 31 7

n. This school has good administrative leadership 5.5% 6.4% 15.7% 25.0% 47.4% 4.02 1.18 4 344 12 8

o.4.5% 5.4% 18.4% 27.7% 44.0% 4.01 1.12 4 332 24 8

Teachers and school leadership are accountable for studentachievement/performance

Extracurricular activities are not emphasized at the expense ofacademics

Initial Expectation Current Experience

Stronglydisagree

Stronglyagree .

This school is meeting students' needs that could not beaddressed at other local schools

Appendix_B6_CSS_Parents_99-00 3 WMU Parent Charter School Survey

School Climate SurveyDescriptive statistics

Informant Group:

What role do you have in the school?Teacher School

adminis-trator

Total Missing

N 159 125 26 310 0

% 51.3 40.3 8.4 100.0

What is your gender?Female Male Total Missing

N 113 60 173 137

% 65.3 34.7 100.0

What is your race/ethnicity?Native

AmericanAsian

AmericanBlack Hispanic White Other Total Missing

N 1 3 39 9 113 4 165 145

% 0.6 1.8 23.6 5.5 68.5 2.4 100.0

Choose the answer that you think most people in your school and community would pick.

Subscale

Mean STD StandardScore

STD Mean StandardScore

1 Teacher-Student Relationships 52.54 5.87 57 47.7 5.9 50

2 Security and Maintenance 28.38 5.01 49 28.4 4.5 50

3 Administration (Principal, Assist. Principal, etc.) 24.26 4.81 54 22.8 4.2 50

4 Student Academic Orientation 14.93 3.16 53 14.1 2.9 50

5 Student Behavioral Values 8.32 2.80 46 9.0 2.3 50

6 Guidance 16.86 2.72 53 16.1 2.4 50

7 Student-Peer Relationships 15.01 2.69 51 14.8 2.4 50

8 Parent and Community-School Relationships 11.81 3.74 45 13.2 3.2 50

9 Instructional Management 27.45 4.36 50 27.4 4.1 50

10 Student Activities 14.66 3.91 43 16.2 2.6 50

Totals for All Pennsylvania Charter School Teachers and Staff, 1999/2000

Teachers/Staff (N=310)

Charter School Data National Norm Data

School staff other thanteacher or administrator

Role in School

0

20

40

60

80

100

Teacher School staff other thanteacher or

administrator

Schooladministrator

Per

cent

Distribution by Race/Ethnicity

0

20

40

60

80

100

NativeAmerican

AsianAmerican

Black Hispanic White Other

Per

cent

Gender

0

20

40

60

80

100

Female MaleP

erce

nt

Appendix_B7_Sch_Climate_Tch-Staff_99-00 1 School Climate Survey

Appendix_B7_Sch_Climate_Tch-Staff_99-00 2 School Climate Survey

School Climate Survey: Teachers and Staff Results

30

40

50

60

70

Teacher-Student

Relationships

Security andMaintenance

Administration(Principal,

Assist.Principal, etc.)

StudentAcademicOrientation

StudentBehavioral

Values

Guidance Student-PeerRelationships

Parent andCommunity-

SchoolRelationships

InstructionalManagement

StudentActivities

Sta

ndar

dS

core

s

National norms designated by the 50th percentile

Choose the answer that you think most people in your school and community would pick.

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

N sing

1 2 3 4 5

Subscale 1: TEACHER-STUDENT RELATIONSHIPS

1 Teachers in this school like their students 0.7% 0.7% 4.3% 36.0% 58.4% 4.51 0.68 3 303 4

2 Teachers in this school are on the side of their students 0.3% 1.7% 9.4% 35.6% 53.0% 4.39 0.75 8 298 4

3 Teachers give students the grades they deserve 0.4% 2.6% 8.1% 43.7% 45.2% 4.31 0.76 33 270 7

4 Teachers help students to be friendly and kind to each other 0.0% 0.7% 6.3% 40.9% 52.2% 4.45 0.64 6 301 3

5 Teachers treat each student as an individual 0.0% 2.3% 4.3% 38.5% 54.8% 4.46 0.69 7 299 4

6 Teachers are willing to help students 0.0% 0.3% 1.3% 30.7% 67.7% 4.66 0.52 3 303 4

7 Teachers are patient when a student has trouble learning 0.3% 2.0% 5.1% 45.9% 46.6% 4.36 0.70 9 296 5

8 Teachers make extra efforts to help students 0.0% 0.7% 4.7% 30.8% 63.9% 4.58 0.62 5 299 6

9 Teachers understand and meet the needs of each student 0.7% 3.4% 13.7% 49.3% 32.9% 4.10 0.81 13 292 5

10 Teachers praise students more often than they scold them 1.1% 5.7% 20.9% 40.8% 31.6% 3.96 0.92 23 282 5

11 Teachers are fair to students 0.3% 1.4% 7.1% 46.4% 44.7% 4.34 0.70 11 295 4

12 Teachers explain carefully so that students can get theirwork done

0.4% 0.7% 5.3% 46.3% 47.4% 4.40 0.66 18 285 7

Subscale 2: SECURITY AND MAINTENANCE

13 Students usually feel safe in the school building 0.3% 1.7% 4.1% 31.4% 62.5% 4.54 0.69 8 296 6

14 Teachers and other workers feel safe in the buildingbefore and after school

0.3% 7.5% 6.1% 29.2% 56.9% 4.35 0.92 12 295 3

15 People are not afraid to come to school for meetings andprograms in the evening

1.1% 6.2% 12.1% 30.0% 50.5% 4.23 0.96 31 273 6

16 Classrooms are usually clean and neat 1.0% 7.2% 11.5% 48.7% 31.6% 4.03 0.90 2 304 4

17 The school building is kept clean and neat 2.0% 9.6% 9.9% 45.5% 33.0% 3.98 1.00 3 303 4

18 The school building is kept in good repair 4.3% 11.3% 18.5% 38.7% 27.2% 3.73 1.11 5 302 3

19 The school grounds are neat and attractive 4.9% 12.8% 21.4% 40.8% 20.1% 3.58 1.10 3 304 3

Subscale 3: ADMINISTRATION (Principal, Assistant Principal, etc.)

20 The administrators in this school listen to student ideas 1.4% 5.6% 15.4% 45.1% 32.5% 4.02 0.91 16 286 8

21 The administrators in this school talk often with teachersand parents

2.4% 7.5% 7.5% 41.1% 41.4% 4.12 1.00 14 292 4

22 The administrators in this school set high standards and letteachers, students, and parents know what these standards are

2.4% 9.3% 17.9% 34.8% 35.5% 3.92 1.06 13 290 7

23 Administrators set a good example by working hard themselves 2.0% 4.8% 9.6% 37.2% 46.4% 4.21 0.95 9 293 8

24 The administrators in this school are willing to hear studentcomplaints and opinions

1.4% 5.1% 7.5% 46.6% 39.4% 4.17 0.88 9 292 9

25 Teachers and students help to decide what happens inthis school

4.2% 10.1% 19.9% 37.6% 28.2% 3.76 1.10 17 287 6

Subscale 4: STUDENT ACADEMIC ORIENTATION

26 Students here understand why they are in school 1.4% 7.1% 13.2% 54.1% 24.3% 3.93 0.88 10 296 4

27 In this school, students are interested in learning new things 1.7% 10.2% 19.0% 48.1% 21.0% 3.77 0.95 11 295 4

28 Students in this school have fun but also work hardon their studies

1.7% 11.9% 20.5% 43.9% 22.1% 3.73 0.99 3 303 4

29 Students work hard to complete their school assignments 1.7% 13.8% 30.3% 45.5% 8.8% 3.46 0.90 9 297 4

Appendix_B7_Sch_Climate_Tch-Staff_99-00 3 School Climate Survey

Choose the answer that you think most people in your school and community would pick.

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStronglyagree

Mean STDDon'tknow

N Missing

1 2 3 4 5

Subscale 5: STUDENT BEHAVIORAL VALUES

30 If one student makes fun of someone, others do not join in 10.5% 34.8% 27.2% 19.9% 7.7% 2.79 1.11 18 287 5

31 Students in this school are well-behaved even when the teachersare not watching them

15.2% 36.9% 27.6% 15.5% 4.8% 2.58 1.07 15 290 5

32 Most students would do their work even if the teacherstepped out of the classroom

11.3% 28.8% 24.7% 29.1% 6.2% 2.90 1.13 14 292 4

Subscale 6: GUIDANCE

33 Teachers or counselors encourage students to thinkabout their future

0.4% 2.8% 11.2% 45.6% 40.0% 4.22 0.78 18 285 7

34 Teachers or counselors help students plan for future classesand for future jobs

1.8% 7.0% 25.0% 40.8% 25.4% 3.81 0.96 32 272 6

35 Teachers or counselors help students with personal problems 0.3% 2.4% 8.6% 38.1% 50.5% 4.36 0.77 9 291 10

36 Students in this school can get help and advice from teachersor counselors

0.3% 2.0% 6.8% 35.4% 55.4% 4.44 0.74 10 294 6

Subscale 7: STUDENT-PEER RELATIONSHIPS

37 Students care about each other 0.7% 8.6% 24.5% 54.1% 12.1% 3.68 0.82 14 290 6

38 Students respect each other 1.7% 19.5% 33.1% 39.6% 6.1% 3.29 0.91 12 293 5

39 Students want to be friends with one another 0.3% 4.1% 19.5% 57.2% 18.8% 3.90 0.76 10 292 8

40 Students have a sense of belonging in this school 0.3% 4.4% 13.9% 47.6% 33.8% 4.10 0.82 9 296 5

Subscale 8: PARENT AND COMMUNITY-SCHOOL RELATIONSHIPS

41 Parents and members of the community attend schoolmeetings and other activities

10.0% 21.1% 20.0% 40.7% 8.2% 3.16 1.15 25 280 5

42 Most people in the community help the school in one way oranother

5.8% 26.0% 30.0% 30.0% 8.3% 3.09 1.06 28 277 5

43 Community attendance at school meetings and programsis good

10.9% 31.9% 27.6% 21.0% 8.6% 2.84 1.13 45 257 8

44 Community groups honor student achievement in learning,music, drama, and sports

18.1% 32.8% 27.9% 17.0% 4.2% 2.56 1.10 40 265 5

Subscale 9: INSTRUCTIONAL MANAGEMENT

45 There is a clear set of rules for students to follow in this school 2.3% 11.0% 10.0% 39.1% 37.5% 3.98 1.06 3 299 8

46 Taking attendance and other tasks do not interfere withclassroom teaching

1.4% 4.7% 11.2% 42.8% 39.9% 4.15 0.90 23 278 9

47 Teachers spend almost all classroom time in learning activities 2.2% 10.5% 13.8% 46.9% 26.5% 3.85 1.00 27 275 8

48 Students in this school usually have assigned schoolwork to do 0.0% 1.4% 8.0% 55.2% 35.3% 4.24 0.66 17 286 7

49 Most classroom time is spent talking about classworkor assignments

2.6% 17.5% 20.9% 38.1% 20.9% 3.57 1.08 35 268 7

50 Teachers use class time to help students learn assigned work 0.4% 3.2% 13.6% 51.6% 31.2% 4.10 0.78 26 279 5

51 Outside interruptions of the classroom are few 6.0% 17.0% 19.9% 39.7% 17.4% 3.45 1.14 22 282 6

Subscale 10: STUDENT ACTIVITIES

52 Students are able to take part in school activities inwhich they are interested

4.6% 10.2% 21.6% 37.5% 26.1% 3.70 1.10 22 283 5

53 Students can be in sports, music, and plays even if theyare not very talented

7.7% 7.7% 23.6% 37.1% 23.9% 3.62 1.16 46 259 5

54 Students are comfortable staying after school for activitiessuch as sports and music

5.5% 9.0% 29.3% 29.7% 26.6% 3.63 1.13 49 256 5

55 Students can take part in sports and other school activitieseven if their families cannot afford it

7.4% 6.3% 24.6% 32.4% 29.3% 3.70 1.17 49 256 5

Appendix_B7_Sch_Climate_Tch-Staff_99-00 4 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 1: TEACHER-STUDENT RELATIONSHIPS

1

Teachers inthis schoollike theirstudents

0.7% 0.7% 4.3% 36.0% 58.4% 4.51 0.68 3 303 4

2

Teachers inthis schoolare on theside of theirstudents

0.3% 1.7% 9.4% 35.6% 53.0% 4.39 0.75 8 298 4

3

Teachersgivestudents thegrades theydeserve

0.4% 2.6% 8.1% 43.7% 45.2% 4.31 0.76 33 270 7

4

Teachershelpstudents tobe friendlyand kind toeach other

0.0% 0.7% 6.3% 40.9% 52.2% 4.45 0.64 6 301 3

5

Teacherstreat eachstudent asan individual

0.0% 2.3% 4.3% 38.5% 54.8% 4.46 0.69 7 299 4

Teachers in this school like their students

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers give students the grades they deserve

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers help students to be friendly and kind

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers treat each student as an individual

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers are on the side of their students

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B7_Sch_Climate_Tch-Staff_99-00 5 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 1: TEACHER-STUDENT RELATIONSHIPS

6

Teachersare willing tohelpstudents

0.0% 0.3% 1.3% 30.7% 67.7% 4.66 0.52 3 303 4

7

Teachersare patientwhen astudent hastroublelearning

0.3% 2.0% 5.1% 45.9% 46.6% 4.36 0.70 9 296 5

8

Teachersmake extraefforts tohelpstudents

0.0% 0.7% 4.7% 30.8% 63.9% 4.58 0.62 5 299 6

9

Teachersunderstandand meetthe needs ofeach student

0.7% 3.4% 13.7% 49.3% 32.9% 4.10 0.81 13 292 5

10

Teacherspraisestudentsmore oftenthan theyscold them

1.1% 5.7% 20.9% 40.8% 31.6% 3.96 0.92 23 282 5

Teachers are willing to help students

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers make extra efforts to help students

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers praise more than they scold students

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers are patient when a student has troublelearning

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers understand & meet needs of eachstudent

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B7_Sch_Climate_Tch-Staff_99-00 6 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 1: TEACHER-STUDENT RELATIONSHIPS

11Teachersare fair tostudents

0.3% 1.4% 7.1% 46.4% 44.7% 4.34 0.70 11 295 4

12

Teachersexplaincarefully sothat studentscan get theirwork done

0.4% 0.7% 5.3% 46.3% 47.4% 4.40 0.66 18 285 7

Subscale 2: SECURITY AND MAINTENANCE

13

Studentsusually feelsafe in theschoolbuilding

0.3% 1.7% 4.1% 31.4% 62.5% 4.54 0.69 8 296 6

14

Teachersand otherworkers feelsafe in thebuildingbefore andafter school

0.3% 7.5% 6.1% 29.2% 56.9% 4.35 0.92 12 295 3

15

People arenot afraid tocome toschool formeetingsandprograms inthe evening

1.1% 6.2% 12.1% 30.0% 50.5% 4.23 0.96 31 273 6

Teachers are fair to students

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

People are not afraid to come to school for eveningmeetings/programs

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers explain carefully so students can get workdone

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers and other workers feel safe in the buildingbefore and after school

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students usually feel safe in the school building

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B7_Sch_Climate_Tch-Staff_99-00 7 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 2: SECURITY AND MAINTENANCE

16

Class-rooms areusually cleanand neat

1.0% 7.2% 11.5% 48.7% 31.6% 4.03 0.90 2 304 4

17

The schoolbuilding iskept cleanand neat

2.0% 9.6% 9.9% 45.5% 33.0% 3.98 1.00 3 303 4

18

The schoolbuilding iskept in goodrepair

4.3% 11.3% 18.5% 38.7% 27.2% 3.73 1.11 5 302 3

19

The schoolgrounds areneat andattractive

4.9% 12.8% 21.4% 40.8% 20.1% 3.58 1.10 3 304 3

Subscale 3: ADMINISTRATION (Principal, Assistant Principal, etc.)

20

The adminis-trators in thisschool listento studentideas

1.4% 5.6% 15.4% 45.1% 32.5% 4.02 0.91 16 286 8

Classrooms are usually clean and neat

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

The school building is kept in good repair

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

The administrators listen to student ideas

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

The school building is kept clean and neat

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

The school grounds are neat and attractive

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B7_Sch_Climate_Tch-Staff_99-00 8 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 3: ADMINISTRATION (Principal, Assistant Principal, etc.)

21

The admin-istrators inthis schooltalk oftenwith teachersand parents

2.4% 7.5% 7.5% 41.1% 41.4% 4.12 1.00 14 292 4

22

Administra-tors set highstandardsand letteachers,students,and parentsknow whatthesestandardsare

2.4% 9.3% 17.9% 34.8% 35.5% 3.92 1.06 13 290 7

23

Administra-tors set agoodexample byworking hardthemselves

2.0% 4.8% 9.6% 37.2% 46.4% 4.21 0.95 9 293 8

24

The admin-istrators inthis schoolare willing tohear studentcomplaintsand opinions

1.4% 5.1% 7.5% 46.6% 39.4% 4.17 0.88 9 292 9

25

Teachersand studentshelp todecide whathappens inthis school

4.2% 10.1% 19.9% 37.6% 28.2% 3.76 1.10 17 287 6

Administrators talk often with teachersand parents

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Administrators set a good example by working hardthemselves

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers and students help to decide whathappens in this school

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Administrators set high standards and let teachers,students & parents know what these standards are

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

The administrators are willing to hear studentcomplaints and opinions

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B7_Sch_Climate_Tch-Staff_99-00 9 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 4: STUDENT ACADEMIC ORIENTATION

26

Studentshereunderstandwhy they arein school

1.4% 7.1% 13.2% 54.1% 24.3% 3.93 0.88 10 296 4

27

In thisschool,students areinterested inlearning newthings

1.7% 10.2% 19.0% 48.1% 21.0% 3.77 0.95 11 295 4

28

Students inthis schoolhave fun butalso workhardon theirstudies

1.7% 11.9% 20.5% 43.9% 22.1% 3.73 0.99 3 303 4

29

Studentswork hard tocompletetheir schoolassignments

1.7% 13.8% 30.3% 45.5% 8.8% 3.46 0.90 9 297 4

Subscale 5: STUDENT BEHAVIORAL VALUES

30

If one stud-ent makesfun ofsomeone,others donot join in

10.5% 34.8% 27.2% 19.9% 7.7% 2.79 1.11 18 287 5

Students here understand why they are in school

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students have fun but also work hard on theirstudies

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

If one student makes fun of someone, others do notjoin in

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students are interested in learning new things

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students work hard to complete their schoolassignments

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B7_Sch_Climate_Tch-Staff_99-00 10 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 5: STUDENT BEHAVIORAL VALUES

31

Students inthis schoolare well-behavedeven whenthe teachersare notwatchingthem

15.2% 36.9% 27.6% 15.5% 4.8% 2.58 1.07 15 290 5

32

Moststudentswould dotheir workeven if theteacherstepped outof theclassroom

11.3% 28.8% 24.7% 29.1% 6.2% 2.90 1.13 14 292 4

Subscale 6: GUIDANCE

33

Teachers orcounselorsencouragestudents tothink abouttheir future

0.4% 2.8% 11.2% 45.6% 40.0% 4.22 0.78 18 285 7

34

Teachers orcounselorshelpstudentsplan forfutureclasses andfor futurejobs

1.8% 7.0% 25.0% 40.8% 25.4% 3.81 0.96 32 272 6

35

Teachers orcounselorshelpstudentswithpersonalproblems

0.3% 2.4% 8.6% 38.1% 50.5% 4.36 0.77 9 291 10

Students in this school are well-behaved even whenthe teachers are not watching them

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers or counselors encourage students to thinkabout their future

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers or counselors help students with personalproblems

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Most students would do their work even if theteacher stepped out of the classroom

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers or counselors help students plan forfuture classes and for future jobs

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B7_Sch_Climate_Tch-Staff_99-00 11 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 6: GUIDANCE

36

Students inthis schoolcan get helpand advicefromteachers orcounselors

0.3% 2.0% 6.8% 35.4% 55.4% 4.44 0.74 10 294 6

Subscale 7: STUDENT-PEER RELATIONSHIPS

37Studentscare abouteach other

0.7% 8.6% 24.5% 54.1% 12.1% 3.68 0.82 14 290 6

38Studentsrespecteach other

1.7% 19.5% 33.1% 39.6% 6.1% 3.29 0.91 12 293 5

39

Studentswant to befriends withone another

0.3% 4.1% 19.5% 57.2% 18.8% 3.90 0.76 10 292 8

40

Studentshave asense ofbelonging inthis school

0.3% 4.4% 13.9% 47.6% 33.8% 4.10 0.82 9 296 5

Students in this school can get help and advice fromteachers or counselors

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students respect each other

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students have a sense of belonging in this school

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students care about each other

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students want to be friends with one another

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B7_Sch_Climate_Tch-Staff_99-00 12 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 8: PARENT AND COMMUNITY-SCHOOL RELATIONSHIPS

41

Parents andmembers ofthecommunityattendschoolmeetingsand otheractivities

10.0% 21.1% 20.0% 40.7% 8.2% 3.16 1.15 25 280 5

42

Most peoplein thecommunityhelp theschool inone way oranother

5.8% 26.0% 30.0% 30.0% 8.3% 3.09 1.06 28 277 5

43

Communityattendanceat schoolmeetingsandprograms isgood

10.9% 31.9% 27.6% 21.0% 8.6% 2.84 1.13 45 257 8

44

Communitygroupshonorstudentachievementin learning,music,drama, andsports

18.1% 32.8% 27.9% 17.0% 4.2% 2.56 1.10 40 265 5

Subscale 9: INSTRUCTIONAL MANAGEMENT

45

There is aclear set ofrules forstudents tofollow in thisschool

2.3% 11.0% 10.0% 39.1% 37.5% 3.98 1.06 3 299 8

Parents and members of the community attendschool meetings and other activities

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Community attendance at school meetings andprograms is good

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

There is a clear set of rules for students to follow inthis school

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Most people in the community help the school inone way or another

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Community groups honor student achievement inlearning, music, drama, and sports

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B7_Sch_Climate_Tch-Staff_99-00 13 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStronglyagree

Mean STDDon'tknow

NMiss-ing

1 2 3 4 5

Subscale 9: INSTRUCTIONAL MANAGEMENT

46

Takingattendanceand othertasks do notinterferewithclassroomteaching

1.4% 4.7% 11.2% 42.8% 39.9% 4.15 0.90 23 278 9

47

Teachersspendalmost allclassroomtime inlearningactivities

2.2% 10.5% 13.8% 46.9% 26.5% 3.85 1.00 27 275 8

48

Students inthis schoolusually haveassignedschoolworkto do

0.0% 1.4% 8.0% 55.2% 35.3% 4.24 0.66 17 286 7

49

Mostclassroomtime is spenttalkingaboutclasswork orassignments

2.6% 17.5% 20.9% 38.1% 20.9% 3.57 1.08 35 268 7

50

Teachersuse classtime to helpstudentslearnassignedwork

0.4% 3.2% 13.6% 51.6% 31.2% 4.10 0.78 26 279 5

Taking attendance and other tasks do not interfere withclassroom teaching

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students in this school usually have assignedschoolwork to do

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers use class time to help students learnassigned work

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers spend almost all classroom time in learningactivities

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Most classroom time is spent talking about classworkor assignments

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B7_Sch_Climate_Tch-Staff_99-00 14 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 9: INSTRUCTIONAL MANAGEMENT

51

Outsideinterruptionsof theclassroomare few

6.0% 17.0% 19.9% 39.7% 17.4% 3.45 1.14 22 282 6

Subscale 10: STUDENT ACTIVITIES

52

Students areable to takepart inschoolactivities inwhich theyareinterested

4.6% 10.2% 21.6% 37.5% 26.1% 3.70 1.10 22 283 5

53

Studentscan be insports,music, andplays even ifthey are notvery talented

7.7% 7.7% 23.6% 37.1% 23.9% 3.62 1.16 46 259 5

54

Students arecomfortablestaying afterschool foractivitiessuch assports andmusic

5.5% 9.0% 29.3% 29.7% 26.6% 3.63 1.13 49 256 5

55

Studentscan takepart insports andother schoolactivitieseven if theirfamiliescannotafford it

7.4% 6.3% 24.6% 32.4% 29.3% 3.70 1.17 49 256 5

Outside interruptions of the classroom are few

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students can be in sports, music, and plays even ifthey are not very talented

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students can take part in sports and other schoolactivities even if their families cannot afford it

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students are able to take part in school activities inwhich they are interested

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students are comfortable staying after school foractivities such as sports and music

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B7_Sch_Climate_Tch-Staff_99-00 15 School Climate Survey

School Climate SurveyDescriptive statistics

Informant Group:

Grade level

Grade level5th 6th 7th 8th 9th 10th 11th 12th Total

N 62 28 78 86 128 115 62 61 620

% 10.0 4.5 12.6 13.9 20.6 18.5 10.0 9.8 100.0

What is your gender?Female Male Total Missing

N 339 389 728 27

% 46.6 53.4 100.0

What is your race/ethnicity?Native

AmericanAsian

AmericanBlack Hispanic White Other Total Missing

N 34 32 296 116 204 40 682 73

% 5.0 4.7 43.4 17.0 29.9 5.9 100.0

Choose the answer that you think most people in your school and community would pick.

Subscale

Mean STD StandardScore

Mean STD StandardScore

1 Teacher-Student Relationships 43.63 10.83 55 39.2 7.9 50

2 Security and Maintenance 23.74 7.28 44 26.5 4.9 50

3 Administration (Principal, Assist. Principal, etc.) 21.57 6.03 52 20.3 4.8 50

4 Student Academic Orientation 14.31 3.98 52 13.5 3.1 50

5 Student Behavioral Values 8.13 3.38 52 7.6 2.6 50

6 Guidance 14.78 4.15 49 15.1 3.2 50

7 Student-Peer Relationships 12.64 4.44 46 13.7 3.2 50

8 Parent and Community-School Relationships 13.20 4.41 47 13.9 3.1 50

9 Instructional Management 25.57 6.12 49 26.2 4.4 50

10 Student Activities 14.01 4.89 48 14.7 3.4 50

Charter School Data National Norm Data

Students (N=755)

Totals for All Pennsylvania Charter School Students, 1999/2000

Distribution by Race/Ethnicity

0

20

40

60

80

100

NativeAmerican

AsianAmerican

Black Hispanic White Other

Per

cent

Gender

0

20

40

60

80

100

Female MaleP

erce

nt

Distribution by Grade

0

20

40

60

80

100

5th 6th 7th 8th 9th 10th 11th 12th

Per

cent

Appendix_B8_Sch_Climate_Students_99-00 1 School Climate Survey

Appendix_B8_Sch_Climate_Students_99-00 2 School Climate Survey

School Climate Survey: Student Results

30

40

50

60

70

Teacher-Student

Relationships

Security andMaintenance

Administration(Principal,

Assist.Principal, etc.)

StudentAcademicOrientation

StudentBehavioral

Values

Guidance Student-PeerRelationships

Parent andCommunity-

SchoolRelationships

InstructionalManagement

StudentActivities

Sta

ndar

dS

core

s

National norms designated by the 50th percentile

Choose the answer that you think most people in your school and community would pick.

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

N Missing

1 2 3 4 5

Subscale 1: TEACHER-STUDENT RELATIONSHIPS

1 Teachers in this school like their students 9.4% 5.5% 22.3% 33.4% 29.4% 3.68 1.22 109 640 6

2 Teachers in this school are on the side of their students 12.9% 12.4% 27.2% 26.8% 20.6% 3.30 1.28 136 611 8

3 Teachers give students the grades they deserve 9.2% 6.1% 15.3% 31.3% 38.1% 3.83 1.26 60 687 8

4 Teachers help students to be friendly and kind to each other 10.5% 9.5% 20.1% 32.0% 27.9% 3.57 1.28 43 703 9

5 Teachers treat each student as an individual 13.5% 9.1% 17.9% 30.1% 29.5% 3.53 1.35 73 672 10

6 Teachers are willing to help students 4.8% 2.7% 11.1% 34.3% 47.2% 4.16 1.05 29 712 14

7 Teachers are patient when a student has trouble learning 8.7% 8.0% 18.5% 30.5% 34.3% 3.74 1.25 53 691 11

8 Teachers make extra efforts to help students 7.0% 5.7% 18.2% 32.7% 36.5% 3.86 1.18 59 688 8

9 Teachers understand and meet the needs of each student 8.9% 9.8% 23.1% 29.9% 28.4% 3.59 1.24 78 666 11

10 Teachers praise students more often than they scold them 18.2% 15.4% 28.3% 21.1% 16.9% 3.03 1.33 123 615 17

11 Teachers are fair to students 12.8% 11.1% 22.7% 29.1% 24.4% 3.41 1.31 42 705 8

12 Teachers explain carefully so that students can get theirwork done

6.9% 7.2% 18.5% 33.6% 33.9% 3.81 1.18 47 699 9

Subscale 2: SECURITY AND MAINTENANCE

13 Students usually feel safe in the school building 12.4% 8.0% 18.0% 31.3% 30.3% 3.59 1.32 110 638 7

14 Teachers and other workers feel safe in the buildingbefore and after school

8.0% 6.7% 18.2% 31.3% 35.8% 3.80 1.22 227 511 17

15 People are not afraid to come to school for meetings andprograms in the evening

8.2% 6.7% 17.1% 29.5% 38.6% 3.84 1.24 152 586 17

16 Classrooms are usually clean and neat 16.1% 13.1% 24.0% 25.0% 22.0% 3.24 1.36 39 697 19

17 The school building is kept clean and neat 14.3% 13.9% 21.4% 30.0% 20.4% 3.28 1.32 34 700 21

18 The school building is kept in good repair 16.1% 15.2% 18.6% 27.7% 22.3% 3.25 1.38 60 682 13

19 The school grounds are neat and attractive 18.6% 16.7% 23.9% 22.8% 17.9% 3.05 1.36 69 681 5

Subscale 3: ADMINISTRATION (Principal, Assistant Principal, etc.)

20 The administrators in this school listen to student ideas 15.9% 10.1% 20.4% 29.0% 24.6% 3.36 1.37 87 662 6

21 The administrators in this school talk often with teachersand parents

5.7% 6.4% 21.4% 35.4% 31.1% 3.80 1.12 92 653 10

22 The administrators in this school set high standards and let teachers,students, and parents know what these standards are

5.3% 7.4% 17.7% 33.9% 35.7% 3.87 1.14 100 638 17

23 Administrators set a good example by working hard themselves 8.4% 10.0% 20.1% 33.4% 28.1% 3.63 1.23 89 641 25

24 The administrators in this school are willing to hear studentcomplaints and opinions

15.0% 8.9% 18.9% 27.6% 29.6% 3.48 1.39 66 673 16

25 Teachers and students help to decide what happens inthis school

16.6% 10.4% 16.9% 29.3% 26.7% 3.39 1.41 82 655 18

Subscale 4: STUDENT ACADEMIC ORIENTATION

26 Students here understand why they are in school 8.2% 5.6% 16.7% 34.1% 35.4% 3.83 1.21 102 646 7

27 In this school, students are interested in learning new things 9.5% 10.8% 21.2% 31.3% 27.3% 3.56 1.26 89 656 10

28 Students in this school have fun but also work hardon their studies

11.4% 11.7% 18.9% 32.1% 25.9% 3.49 1.30 76 667 12

29 Students work hard to complete their school assignments 10.7% 8.1% 27.0% 32.4% 21.8% 3.46 1.22 88 652 15

Appendix_B8_Sch_Climate_Students_99-00 3 School Climate Survey

Choose the answer that you think most people in your school and community would pick.

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStronglyagree

Mean STDDon'tknow

N issing

1 2 3 4 5

Subscale 5: STUDENT BEHAVIORAL VALUES

30 If one student makes fun of someone, others do not join in 33.0% 19.6% 19.5% 15.6% 12.2% 2.54 1.40 67 678 10

31 Students in this school are well-behaved even when the teachersare not watching them

28.9% 20.8% 24.6% 17.5% 8.2% 2.55 1.29 48 698 9

32 Most students would do their work even if the teacherstepped out of the classroom

22.0% 14.8% 24.1% 25.1% 13.9% 2.94 1.35 44 704 7

Subscale 6: GUIDANCE

33 Teachers or counselors encourage students to thinkabout their future

6.5% 7.8% 16.4% 31.7% 37.7% 3.86 1.19 77 666 12

34 Teachers or counselors help students plan for future classesand for future jobs

8.8% 11.1% 19.9% 29.2% 30.9% 3.62 1.27 84 657 14

35 Teachers or counselors help students with personal problems 12.6% 9.1% 19.3% 30.3% 28.8% 3.54 1.33 110 618 27

36 Students in this school can get help and advice from teachersor counselors

8.9% 8.2% 16.2% 35.0% 31.7% 3.72 1.24 84 649 22

Subscale 7: STUDENT-PEER RELATIONSHIPS

37 Students care about each other 20.4% 12.6% 24.1% 24.1% 18.9% 3.08 1.39 94 652 9

38 Students respect each other 21.7% 15.6% 26.2% 26.6% 10.0% 2.88 1.29 57 688 10

39 Students want to be friends with one another 12.7% 8.6% 26.3% 33.3% 19.2% 3.38 1.25 68 678 9

40 Students have a sense of belonging in this school 16.3% 9.6% 24.2% 29.0% 20.9% 3.29 1.34 121 627 7

Subscale 8: PARENT AND COMMUNITY-SCHOOL RELATIONSHIPS

41 Parents and members of the community attend schoolmeetings and other activities

12.6% 9.3% 24.6% 30.6% 22.8% 3.42 1.28 141 601 13

42 Most people in the community help the school in one way oranother

14.2% 15.1% 23.9% 26.8% 20.0% 3.23 1.32 146 590 19

43 Community attendance at school meetings and programsis good

12.7% 12.3% 24.6% 29.4% 21.0% 3.34 1.29 191 528 36

44 Community groups honor student achievement in learning,music, drama, and sports

20.0% 12.7% 20.6% 25.6% 21.1% 3.15 1.42 166 559 30

Subscale 9: INSTRUCTIONAL MANAGEMENT

45 There is a clear set of rules for students to follow in this school 7.5% 6.4% 12.8% 31.4% 41.9% 3.94 1.21 66 671 18

46 Taking attendance and other tasks do not interfere withclassroom teaching

8.4% 8.0% 19.6% 36.3% 27.8% 3.67 1.20 95 634 26

47 Teachers spend almost all classroom time in learning activities 8.1% 9.8% 22.4% 32.9% 26.7% 3.60 1.21 77 651 27

48 Students in this school usually have assigned schoolwork to do 5.2% 6.0% 15.7% 37.6% 35.5% 3.92 1.10 50 668 37

49 Most classroom time is spent talking about classworkor assignments

9.7% 12.4% 25.5% 32.1% 20.4% 3.41 1.22 66 663 26

50 Teachers use class time to help students learn assigned work 7.3% 8.8% 20.4% 37.5% 26.0% 3.66 1.16 67 661 27

51 Outside interruptions of the classroom are few 14.5% 11.8% 21.3% 30.0% 22.4% 3.34 1.33 83 643 29

Subscale 10: STUDENT ACTIVITIES

52 Students are able to take part in school activities inwhich they are interested

16.1% 8.3% 15.9% 26.8% 33.0% 3.52 1.43 79 654 22

53 Students can be in sports, music, and plays even if theyare not very talented

16.6% 9.5% 16.3% 26.9% 30.6% 3.45 1.43 86 650 19

54 Students are comfortable staying after school for activitiessuch as sports and music

15.9% 7.8% 18.0% 26.5% 31.7% 3.50 1.41 135 599 21

55 Students can take part in sports and other school activitieseven if their families cannot afford it

18.4% 6.7% 18.7% 24.9% 31.3% 3.44 1.45 170 566 19

Appendix_B8_Sch_Climate_Students_99-00 4 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 1: TEACHER-STUDENT RELATIONSHIPS

1

Teachers inthis schoollike theirstudents

9.4% 5.5% 22.3% 33.4% 29.4% 3.68 1.22 109 640 6

2

Teachers inthis schoolare on theside of theirstudents

12.9% 12.4% 27.2% 26.8% 20.6% 3.30 1.28 136 611 8

3

Teachersgivestudents thegrades theydeserve

9.2% 6.1% 15.3% 31.3% 38.1% 3.83 1.26 60 687 8

4

Teachershelpstudents tobe friendlyand kind toeach other

10.5% 9.5% 20.1% 32.0% 27.9% 3.57 1.28 43 703 9

5

Teacherstreat eachstudent asan individual

13.5% 9.1% 17.9% 30.1% 29.5% 3.53 1.35 73 672 10

Teachers in this school like their students

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers give students the grades they deserve

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers help students to be friendly and kind

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers treat each student as an individual

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers are on the side of their students

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B8_Sch_Climate_Students_99-00 5 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 1: TEACHER-STUDENT RELATIONSHIPS

6

Teachersare willing tohelpstudents

4.8% 2.7% 11.1% 34.3% 47.2% 4.16 1.05 29 712 14

7

Teachersare patientwhen astudent hastroublelearning

8.7% 8.0% 18.5% 30.5% 34.3% 3.74 1.25 53 691 11

8

Teachersmake extraefforts tohelpstudents

7.0% 5.7% 18.2% 32.7% 36.5% 3.86 1.18 59 688 8

9

Teachersunderstandand meetthe needs ofeach student

8.9% 9.8% 23.1% 29.9% 28.4% 3.59 1.24 78 666 11

10

Teacherspraisestudentsmore oftenthan theyscold them

18.2% 15.4% 28.3% 21.1% 16.9% 3.03 1.33 123 615 17

Teachers are willing to help students

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers make extra efforts to help students

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers praise more than they scold students

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers are patient when a student has troublelearning

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers understand & meet needs of eachstudent

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B8_Sch_Climate_Students_99-00 6 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 1: TEACHER-STUDENT RELATIONSHIPS

11Teachersare fair tostudents

12.8% 11.1% 22.7% 29.1% 24.4% 3.41 1.31 42 705 8

12

Teachersexplaincarefully sothat studentscan get theirwork done

6.9% 7.2% 18.5% 33.6% 33.9% 3.81 1.18 47 699 9

Subscale 2: SECURITY AND MAINTENANCE

13

Studentsusually feelsafe in theschoolbuilding

12.4% 8.0% 18.0% 31.3% 30.3% 3.59 1.32 110 638 7

14

Teachersand otherworkers feelsafe in thebuildingbefore andafter school

8.0% 6.7% 18.2% 31.3% 35.8% 3.80 1.22 227 511 17

15

People arenot afraid tocome toschool formeetingsandprograms inthe evening

8.2% 6.7% 17.1% 29.5% 38.6% 3.84 1.24 152 586 17

Teachers are fair to students

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

People are not afraid to come to school for eveningmeetings/programs

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers explain carefully so students can get workdone

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers and other workers feel safe in the buildingbefore and after school

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students usually feel safe in the school building

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B8_Sch_Climate_Students_99-00 7 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 2: SECURITY AND MAINTENANCE

16

Class-rooms areusually cleanand neat

16.1% 13.1% 24.0% 25.0% 22.0% 3.24 1.36 39 697 19

17

The schoolbuilding iskept cleanand neat

14.3% 13.9% 21.4% 30.0% 20.4% 3.28 1.32 34 700 21

18

The schoolbuilding iskept in goodrepair

16.1% 15.2% 18.6% 27.7% 22.3% 3.25 1.38 60 682 13

19

The schoolgrounds areneat andattractive

18.6% 16.7% 23.9% 22.8% 17.9% 3.05 1.36 69 681 5

Subscale 3: ADMINISTRATION (Principal, Assistant Principal, etc.)

20

The adminis-trators in thisschool listento studentideas

15.9% 10.1% 20.4% 29.0% 24.6% 3.36 1.37 87 662 6

Classrooms are usually clean and neat

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

The school building is kept in good repair

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

The administrators listen to student ideas

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

The school building is kept clean and neat

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

The school grounds are neat and attractive

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B8_Sch_Climate_Students_99-00 8 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 3: ADMINISTRATION (Principal, Assistant Principal, etc.)

21

The admin-istrators inthis schooltalk oftenwith teachersand parents

5.7% 6.4% 21.4% 35.4% 31.1% 3.80 1.12 92 653 10

22

Administra-tors set highstandardsand letteachers,students,and parentsknow whatthesestandardsare

5.3% 7.4% 17.7% 33.9% 35.7% 3.87 1.14 100 638 17

23

Administra-tors set agoodexample byworking hardthemselves

8.4% 10.0% 20.1% 33.4% 28.1% 3.63 1.23 89 641 25

24

The admin-istrators inthis schoolare willing tohear studentcomplaintsand opinions

15.0% 8.9% 18.9% 27.6% 29.6% 3.48 1.39 66 673 16

25

Teachersand studentshelp todecide whathappens inthis school

16.6% 10.4% 16.9% 29.3% 26.7% 3.39 1.41 82 655 18

Administrators talk often with teachersand parents

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Administrators set a good example by working hardthemselves

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers and students help to decide whathappens in this school

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Administrators set high standards and let teachers,students & parents know what these standards are

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

The administrators are willing to hear studentcomplaints and opinions

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B8_Sch_Climate_Students_99-00 9 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 4: STUDENT ACADEMIC ORIENTATION

26

Studentshereunderstandwhy they arein school

8.2% 5.6% 16.7% 34.1% 35.4% 3.83 1.21 102 646 7

27

In thisschool,students areinterested inlearning newthings

9.5% 10.8% 21.2% 31.3% 27.3% 3.56 1.26 89 656 10

28

Students inthis schoolhave fun butalso workhardon theirstudies

11.4% 11.7% 18.9% 32.1% 25.9% 3.49 1.30 76 667 12

29

Studentswork hard tocompletetheir schoolassignments

10.7% 8.1% 27.0% 32.4% 21.8% 3.46 1.22 88 652 15

Subscale 5: STUDENT BEHAVIORAL VALUES

30

If one stud-ent makesfun ofsomeone,others donot join in

33.0% 19.6% 19.5% 15.6% 12.2% 2.54 1.40 67 678 10

Students here understand why they are in school

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students have fun but also work hard on theirstudies

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

If one student makes fun of someone, others do notjoin in

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students are interested in learning new things

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students work hard to complete their schoolassignments

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B8_Sch_Climate_Students_99-00 10 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 5: STUDENT BEHAVIORAL VALUES

31

Students inthis schoolare well-behavedeven whenthe teachersare notwatchingthem

28.9% 20.8% 24.6% 17.5% 8.2% 2.55 1.29 48 698 9

32

Moststudentswould dotheir workeven if theteacherstepped outof theclassroom

22.0% 14.8% 24.1% 25.1% 13.9% 2.94 1.35 44 704 7

Subscale 6: GUIDANCE

33

Teachers orcounselorsencouragestudents tothink abouttheir future

6.5% 7.8% 16.4% 31.7% 37.7% 3.86 1.19 77 666 12

34

Teachers orcounselorshelpstudentsplan forfutureclasses andfor futurejobs

8.8% 11.1% 19.9% 29.2% 30.9% 3.62 1.27 84 657 14

35

Teachers orcounselorshelpstudentswithpersonalproblems

12.6% 9.1% 19.3% 30.3% 28.8% 3.54 1.33 110 618 27

Students in this school are well-behaved even whenthe teachers are not watching them

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers or counselors encourage students to thinkabout their future

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers or counselors help students with personalproblems

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Most students would do their work even if theteacher stepped out of the classroom

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers or counselors help students plan forfuture classes and for future jobs

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B8_Sch_Climate_Students_99-00 11 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 6: GUIDANCE

36

Students inthis schoolcan get helpand advicefromteachers orcounselors

8.9% 8.2% 16.2% 35.0% 31.7% 3.72 1.24 84 649 22

Subscale 7: STUDENT-PEER RELATIONSHIPS

37Studentscare abouteach other

20.4% 12.6% 24.1% 24.1% 18.9% 3.08 1.39 94 652 9

38Studentsrespecteach other

21.7% 15.6% 26.2% 26.6% 10.0% 2.88 1.29 57 688 10

39

Studentswant to befriends withone another

12.7% 8.6% 26.3% 33.3% 19.2% 3.38 1.25 68 678 9

40

Studentshave asense ofbelonging inthis school

16.3% 9.6% 24.2% 29.0% 20.9% 3.29 1.34 121 627 7

Students in this school can get help and advice fromteachers or counselors

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students respect each other

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students have a sense of belonging in this school

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students care about each other

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students want to be friends with one another

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B8_Sch_Climate_Students_99-00 12 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 8: PARENT AND COMMUNITY-SCHOOL RELATIONSHIPS

41

Parents andmembers ofthecommunityattendschoolmeetingsand otheractivities

12.6% 9.3% 24.6% 30.6% 22.8% 3.42 1.28 141 601 13

42

Most peoplein thecommunityhelp theschool inone way oranother

14.2% 15.1% 23.9% 26.8% 20.0% 3.23 1.32 146 590 19

43

Communityattendanceat schoolmeetingsandprograms isgood

12.7% 12.3% 24.6% 29.4% 21.0% 3.34 1.29 191 528 36

44

Communitygroupshonorstudentachievementin learning,music,drama, andsports

20.0% 12.7% 20.6% 25.6% 21.1% 3.15 1.42 166 559 30

Subscale 9: INSTRUCTIONAL MANAGEMENT

45

There is aclear set ofrules forstudents tofollow in thisschool

7.5% 6.4% 12.8% 31.4% 41.9% 3.94 1.21 66 671 18

Parents and members of the community attendschool meetings and other activities

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Community attendance at school meetings andprograms is good

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

There is a clear set of rules for students to follow inthis school

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Most people in the community help the school inone way or another

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Community groups honor student achievement inlearning, music, drama, and sports

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B8_Sch_Climate_Students_99-00 13 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 9: INSTRUCTIONAL MANAGEMENT

46

Takingattendanceand othertasks do notinterfere withclassroomteaching

8.4% 8.0% 19.6% 36.3% 27.8% 3.67 1.20 95 634 26

47

Teachersspendalmost allclassroomtime inlearningactivities

8.1% 9.8% 22.4% 32.9% 26.7% 3.60 1.21 77 651 27

48

Students inthis schoolusually haveassignedschoolworkto do

5.2% 6.0% 15.7% 37.6% 35.5% 3.92 1.10 50 668 37

49

Mostclassroomtime is spenttalking aboutclasswork orassignments

9.7% 12.4% 25.5% 32.1% 20.4% 3.41 1.22 66 663 26

50

Teachersuse classtime to helpstudentslearnassignedwork

7.3% 8.8% 20.4% 37.5% 26.0% 3.66 1.16 67 661 27

Taking attendance and other tasks do not interferewith classroom teaching

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students in this school usually have assignedschoolwork to do

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers use class time to help students learnassigned work

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Teachers spend almost all classroom time inlearning activities

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Most classroom time is spent talking aboutclasswork or assignments

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B8_Sch_Climate_Students_99-00 14 School Climate Survey

Stronglydisagree

DisagreeNeither

agree nordisagree

AgreeStrongly

agreeMean STD

Don'tknow

NMiss-ing

1 2 3 4 5

Subscale 9: INSTRUCTIONAL MANAGEMENT

51

Outsideinterruptionsof theclassroomare few

14.5% 11.8% 21.3% 30.0% 22.4% 3.34 1.33 83 643 29

Subscale 10: STUDENT ACTIVITIES

52

Students areable to takepart inschoolactivities inwhich theyareinterested

16.1% 8.3% 15.9% 26.8% 33.0% 3.52 1.43 79 654 22

53

Studentscan be insports,music, andplays even ifthey are notvery talented

16.6% 9.5% 16.3% 26.9% 30.6% 3.45 1.43 86 650 19

54

Students arecomfortablestaying afterschool foractivitiessuch assports andmusic

15.9% 7.8% 18.0% 26.5% 31.7% 3.50 1.41 135 599 21

55

Studentscan takepart insports andother schoolactivitieseven if theirfamiliescannotafford it

18.4% 6.7% 18.7% 24.9% 31.3% 3.44 1.45 170 566 19

Outside interruptions of the classroom are few

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students can be in sports, music, and plays even ifthey are not very talented

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students can take part in sports and other schoolactivities even if their families cannot afford it

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students are able to take part in school activities inwhich they are interested

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Students are comfortable staying after school foractivities such as sports and music

0%

20%

40%

60%

80%

100%

Stronglydisagree

Disagree Neitheragree nordisagree

Agree Stronglyagree

Appendix_B8_Sch_Climate_Students_99-00 15 School Climate Survey

Appendix C:1 Charter School Revenue Sources, 1998/99Charter School Total revenue % Local % State % Federal LEAs % Other TotalManchester Academy 1,251,716$ 18% 11% 0% 71% 0% 100%Northside Urban Pathways 1,344,106$ 13% 8% 4% 75% 0% 100%Urban League of Pittsburgh 1,110,305$ 39% 2% 7% 53% 0% 100%Mosaica Academy 3,294,377$ 0% 9% 6% 85% 0% 100%Centre Learning Community 352,762$ 1% 2% 11% 86% 0% 100%Nittany Valley 369,153$ 5% 2% 10% 83% 0% 100%Chester County Family Academy 371,663$ 10% 3% 7% 80% 0% 100%Sylvan Heights Science 610,547$ 2% 3% 7% 88% 0% 100%Archway CS of Chester (Chester Community CS) 1,218,790$ 0% 4% 4% 72% 19% 81%Chester CS 1,997,853$ 0% 3% 16% 81% 0% 100%Creative Educational Concepts

Village CS of Chester-Upland 1,836,537$ 1% 3% 7% 89% 0% 100%GECAC Community CS 1,992,925$ 19% 3% 9% 63% 7% 93%Northeast CS 326,625$ 1% 3% 26% 70% 0% 100%La Academia Partnership CS 602,357$ 2% 3% 16% 79% 0% 100%Keystone Education Center 1,588,566$ 0% 5% 10% 85% 0% 100%SUSQ-CYBER CS 275,856$ 0% 9% 4% 87% 0% 100%Center for Economics and Law CS 1,054,832$ 0% 6% 18% 76% 0% 100%Eugenio Maria De Hosts CS 739,978$ 0% 12% 2% 86% 0% 100%Family CS 746,711$ 35% 4% 11% 51% 0% 100%Imhotep CS 1,422,896$ 1% 3% 15% 82% 0% 100%Multi-cultural Academy CS 809,056$ 2% 11% 5% 82% 0% 100%Philadelphia Comm Acad CS 1,601,662$ 4% 5% 5% 86% 0% 100%Philadelphia Harambee Institute 1,445,592$ 4% 6% 6% 84% 0% 100%The Alliance for Progress CS 1,303,515$ 1% 3% 28% 68% 0% 100%The Laboratory CS 2,180,533$ 0% 10% 8% 82% 0% 100%The Preparatory CS 1,015,184$ 3% 3% 14% 80% 0% 100%West Oak Lane CS 5,124,463$ 0% 1% 13% 56% 29% 71%World Communications CS 2,846,764$ 1% 1% 17% 82% 0% 100%YouthBuild Philadelphia CS 4,132,014$ 57% 2% 24% 17% 0% 100%Ridgeview Academy CS 2,214,724$ 30% 3% 12% 56% 0% 100%

Mean 1,506,069$ 8% 5% 11% 74% 2%Median 1,323,811$ 1% 3% 10% 81% 0%

Standard deviation 1,136,440$ 14% 3% 7% 15% 6%

Appendix C:2 Charter School and Host District Expenditure Patterns Compared, 1998/99

Charter SchoolTotal perpupil instruct.

supp.serv.

non-instr. facil. debt serv other

Total perpupilexpenditure instruct

supp.serv.

non-instr.

Total perpupilexpenditure instruct.

supp.serv.

non-instr.

Manchester Academy 9,164$ 73% 27% 0% 0% 0% 0% 10,059$ 64% 35% 1% (895)$ 9% -8% -1%Northside Urban Pathways 9,852$ 70% 28% 0% 0% 0% 2% 10,059$ 64% 35% 1% (207)$ 6% -7% -1%Urban League of Pittsburgh 12,077$ 70% 30% 0% 0% 0% 0% 10,059$ 64% 35% 1% 2,017$ 6% -5% -1%Mosaica Academy 7,243$ 40% 38% 0% 16% 6% 6% 8,917$ 65% 33% 2% (1,674)$ -25% 5% -2%Centre Learning Community 5,455$ 98% 0% 0% 0% 2% 2% 6,653$ 65% 34% 2% (1,198)$ 33% -34% -2%Nittany Valley 6,961$ 64% 35% 1% 0% 0% 0% 7,418$ 64% 35% 2% (457)$ 0% 1% -1%Chester County Family Academy 6,682$ 82% 17% 1% 0% 0% 0% 7,877$ 66% 32% 2% (1,195)$ 16% -15% -1%Sylvan Heights Science 5,814$ 52% 47% 0% 0% 0% 1% 70% 30% 1% -18% 17% -1%Archway CS of Chester (Chester Comm 4,578$ 39% 41% 0% 18% 3% 3% 8,691$ 66% 33% 1% (4,113)$ -28% 8% -1%Chester CS 13,814$ 62% 38% 0% 0% 0% 0% 8,691$ 66% 33% 1% 5,123$ -5% 5% -1%Creative Educational Concepts 8,691$ 66% 33% 1%Village CS of Chester-Upland 5,526$ 56% 43% 1% 0% 0% 0% 8,691$ 66% 33% 1% (3,165)$ -11% 10% 0%GECAC Community CS 9,605$ 48% 47% 0% 1% 4% 4% 7,175$ 66% 33% 1% 2,429$ -18% 14% -1%Northeast CS 9,310$ 43% 57% 0% 0% 0% 0% 6,768$ 66% 32% 2% 2,542$ -22% 25% -2%La Academia Partnership CS 5,828$ 53% 45% 0% 0% 0% 2% 6,749$ 67% 30% 3% (922)$ -14% 15% -3%Keystone Education Center 7,011$ 63% 36% 0% 0% 0% 1% 6,670$ 66% 31% 3% 341$ -3% 5% -3%SUSQ-CYBER CS 4,582$ 61% 39% 0% 0% 0% 0% 6,472$ 70% 28% 2% (1,890)$ -8% 10% -2%Center for Economics and Law CS 4,265$ 64% 36% 0% 0% 0% 0% 6,922$ 65% 32% 3% (2,656)$ -2% 5% -3%Eugenio Maria De Hosts CS 6,438$ 44% 55% 1% 0% 0% 0% 6,922$ 65% 32% 3% (484)$ -21% 24% -2%Family CS 8,581$ 64% 36% 0% 0% 0% 0% 6,922$ 65% 32% 3% 1,660$ -1% 5% -3%Imhotep CS 7,604$ 40% 60% 0% 0% 0% 0% 6,922$ 65% 32% 3% 682$ -25% 29% -3%Multi-cultural Academy CS 5,047$ 56% 44% 0% 0% 0% 0% 6,922$ 65% 32% 3% (1,875)$ -10% 13% -3%Philadelphia Comm Acad CS 7,062$ 42% 51% 7% 0% 0% 0% 6,922$ 65% 32% 3% 141$ -23% 20% 4%Philadelphia Harambee Institute 7,425$ 57% 41% 0% 2% 0% 0% 6,922$ 65% 32% 3% 503$ -8% 10% -3%The Alliance for Progress CS 6,669$ 100% 0% 0% 0% 0% 0% 6,922$ 65% 32% 3% (253)$ 35% -32% -3%The Laboratory CS 3,668$ 51% 49% 0% 0% 0% 0% 6,922$ 65% 32% 3% (3,253)$ -14% 17% -3%The Preparatory CS 5,240$ 45% 51% 4% 0% 0% 0% 6,922$ 65% 32% 3% (1,682)$ -20% 19% 1%West Oak Lane CS 10,722$ 35% 25% 0% 37% 3% 3% 6,922$ 65% 32% 3% 3,801$ -31% -6% -3%World Communications CS 7,010$ 45% 55% 0% 0% 0% 0% 6,922$ 65% 32% 3% 88$ -21% 24% -3%YouthBuild Philadelphia CS 29,879$ 81% 19% 0% 0% 0% 0% 6,922$ 65% 32% 3% 22,957$ 15% -12% -3%Ridgeview Academy CS 10,117$ 82% 14% 0% 0% 0% 4% 6,099$ 68% 30% 2% 4,018$ 14% -16% -2%means 8,108$ 59% 37% 0% 2% 1% 1% 7,524$ 66% 32% 2% 703$ -6% 5% -2%medians 7,010$ 56% 38% 0% 0% 0% 0% 6,922$ 65% 32% 2% (253)$ -9% 7% -2%standard deviations 4,754$ 17% 15% 1% 8% 1% 2% 1,136$ 1% 2% 1% 4,825$ 17% 16% 2%

Charter Schools Host Districts CS minus HD

Note: For charter schools with more than one host district we generated a composite value by taking the average of all host districts. Per pupil expenditures were calculated using estimates for averagedaily membership. Host district per pupil expenditures were calculated by dividing the current year total expenditures by the total average daily membership for the district. Finally, the difference betweenthe aggregate mean per pupil expenditures and the school-to-school comparisons of the per pupil expenditure are not equal due to missing values.

Appendix C:3 Charter School Variances and End of Year Balances, 1998/99

Charter School Variance ($)% BudgetedRevenues Variance ($)

% BudgetedExpenditures $

% Totalexpenditures

Manchester Academy* 1,251,716$ 1,118,040$ 133,676$ 12%Northside Urban Pathways 341,355$ 34% 209,026$ 21% 132,329$ 11%Urban League of Pittsburgh (87,031)$ -7% (231,215)$ -19% 144,184$ 15%Mosaica Academy (356,306)$ -10% (399,804)$ -11% 27,670$ 1%Centre Learning Community 19,590$ 6% (71,309)$ -21% 90,899$ 35%Nittany Valley 42,889$ 13% 37,949$ 13% 35,039$ 10%Chester County Family Academy 17,163$ 5% (60,482)$ -17% 77,645$ 26%Sylvan Heights Science (3,219)$ -1% (136,717)$ -22% 133,839$ 28%Archway CS of Chester (Chester Comm. CS) 124,170$ 11% 307,620$ 35% 33,062$ 3%Chester CS (246,973)$ -11% (449,820)$ -20% 243,511$ 14%Creative Educational Concepts

Village CS of Chester-Upland (393,310)$ -18% (890,571)$ -38% 383,261$ 26%GECAC Community CS 880,731$ 79% 703,067$ 63% 177,664$ 10%Northeast CS (295,543)$ -48% (286,994)$ -46% (8,549)$ -3%La Academia Partnership CS 10,127$ 2% (67,279)$ -11% 77,860$ 15%Keystone Education Center 246,205$ 18% (24,567)$ -2% 109,302$ 7%SUSQ-CYBER CS 135,585$ 97% 61,342$ 44% 74,243$ 37%Center for Economics and Law CS (263,168)$ -20% (405,136)$ -31% 142,084$ 16%Eugenio Maria De Hosts CS (82,555)$ -10% (120,807)$ -15% 38,252$ 5%Family CS 114,153$ 18% 88,338$ 14% 25,884$ 4%Imhotep CS (28,944)$ -2% 75,667$ 5% (82,608)$ -5%Multi-cultural Academy CS 23,198$ 3% (154,989)$ -20% 178,187$ 28%Philadelphia Comm Acad CS (21,599)$ -1% 36,578$ 2% (192,177)$ -11%Philadelphia Harambee Institute* 1,445,592$ 1,611,233$ (165,641)$ -10%The Alliance for Progress CS* 1,303,515$ 987,018$ 316,497$ 32%The Laboratory CS (183,317)$ -8% (984,580)$ -42% 801,263$ 58%The Preparatory CS (4,840)$ 0% (122,102)$ -14% 234,442$ 30%West Oak Lane CS 1,256,252$ 32% 1,664,510$ 43% (408,258)$ -7%World Communications CS 223,884$ 9% 252,675$ 10% (20,197)$ -1%YouthBuild Philadelphia CS 360,790$ 10% 1,446,411$ 52% (80,864)$ -2%Ridgeview Academy CS (49,298)$ -2% (262,799)$ -12% 282,308$ 15%

Mean 192,694$ 7.4% 131,010$ -1.4% 97,827$ 13.3%Median 18,377$ 1.7% (42,525)$ -11.4% 84,380$ 11.4%

Standard deviation 510,377$ 28.4% 655,597$ 29.1% 206,902$ 16.1%*School budgeted for no revenues or expenditures

Revenues Expenditures Balance

D-1

Appendix D:1Details on Change Scores Analysis

To encourage greater exploration of the PSSA data, we present in this appendix more detailedinformation on the change score comparisons between charter schools and their host districts. TableD:1 provides PSSA scores for both the 1997/98 and schools and each of their host districts. Thetable also includes information on the concentration of low income students and PSSA participationrates in these schools.

Table D:1 Charter School and Host District Two-Year Trends ComparedCharter Schools Host Districts Charter schools minus host

districtsName of School Test 2-year

average %low income(1998/99)

2-yearaverage %

participation

1997/1998

1998/1999

Diff. Name ofDistrict

2-yearaverage %low income(1998/99)

2-yearaverage %

partici-pation

1997/1998

1998/1999

Diff. Changein % lowincome

Change inpercentpartici-pation

Changein PSSA

mean

Keystone EducationCenter(Greenville comparison)

Math 8 68 76 1080 1120 40 Greenville Area 33 92 1310 1380 70 5 -30Reading 8 68 76 1010 1100 90 Greenville Area 33 90 1300 1360 60 5 30Math 11 68 73 1000 1170 170 Greenville Area 33 95 1360 1350 -10 5 180Reading 11 68 73 1070 1220 150 Greenville Area 33 93 1330 1330 0 5 150

Keystone Education Center(Reynolds comparison)

Math 8 68 76 1080 1120 40 Reynolds 34 92 1240 1260 20 5 20Reading 8 68 76 1010 1100 90 Reynolds 34 90 1230 1250 20 5 70Math 11 68 73 1000 1170 170 Reynolds 34 95 1260 1220 -40 5 210Reading 11 68 73 1070 1220 150 Reynolds 34 93 1300 1220 -80 5 230

Philadelphia Community Academy

Math 11 87 88 1000 1040 40 Philadelphia 82 78 1120 1140 20 13 3 20

Reading 11 87 93 1010 1130 120 Philadelphia 82 78 1140 1140 0 13 -1 120

Philadelphia Harambee Institute

Math 5 91 1140 1100 -40 Philadelphia 82 94 1140 1140 0 15 -40Reading 5 88 1030 1150 120 Philadelphia 82 95 1090 1120 30 23 90Writing 6 1170 1070 70 Philadelphia 82 1090 1110 20 50Math 8 100 1000 1000 250 Philadelphia 82 91 1120 1140 20 0 230Reading 8 97 1000 1000 350 Philadelphia 82 92 1120 1130 10 7 340

World Communications

Writing 6 60 1170 1180 10 Philadelphia 82 1090 1110 20 0 -10Math 8 60 98 1110 1160 50 Philadelphia 82 91 1120 1140 20 0 2 30Reading 8 60 99 1150 1160 10 Philadelphia 82 92 1120 1130 10 0 -2 0Writing 9 60 1200 1250 50 Philadelphia 82 1140 1200 60 0 -10

Note: Where one year of participation or income data are missing, we use the one year of available data as an estimate. Missing values in change scores reflect eitherthe fact that no participation rates are reported for writing or that we were missing a value for one of the two years.

D-3

Table D:2 Regression Estimates From Full Model

Predictor Coefficient Standarderror

P>|t|

% low income -2.2 0.05 0.00

% nonwhite -1.7 0.07 0.00

% single parent 2.0 0.17 0.00

% college degree 2.7 0.10 0.00

Intercept 1325 3.65 0.00

R2 0.73

P > |F| 0.00

Table D:3 Regression Estimates from Income-Only Model

Predictor Coefficient Standarderror

P>|t|

% low income -3.15 0.03 0.00

Intercept 1412 1.24 0.00

R2 0.69

P > |F| 0.00

Appendix D:2Details on the Analysis of Regression Residuals

As discussed in Section 11.5, the regression residual analysis proceeds on the assumption that theanalyst has included all relevant demographic factors in the model. Unfortunately, the paucity ofcomparable data sources on both charter and noncharter schools has limited our analysis to the useof income only. The vast literature on student achievement demonstrates that while income is amongthe most powerful predictors of student achievement, it is far from the only one. Hence, a model thatpredicts student achievement based only on student income is likely to produce residuals that arestatistically biased. In this appendix we briefly explore the extent and consequences of this statisticalbias.

We began by estimating a more fullyspecified model using data from allnoncharter schools in Pennsylvania(we lacked comparable data on allbut achievement and income forcharter schools). The full modelincludes, in addition to income, theproportion of nonwhite students, theproportion of single-parent families,and the proportion of adults in thecommunity with at least a collegedegree. Ideally, we would includemany other variables, including theconcentration of special educationstudents, pretest scores on studentachievement tests, and a host of factors describing students’ prior educational experiences. Data forthe additional student background factors came from the 1990 U.S. census and are measured at thedistrict level. Using district level predictors in a model of school-level student achievement is likelyto bias the standard errors (measureof the variability in the coefficients)but should not bias the coefficientsthemselves (from which wecalculate the residuals). In order tosimplify the analysis, we estimatedthe model on data from the1998/1999 5th grade math portion ofthe PSSA only. The results will bedifferent for other years and otherportions of the PSSA. Table D:2

1 The bivariate correlation between education and income is negative (r=-0.38).

D-4

Res

idua

ls fr

om fu

ll m

odel

Residuals from income model-200 0 200 400

-200

0

200

400

Figure D:1 Residuals From Income=Only and Full Model Compared, R=0.94

presents coefficients for both the full model and the income-only model. Table 11:3 providesestimates from the income-only model.

Clearly, the addition of the extra student background variables changes the slope on the incomevariable, reducing it by approximately 1 point. Given that residuals are calculated by comparingpredicted and observed values, omission of these variables will clearly statistically bias the residuals.However, the bias is attenuated by the fact that income is strongly and positively correlated with race(r=0.70) and single parent status (r=0.79).1 Hence, income in the constrained model picks up muchof the unmeasured influence of race and single parent status. Consequently, the ultimate impact onthe residuals is quite modest. Indeed, the correlation between residuals from each model is very high(r=0.94). The strong relationship between the residuals is illustrated by scatterplot in Figure D:1.

D-5

Table D:4 Regression Residuals for 5th Grade MathYear Name of School Percent

Low IncomeMean PSSA Predicted

Mean PSSADifference

1999 Manchester Academy 60 1090 1230 -1401999 Mosaica Academy 19 1240 1360 -1201999 Centre Learning Community 10 1370 1380 -101999 Nittany Valley Charter School 4 1520 1400 1201999 Chester Charter School 67 1000 1210 -2101999 Creative Educational Concepts 1020 1410 1999 Village CS of Chester-Upland 73 1110 1190 -801999 Eugenio Maria de Hosts CS 100 1050 1110 -601998 Philadelphia Harambee Institute 1140 1410 1999 Philadelphia Harambee Institute 1100 1410 1999 The Laboratory Charter School 97 1080 1120 -40

Mean (weighted) 54 1156 1300 -88Median 64 1100 1360 -70St. Dev. 38 160 120 98Note: Predicted PSSA scores rounded to the nearest 10 to conform with scoring practices

Table D:5 Regression Residuals for 5th Grade ReadingYear Name of School Percent

Low IncomeMean PSSA Predicted

Mean PSSADifference

1999 Manchester Academy 60 1140 1230 -901999 Mosaica Academy 19 1220 1360 -1401999 Centre Learning Community 10 1400 1390 101999 Nittany Valley Charter School 4 1460 1410 501999 Chester Charter School 67 1070 1210 -1401999 Creative Educational Concepts 1060 1999 Village CS of Chester-Upland 73 1130 1190 -601999 Eugenio Maria de Hosts CS 100 1070 1100 -301998 Philadelphia Harambee Institute 1030 1999 Philadelphia Harambee Institute 1150 1999 The Laboratory Charter School 97 1140 1110 30

Mean (weighted) 54 1170 1250 -63Median 64 1140 1220 -45St. Dev. 38 140 120 74Note: Predicted PSSA scores rounded to the nearest 10 to conform with scoring practices

D-6

Table D:6 Regression Residuals for 6th Grade WritingYear Name of School Percent

Low IncomeMean PSSA Predicted

Mean PSSADifference

1999 Mosaica Academy 19 1290 1290 01999 Centre Learning Community 10 1250 1310 -601999 Nittany Valley Charter School 4 1030* 1320 -2901999 Creative Educational Concepts 1050 1999 Village CS of Chester-Upland 73 1110 1150 -401999 La Academia Partnership CS 68 1120 1160 -401999 Eugenio Maria de Hosts CS 100 1110 1090 201998 Philadelphia Harambee Institute 1070 1999 Philadelphia Harambee Institute 1140 1998 World Communications CS 60 1170 1180 -101999 World Communications CS 60 1180 1180 0

Mean (weighted) 49 1135 1220 -20Median 60 1120 1180 -25St. Dev. 34 85 90 110Note: Predicted PSSA scores rounded to the nearest 10 to conform with scoring practices*Reported as a 1000 on public PDE raw data files. Correspondence with PDE staff indicates that the reported numberis in error. Also, the scaled score incorporates 33 percent of the examinations that were graded “unscorable,” whichcount as zeros in generating school means.

Table D:7 Regression Residuals for 8th Grade MathYear Name of School Percent

Low IncomeMean PSSA Predicted

Mean PSSADifference

1999 Nittany Valley Charter School 4 1580 1380 2001999 Village CS of Chester-Upland 73 1000 1170 -1701999 La Academia Partnership CS 68 1060 1180 -1201998 Keystone Education Center 65 1080 1190 -1101999 Keystone Education Center 70 1120 1180 -601999 Philadelphia Comm Acad CS 93 1050 1100 -501998 Philadelphia Harambee Institute 1000 1999 Philadelphia Harambee Institute 1250 1998 World Communications CS 60 1110 1190 -801999 World Communications CS 60 1160 1210 -501999 Ridgeview Academy CS 88 1130 1120 10

Mean (weighted) 65 1140 1190 -69Median 68 1110 1180 -60St. Dev. 25 163 80 106Note: Predicted PSSA scores rounded to the nearest 10 to conform with scoring practices

D-7

Table D:8 Regression Residuals for 8th Grade ReadingYear Name of School Percent

Low IncomeMean PSSA Predicted

Mean PSSADifference

1999 Nittany Valley Charter School 4 1450 1380 701999 Village CS of Chester-Upland 73 1080 1170 -901999 La Academia Partnership CS 68 1030 1180 -1501998 Keystone Education Center 65 1010 1190 -1801999 Keystone Education Center 70 1100 1180 -801999 Philadelphia Comm Acad CS 93 1050 1110 -601998 Philadelphia Harambee Institute 1000 1999 Philadelphia Harambee Institute 1350 1998 World Communications CS 60 1150 1200 -501999 World Communications CS 60 1160 1210 -501999 Ridgeview Academy CS 88 1140 1120 20

Mean (weighted) 65 1138 1193 -69Median 68 1100 1180 -60St. Dev. 25 142 78 77Note: Predicted PSSA scores rounded to the nearest 10 to conform with scoring practices

Table D:9 Regression Residuals for 9th Grade WritingYear Name of School Percent

Low IncomeMean PSSA Predicted

Mean PSSADifference

1999 Village CS of Chester-Upland 73 1200 1210 -10

1999 SUSQ-CYBER CS 0 1200 1410 -210

1999 Center for Economics & Law CS 70 1230 1210 20

1999 Imhotep Charter School 75 1230 1200 30

1999 Multi-cultural Academy CS 89 1330 1160 170

1999 The Preparatory CS 61 1330 1240 90

1998 World Communications CS 60 1200 1230 -30

1999 World Communications CS 60 1250 1240 10

Mean (weighted) 61 1246 1240 18

Median 66 1230 1220 15Note: Predicted PSSA scores rounded to the nearest 10 to conform with scoring practices

D-8

Table D:10 Regression Residuals for 11th Grade MathYear Name of School Percent

Low IncomeMean PSSA Predicted

Mean PSSADifference

1999 Northeast Charter School 30 1140 1260 -120

1998 Keystone Education Center 65 1000 1130 -130

1999 Keystone Education Center 70 1170 1110 60

1999 SUSQ-CYBER CS 0 1140 1370 -230

1998 Philadelphia Comm Acad CS 80 1000 1050 -50

1999 Philadelphia Comm Acad CS 93 1040 1030 10

1999 Ridgeview Academy CS 88 1230 1050 180

Mean (weighted) 61 1103 1140 -4

Median 70 1140 1110 -50

St. Dev. 34 90 130 136Note: Predicted PSSA scores rounded to the nearest 10 to conform with scoring practices

Table D:11 Regression Residuals for 11th Grade ReadingYear Name of School % low

incomeMean PSSA Predicted

Mean PSSADifference

1999 Northeast Charter School 30 1110 1260 -150

1998 Keystone Education Center 65 1070 1140 -70

1999 Keystone Education Center 70 1220 1120 100

1999 SUSQ-CYBER CS 0 1220 1360 -140

1998 Philadelphia Comm Acad CS 80 1010 1110 -100

1999 Philadelphia Comm Acad CS 93 1130 1050 80

1999 Ridgeview Academy CS 88 1290 1060 230

Mean (weighted) 61 1150 1160 28

Median 70 1130 1120 -70

St. Dev. 34 98 110 145Note: Predicted PSSA scores rounded to the nearest 10 to conform with scoring practices

D-9

Appendix D:3Host District Comparisons

Tables D:12 through D:19 provide detailed information on the host district comparisons. Since sixcharter schools had been open two years at the time these data were collected, we have included acolumn of “year.” We have also included separate comparisons for each host district where chartersare sponsored by more than one (regional charters).

The most important information in the tables is found in the gray area toward the middle of thetables. The righthand column in this area shows the difference between the charter school’s and hostdistrict’s score on the relevant portion and year of the PSSA. The numbers are derived bysubtracting the host district’s score from that of the charter school. Hence, positive numbers indicatethat the charter school outperformed the host district. Negative numbers, by contrast, indicate thatthe host district outperformed the charter school. To facilitate interpretation of the table, we haveprinted positive values in green text and negative values in red text. The mean value at the bottomis the average of all differences weighted by charter schools’ average daily membership (ADM).

In addition to this information, the tables also provide information that will facilitate carefulinterpretation of the difference scores. The columns to the right of the gray area include detailedbreakdowns of PSSA scores by quartiles. Quartiles, as the name suggests, divide a distribution ofdata points into four groups, each containing 25 percent of the data points. These, in combinationwith the mean PSSA scores, provide readers with a fairly detailed understanding of both the centraltendency and dispersion of scores within a school. The tables also provide information onparticipation rates and concentration of low income students. Countless studies have demonstratedthat both factors exert an independent influence on student achievement scores. Readers areencouraged to consider both factors when assessing differences between particular charter schoolsand their host districts. To facilitate such comparisons, the left column in the gray area provides thedifference between charter schools and their host districts in the concentration of low incomestudents. Space constraints prevented us from including such comparisons for participation rates.

Table D:12 Comparisons Between Charter Schools and Their Host Districts: PSSA 5th Grade MathDetailed Score Breakdown

Charter Schools Host Districts Charter MinusHost

Charter School Quartile Host District Quartile

Year Name of School % LowIncome

%Partic

MeanPSSA

Name of District % LowIncome

MeanPSSA

% LowIncome

MeanPSSA

Top Hi-mid Lo-mid Bottom Top Hi-mid Lo-mid Bottom

1999 Manchester Academy 60 1090 Pittsburgh 76 1250 -16 -160 0 0 21 79 15 19 33 33

1999 Mosaica Academy 19 100 1240 Bensalem Township 24 1330 -5 -90 12 16 42 30 25 28 33 14

1999 Centre LearningCommunity

10 89 1370 Bald Eagle Area 43 1260 -33 110 38 29 8 25 13 24 37 26

1999 Centre LearningCommunity

10 89 1370 Bellefonte Area 15 1380 -5 -10 38 29 8 25 37 28 23 12

1999 Centre LearningCommunity

10 89 1370 State College Area 9 1400 1 -30 38 29 8 25 36 30 25 9

1999 Nittany Valley 4 100 1520 State College Area 9 1400 -5 120 50 33 0 17 36 30 25 9

1999 Chester CS 37 1000 Chester-Upland 63 1110 -26 -110 0 3 7 90 1 7 18 75

1999 Creative EducationalConcepts

80 1020 Chester-Upland 63 1110 -90 0 0 0 100 1 7 18 75

1999 Village CS of Chester-Upland

73 86 1110 Chester-Upland 63 1110 10 0 0 0 28 72 1 7 18 75

1999 Eugenio Maria DeHosts CS

100 93 1050 Philadelphia 82 1140 18 -90 0 0 19 81 3 9 24 64

1998 PhiladelphiaHarambee Institute

83 1140 Philadelphia 82 1140 0 0 11 17 72 4 9 22 66

1999 PhiladelphiaHarambee Institute

98 1100 Philadelphia 82 1140 -40 0 3 21 77 3 9 24 64

1999 The Laboratory CS 97 90 1080 Philadelphia 82 1140 15 -60 0 11 8 81 3 9 24 64

Average Difference (Weighted by Av. Daily Membership) -61

Table D:13 Comparisons Between Charter Schools and Their Host Districts: PSSA 5th Grade ReadingDetailed Score Breakdown

Charter Schools Host Districts Charter MinusHost

Charter School Quartile Host District Quartile

Year Name of School % LowIncome

%Partic

MeanPSSA Name of District

% LowIncome

MeanPSSA

% LowIncome

MeanPSSA

Top Hi-mid Lo-mid Bottom Top Hi-mid Lo-mid Bottom

1999 Manchester Academy 60 1140 Pittsburgh 76.3 1230 -16 -90 7 29 64 14 19 31 36

1999 Mosaica Academy 19 100 1220 Bensalem Township 24.4 1320 -5 -100 8 28 24 40 25 30 28 17

1999 Centre LearningCommunity

10 85 1400 Bald Eagle Area 43 1310 -33 90 44 30 9 17 26 29 22 23

1999 Centre LearningCommunity

10 85 1400 Bellefonte Area 14.7 1340 -5 60 44 30 9 17 31 26 28 15

1999 Centre LearningCommunity

10 85 1400 State College Area 8.7 1390 1 10 44 30 9 17 42 28 20 10

1999 Nittany Valley 4 100 1460 State College Area 8.7 1390 -5 70 67 17 17 42 28 20 10

1999 Chester CS 67 82 1070 Chester-Upland 63 1090 4 -20 7 4 14 75 4 6 22 69

1999 Creative EducationalConcepts

90 1060 Chester-Upland 63 1090 -30 11 89 4 6 22 69

1999 Village CS of Chester-Upland

73 86 1130 Chester-Upland 63 1090 10 40 6 39 56 4 6 22 69

1999 Eugenio Maria DeHosts CS

100 93 1070 Philadelphia 81.6 1120 18 -50 3 27 70 4 10 22 64

1998 PhiladelphiaHarambee Institute

75 1030 Philadelphia 81.5 1090 -60 5 5 20 70 5 10 21 65

1999 PhiladelphiaHarambee Institute

98 1150 Philadelphia 81.6 1120 30 5 5 39 51 4 10 22 64

1999 The Laboratory CS 97 93 1140 Philadelphia 81.6 1120 15 20 5 13 24 58 4 10 22 64

Average Difference (Weighted by Av. Daily Membership) -25

Table D:14 Comparisons Between Charter Schools and Their Host Districts: PSSA 8th Grade MathDetailed Score Breakdown

Charter Schools Host Districts Charter MinusHost

Charter School Quartile Host District Quartile

Year Name of School % LowIncome

%Partic

MeanPSSA

Name of District % LowIncome

MeanPSSA

% LowIncome

MeanPSSA

Top Hi-mid Lo-mid Bottom Top Hi-mid Lo-mid Bottom

1999 Nittany Valley 4 100 1580 State College Area 9 1420 -5 160 50 0 0 50 42 31 19 81999 Village CS of Chester-

Upland73 47 1000 Chester-Upland 63 1120 10 -120 0 14 14 71 2 9 25 64

1999 La AcademiaPartnership CS

68 52 1060 Lancaster 59 1170 9 -110 0 0 25 75 10 12 23 55

1998 Keystone EducationCenter

65 76 1080 Greenville Area 38 1310 27 -230 3 0 31 66 23 25 29 23

1998 Keystone EducationCenter

65 76 1080 Reynolds 38 1240 27 -160 3 0 31 66 9 24 39 28

1999 Keystone EducationCenter

70 1120 Greenville Area 27 1380 43 -260 4 9 22 65 34 34 23 9

1999 Keystone EducationCenter

70 1120 Reynolds 39 1260 31 -140 4 9 22 65 10 32 31 28

1999 Philadelphia CommAcad CS

93 100 1050 Philadelphia 82 1140 12 -90 0 0 17 83 5 11 25 59

1998 PhiladelphiaHarambee Institute

100 1000 Philadelphia 82 1120 -120 0 0 21 79 5 10 24 61

1999 PhiladelphiaHarambee Institute

100 1250 Philadelphia 82 1140 110 0 31 54 15 5 11 25 59

1998 WorldCommunications CS

60 97 1110 Philadelphia 82 1120 -22 -10 0 9 36 55 5 10 24 61

1999 WorldCommunications CS

60 99 1160 Philadelphia 82 1140 -22 20 1 8 40 51 5 11 25 59

1999 Ridgeview AcademyCS

88 1130 Greater Latrobe 26 1370 62 -240 4 11 21 64 31 33 25 11

Table D:15 Comparisons Between Charter Schools and Their Host Districts: PSSA 8th Grade ReadingDetailed Score Breakdown

Charter Schools Host Districts Charter MinusHost

Charter School Quartile Host District Quartile

Year Name of School % LowIncome

%Partic

MeanPSSA

Name of District % LowIncome

MeanPSSA

% LowIncome

MeanPSSA

Top Hi-mid Lo-mid Bottom Top Hi-mid Lo-mid Bottom

1999 Nittany Valley 4 100 1450 State College Area 9 1400 -5 50 50 0 0 50 43 28 18 111999 Village CS of Chester-

Upland73 53 1080 Chester-Upland 63 1110 10 -30 0 0 38 63 4 11 24 62

1999 La AcademiaPartnership CS

68 57 1030 Lancaster 59 1170 9 -140 0 8 23 69 11 15 24 50

1998 Keystone EducationCenter

65 76 1010 Greenville Area 38 1300 27 -290 0 0 21 79 24 28 25 24

1998 Keystone EducationCenter

65 76 1010 Reynolds 38 1230 27 -220 0 0 21 79 14 23 27 36

1999 Keystone EducationCenter

70 1100 Greenville Area 27 1360 43 -260 4 13 9 74 30 31 28 11

1999 Keystone EducationCenter

70 1100 Reynolds 39 1250 31 -150 4 13 9 74 17 25 23 36

1999 Philadelphia CommAcad CS

93 87 1050 Philadelphia 82 1130 11 -80 0 5 15 80 7 12 24 57

1998 PhiladelphiaHarambee Institute

93 1000 Philadelphia 82 1120 -120 0 0 13 87 7 13 25 55

1999 PhiladelphiaHarambee Institute

100 1350 Philadelphia 82 1130 220 15 31 46 8 7 12 24 57

1998 WorldCommunications CS

60 100 1150 Philadelphia 82 1120 -22 30 5 16 32 46 7 13 25 55

1999 WorldCommunications CS

60 98 1160 Philadelphia 82 1130 -22 30 4 17 26 54 7 12 24 57

1999 Ridgeview AcademyCS

88 1140 Greater Latrobe 26 1370 62 -230 4 7 32 57 34 33 22 12

Table D:16 Comparisons Between Charter Schools and Their Host Districts: PSSA 11th Grade MathDetailed Score Breakdown

Charter Schools Host Districts Charter MinusHost

Charter School Quartile Host District Quartile

Year Name of School % LowIncome

%Partic

MeanPSSA

Name of District % LowIncome

MeanPSSA

% LowIncome

MeanPSSA

Top Hi-mid Lo-mid Bottom Top Hi-mid Lo-mid Bottom

1999 Northeast CS 30 1140 Abington Heights 6 1340 24 -200 0 50 50 0 24 32 24 201999 Northeast CS 30 1140 Carbondale Area 53 1220 -23 -80 0 50 50 0 9 23 35 341999 Northeast CS 30 1140 Dunmore 26 1280 4 -140 0 50 50 0 18 26 36 201999 Northeast CS 30 1140 Lakeland 14 1280 16 -140 0 50 50 0 17 31 18 341999 Northeast CS 30 1140 North Pocono 28 1400 2 -260 0 50 50 0 36 30 23 121999 Northeast CS 30 1140 Old Forge 13 1290 17 -150 0 50 50 0 18 32 22 281999 Northeast CS 30 1140 Riverside 39 1280 -9 -140 0 50 50 0 19 26 26 301999 Northeast CS 30 1140 Scranton 32 1240 -2 -100 0 50 50 0 11 24 32 331999 Northeast CS 30 1140 Valley View 17 1330 13 -190 0 50 50 0 21 31 34 141999 Northeast CS 30 1140 Forest City Regional 40 1270 -10 -130 0 50 50 0 15 25 33 271999 Northeast CS 30 1140 Wallenpaupack Area 29 1310 1 -170 0 50 50 0 17 36 29 181999 Northeast CS 30 1140 Wayne Highlands 39 1320 -9 -180 0 50 50 0 22 31 35 121999 Northeast CS 30 1140 Western Wayne 47 1260 -17 -120 0 50 50 0 10 31 36 231998 Keystone Edu. Center 65 73 1000 Greenville Area 38 1360 27 -360 0 0 9 91 30 33 13 241998 Keystone Edu. Center 65 73 1000 Reynolds 38 1260 27 -260 0 0 9 91 11 26 44 191999 Keystone Edu. Center 70 1170 Greenville Area 27 1350 43 -180 0 31 24 45 33 27 22 181999 Keystone Edu. Center 70 1170 Reynolds 39 1220 31 -50 0 31 24 45 13 19 29 391999 SUSQ-Cyber CS 0 1140 Berwick Area 51 1270 -51 -130 0 8 42 50 14 24 38 231999 SUSQ-Cyber CS 0 1140 Bloomsburg Area 44 1310 -44 -170 0 8 42 50 20 34 28 181999 SUSQ-Cyber CS 0 1140 Milton Area 29 1240 -29 -100 0 8 42 50 11 25 32 331998 Phila. Comm Acad CS 80 86 1000 Philadelphia 82 1120 -2 -120 0 0 0 100 8 12 22 591999 Phila. Comm Acad CS 93 89 1040 Philadelphia 82 1140 11 -100 0 2 14 83 7 13 23 571999 Ridgeview Academy 88 1230 Greater Latrobe 26 1330 62 -100 13 31 13 44 24 31 24 21

Table D:17 Comparisons Between Charter Schools and Their Host Districts: PSSA 11th Grade ReadingDetailed Score Breakdown

Charter Schools Host Districts Charter MinusHost

Charter School Quartile Host District Quartile

Year Name of School % LowIncome

%Partic

MeanPSSA

Name of District % LowIncome

MeanPSSA

% LowIncome

MeanPSSA

Top Hi-mid Lo-mid Bottom Top Hi-mid Lo-mid

Bottom

1999 Northeast CS 30 1110 Abington Heights 6 1300 24 -190 0 0 33 67 21 31 23 251999 Northeast CS 30 1110 Carbondale Area 53 1260 -23 -150 0 0 33 67 18 28 24 301999 Northeast CS 30 1110 Dunmore 26 1260 4 -150 0 0 33 67 13 32 30 261999 Northeast CS 30 1110 Lakeland 14 1270 16 -160 0 0 33 67 19 31 24 251999 Northeast CS 30 1110 North Pocono 28 1430 2 -320 0 0 33 67 47 28 20 51999 Northeast CS 30 1110 Old Forge 13 1320 17 -210 0 0 33 67 25 28 30 171999 Northeast CS 30 1110 Riverside 39 1280 -9 -170 0 0 33 67 21 23 28 271999 Northeast CS 30 1110 Scranton 32 1280 -2 -170 0 0 33 67 18 32 27 241999 Northeast CS 30 1110 Valley View 17 1350 13 -240 0 0 33 67 32 23 30 151999 Northeast CS 30 1110 Forest City Regional 40 1310 -10 -200 0 0 33 67 25 27 30 181999 Northeast CS 30 1110 Wallenpaupack Area 29 1300 1 -190 0 0 33 67 23 33 26 191999 Northeast CS 30 1110 Wayne Highlands 39 1290 -9 -180 0 0 33 67 21 29 24 251999 Northeast CS 30 1110 Western Wayne 47 1280 -17 -170 0 0 33 67 13 38 26 231998 Keystone Edu. Center 65 73 1070 Greenville Area 38 1330 27 -260 0 0 36 64 33 25 25 171998 Keystone Edu. Center 65 73 1070 Reynolds 38 1300 27 -230 0 0 36 64 23 28 31 181999 Keystone Edu. Center 70 1220 Greenville Area 27 1330 43 -110 10 20 30 40 30 31 19 201999 Keystone Edu. Center 70 1220 Reynolds 39 1220 31 0 10 20 30 40 13 30 19 381999 SUSQ-Cyber CS 0 1220 Berwick Area 51 1290 -51 -70 0 17 67 17 20 31 28 221999 SUSQ-Cyber CS 0 1220 Bloomsburg Area 44 1330 -44 -110 0 17 67 17 26 34 26 141999 SUSQ-Cyber CS 0 1220 Milton Area 29 1220 -29 0 0 17 67 17 11 27 26 361998 Phila.Comm. Acad CS 80 93 1010 Philadelphia 82 1140 -2 -130 0 0 17 83 9 15 24 531999 Phila.Comm. Acad CS 93 92 1130 Philadelphia 82 1140 11 -10 2 12 37 49 9 15 23 531999 Ridgeview Academy 88 1290 Greater Latrobe 26 1320 62 -30 25 25 13 38 27 31 22 21

Table D:18 Comparisons Between Charter Schools and Their Host Districts: PSSA 6th Grade WritingDetailed Score Breakdown

Charter Schools Host Districts Charter Minus Host Charter School QualitativeRating (%)

Host District QualitativeRating (%)

Year Name of School % LowIncome

MeanPSSA

Name of District % LowIncome

MeanPSSA

% LowIncome

Mean PSSA Excel-lent

Good Fair Weak Poor Excel-lent

Good Fair Weak Poor

1999 Mosaica Academy 19 1290 Bensalem Township 24 1300 -5 -10 8 38 46 8 0 6 51 39 4 01999 Centre Learning

Community10 1250 Bald Eagle Area 43 1190 -33 60 5 43 43 10 0 1 28 55 13 1

1999 Centre LearningCommunity

10 1250 Bellefonte Area 15 1260 -5 -10 5 43 43 10 0 11 37 40 7 4

1999 Centre LearningCommunity

10 1250 State College Area 9 1 5 43 43 10 0 0 0 0 0 0

1999 Nittany Valley 4 1030 State College Area 9 -5 0 50 17 0 01999 Creative Ed.Concepts 1050 Chester-Upland 63 1170 -120 0 13 60 20 0 3 24 52 18 21999 Village CS of Chester-

Upland73 1110 Chester-Upland 63 1170 10 -60 0 13 38 50 0 3 24 52 18 2

1999 La AcademiaPartnership CS

68 1120 Lancaster 59 1170 9 -50 0 0 81 13 6 1 27 52 18 1

1999 Eugenio Maria de HostsCS

100 1110 Philadelphia 82 1110 18 0 0 9 69 20 0 1 16 48 29 3

1998 Philadelphia HarambeeInstitute

1070 Philadelphia 82 1090 -82 -20 0 17 25 58 0 1 14 46 32 4

1999 Philadelphia HarambeeInstitute

1140 Philadelphia 82 1110 -82 30 0 15 66 19 0 1 16 48 29 3

1998 World CommunicationsCS

60 1170 Philadelphia 82 1090 -22 80 0 18 64 16 0 1 14 46 32 4

1999 World CommunicationsCS

60 1180 Philadelphia 82 1110 -22 70 0 27 58 12 1 1 16 48 29 3

Note: Not All Schools Were Required to Administer the Writing Portion of the PSSA During the 1997/98 and 1998/99 Academic Years.

Table D:19 Comparisons Between Charter Schools and Their Host Districts: PSSA 9th Grade WritingDetailed Score Breakdown

Charter Schools Host Districts CharterMinus Host

Charter School Qualitative Rating(%)

Host District Qualitative Rating (%)

Year Name of School % LowIncome

MeanPSSA

Name ofDistrict

% LowIncome

MeanPSSA

% LowIncome

MeanPSSA

Excellent Good Fair Weak Poor Excellent Good Fair Weak Poor

1999 Village CS of Chester-Upland

73 1200 Chester-Upland

63 1100 10 100 0 7 43 43 7 0 2 44 37 6

1999 SUSQ-Cyber CS 0 1200 Berwick Area 51 1320 -51 -120 0 71 14 14 0 1 16 68 13 0

1999 SUSQ-Cyber CS 0 1200 BloomsburgArea

44 1400 -44 -200 0 71 14 14 0 8 31 55 6 1

1999 SUSQ-Cyber CS 0 1200 Milton Area 29 1390 -29 -190 0 71 14 14 0 13 26 37 23 1

1999 Center for Economicsand Law CS

70 1230 Philadelphia 82 1200 -12 30 0 8 60 25 5 2 13 42 29 8

1999 Imhotep CS 75 1230 Philadelphia 82 1200 -7 30 3 24 44 15 3 2 13 42 29 8

1999 Multi-culturalAcademy CS

89 1330 Philadelphia 82 1200 7 130 4 24 52 20 0 2 13 42 29 8

1999 The Preparatory CS 61 1330 Philadelphia 82 1200 -21 130 1 15 71 14 0 2 13 42 29 8

1998 WorldCommunications CS

60 1200 Philadelphia 82 1140 -22 60 0 6 37 50 7 1 8 35 36 11

1999 WorldCommunications CS

60 1250 Philadelphia 82 1200 -22 50 0 4 65 31 0 2 13 42 29 8