Alumni Giving Implications

7/31/2019 Alumni Giving Implications

1/26

International Journal of Educational Advancement. Vol.7 No.4

INTERNATIONAL JOURNAL OF EDUCATIONAL ADVANCEMENT. VOL.7 NO.4 307332

2007 PALGRAVE MACMILLAN LTD. ISSN 17446503 $30.00

307

www.palgrave-journals.com/ijea

AbstractDespite readily available alumni surveydata warehoused at many alumniassociations and foundations acrosscolleges and universities, researchershave underutilized the abundant

available data to identify keypredictors of alumni donation,including factors that trigger alumnidonation behavior. Utilizing the datafrom a two-year alumni surveyconducted at a Midwest publicuniversity, a multivariate causal modelthat captures the determinants ofalumni donation was applied to thedata. Four hypotheses were tested.Three were found to be significant.

Based on a multivariate causalmodel that analyzed data from atwo-year alumni survey, the findingssuggest that alumni fundraisers andhigher education administrators mayincrease alumni solicitations if they

collaboratively create a comprehensivecommunication strategy to reachalumni; focus on current students asfuture funders, provide qualityeducational experiences to students;encourage and support relationshipbuilding between faculty and currentstudents and graduates; enhancealumni services based on stakeholdersneeds; and most importantly, redirectand expand efforts to connect witholder female alumni.International Journal of Educational

Advancement(2007) 7, 307332.doi:10.1057/palgrave.ijea.2150073

Keywords:alumni giving, communication strategy,

fund raising

A Multivariate Causal Model of

Alumni Giving: Implications forAlumni FundraisersReceived (in revised form): October 29, 2007

Xiaogeng Sunearned a Ph.D. in educational studies from the University of Nebraska-Lincoln. He is the AssistantDirector of Program Evaluation in the Anchorage School District, Alaska.

Sharon C. Hoffmanis a Ph.D. candidate at the University of Nebraska-Lincoln in educational leadership. Her researchinterests are in leadership theory and developing organizational strategic capacity.

Marilyn L. Gradyearned a Ph.D. and is Professor of Educational Administration at the University of Nebraska-Lincoln. Her research interests are in leadership theory, leader preparation, and advocacy.

Authors Contact Address:

Xiaogeng Sun

Anchorage School District

5530 E. Northern Lights Blvd.

Anchorage AK 99504, USA

Phone: +1 907 743 0599

Fax: +1 907 743 0599

Email: [email protected]


2/26

Xiaogeng Sun, Sharon C. Hoffman and Marilyn L. Grady



308

IntroductionUS higher education institutionsimproved fund-raising results overall in

2006, a 9.4 percent or $2.4 billionincrease from 2005.1 This was thelargest increase since 2000, due tolarger single donation amounts fromalumni and others. What did notincrease, however, was the number ofalumni donors. Even though alumnidonations improved 18.35 percent, thenumber of alumni donors remainedflat, or slightly decreased by 0.02percent (Strout, 2007a, p. A1).

Alumni fundraisers are discovering

that their efforts compete more thanever before with larger athleticdepartments as they vie for donations.In 1998, athletic gifts accounted for14.7 percent of all contributions.By 2003, alumni and other donorcontributions to sports accountedfor 26 percent of all donations(Wolverton, 2007, p. A1).

Given the static number of alumnicontributors and the competition foralumni dollars, the results of this study

have important implications for highereducational institutions, and alumniassociations in particular. If universityadministrators and alumni fundraisersprioritize fund-raising strategies andexpenditures, these efforts may resultin more cost-effective approaches withincreased alumni contributions.

The purpose of this study was toinvestigate the impact of studentexperience, alumni experience, alumnimotivation, and demographic variables

associated with alumni on alumni giving.

Alumni-Giving Decision ModelThe researchers proposed the alumni-giving decision model featured inFigure 1.

Student experienceAn argument exists that alumni whowere treated favorably as students,who were satisfied with their academic

experiences, and who believe theircollege education contributed to theircareer success are more inclined togive as alumni than those with lessfavorable feelings and beliefs.

Alumni experienceAlumni experience is alumniperceptions of their interactions withtheir alma maters after graduation.Alumni experience can be perceived

as marketing efforts of the alumniassociation because most alumniinteract with the alma materthrough the alumni association.

Alumni motivationThe research results indicated thatalumni motivation to give wassignificantly related to actual giving(Gardner, 1975; Beeler, 1982; Oglesby,1991; Halfpenny, 1999; Clotfelter,

2003). Motivation is defined as aninternal state or desire that serves toactivate behavior (Kleinginna andKleinginna, 1981). Alumni motivationis the internal desire that is rooted deeplyenough in ones awareness to induce adesire to give to the alma mater.

Student Experience

Alumni Experience

Alumni Motivation

Demographic Variables

Alumni Donation

Figure 1: Alumni-giving decision model


3/26

A Multivariate Causal Model of Alumni Giving



309

Demographic variablesSpecific demographic variables havebeen found to be significantly related

to alumni giving. According to Bristol(1990), the number of years betweengraduation and onset of giving had asubstantial effect on the magnitude ofalumni giving. The participation ratein alumni donation rises with theincrease of class age. Research byGraham and Husted (1993)demonstrated a high correlationbetween alumni wealth and theirdonations to the alma mater.

General HypothesesThe following research hypotheseswere proposed in this study.

Hypothesis 1: Student experience

significantly distinguishes

alumni donors from

nondonors.

Hypothesis 2: Alumni experience


alumni donors from

nondonors.

Hypothesis 3: Alumni motivation


alumni donors from

nondonors.

Hypothesis 4: Demographic variables

(graduation year, gender,

ethnicity, type of degree, in

or out of state, membership

status) significantly

distinguish alumni donorsfrom nondonors.

Delimitations and limitationsOne delimitation applies to thisstudy. The study was based on theexperiences reported by alumni at one

Midwest university on the survey datain each of two years: 2001 and 2002.

LimitationsThe following limitations wereidentified for the study:

1. This research assumes the sampleto be as representative of thepopulation as possible. Samplingerrors, however, may not provide anextensive picture of the experiencesof the study population.

2. Because the data collection methodwas a self-report survey, respondents

may have compromised their responsesand provided socially desirable answers.

3. Responses to questions related tostudent experience and alumniexperience may also be susceptiblebecause of the reliance on thememories, biases, prejudices, andperspectives of respondents.

4. Self-selection of the respondentsresulted in a greater number ofalumni donor participants in thissurvey than alumni nondonors.

5. The survey data were collected afterSeptember 11, 2001. This tragic eventmay have influenced the donationbehavior of American citizens. Noattempt, due to data constraints, wasmade to adjust for this limitation.

Significance of the studyThe results of this study are importantfor several reasons. First, throughexamination of the findings, alumni

fundraisers may improve their abilityto distinguish alumni donors fromnondonors and target likely donors intheir fund-raising efforts. Alumni fundraising may improve with reliableapproaches to identifying potentialdonors from nondonors.


4/26




310

Secondly, fund raising is generallyregarded as relationship marketing.Alumni who believe or feel they are

well treated are more inclined togive than alumni without similarbeliefs or feelings. The results mayidentify several alumni services thatalumni deem important but the alumniassociation may not adequatelyprovide. The studys results may helpthe alumni association identifyimproved services in order to increasealumni satisfaction and subsequentgiving.

Finally, the results may enrich the

body of alumni research by specificallytargeting alumni at a Midwest, public,research-extensive university. Otherhigher education institutions may finetune this model according to theirspecific situations and identify importantdeterminants of alumni giving.

Theoretical Frameworks onCharitable GivingResearch development on charitable

giving in general, and alumni giving inparticular, is attributable to thetheoretical and conceptual frameworksproposed by both economists andsociologists. Some of the most prominenteconomic frameworks are proposed byBecker (1974), Sugden (1984), Andreoni(1989, 1990), and Steinberg (1997).Among sociologists, important theoriesare proposed by Thibaut and Kelley(1959), Blumer (1969), Walster andWalster (1978), Foa and Foa (1974),

and Bar-Tal (1976).

Neoclassical microeconomictheories on charitable givingNeoclassical microeconomic theorymakes the following three assumptions

(Weintraub, 1985).

1. Individuals have preferences foroutcomes.

2. Utility was maximized byindividuals.

3. Individuals act independently basedon full and relevant information.

Public good theoryIn public good theory, the purpose ofcharitable donations is the collectiveinterest of the donors and donees. Thedonor gains utility from the totalutility out of donations from alldonors, and not from an individual

recipients additional utility.

Social exchange theoryHalfpenny (1999) maintains that somecommon statements from the socialexchange theory pertain to thedonation process. First, the socialexchange theory focuses on the humaninteraction during social exchange.An exchange occurs only when bothparties in the exchange find theirrewards attractive. The question Why

do donors give money and time toothers without rewards? providesopportunity for exploration. Oneexplanation is that a reward isobvious to the giver, but not obviousto the observer.

Equity theoryThe equity theory (Walster andWalster, 1978) assumes that societyrewards individuals for equity in theirinteractions with others. The

significance of the equity theory lies inthe feeling of distress that derives frominequitable relationships. Greaterunderstanding of individuals donationdecisions is more likely if there isgreater knowledge about howindividuals perceive rewards and


5/26




311

distress in a donation process(Halfpenny, 1999).

Symbolic interactionismSymbolic interactionism (Blumer, 1969)asserts that individuals get the meaningabout the world through interactionwith the surrounding environment.Symbolic interactionism, when appliedto the donation process, suggests thatthrough interaction with the social andphysical environment, individualsassume or reject the role of donoror donee (Halfpenny, 1999).

Supply and demand analysisof alumni givingYoo and Harrison (1989) appliedsupply and demand analysis to alumnigiving and claimed that it seems morelogical to deem donors as conventionalbuyers who purchase some servicesfrom donees. Therefore, Yoo andHarrison (1989) suggested using theaverage amount of donation per donoras the price donors pay for the services

they receive from their alma mater.Donors derive utility from the

services provided by the recipients. Thereciprocity between the donor anddonee may be in the form of honorsand alumni services. Yoo and Harrison(1989) also asserted that donorsshould be regarded as buyers whoreceive tangible and intangible benefitsfrom the recipients. They classifiedgifts as one form of market exchangein which both donors and donees are

motivated by self-interest.

Organizational theories onalumni givingMael and Ashforth (1992) appliedOrganizational Identity Theory (OID)

to the study of alumni giving.Organizational identification is definedas a perception that one belongs to an

organization and shares the successand failure of the organization.According to the social identity theory,the self-concept is characterized by apersonal identity that includescharacteristics such as abilities andinterests, and a social identity thatincludes group classification. Thesocial identity theory predicts thatindividuals tend to participate inactivities that match their socialidentities and support the institutions

representing these identities. Theresearchers considered alumni givinga possible outcome of organizationalidentification (Mael and Ashforth,1992).

Other theories on alumni givingCoelho (1985) suggests that the needfor status is the motivation fordonations because the need to seekstatus explains many actions. Alumni

donations may be alternative sourcesof funding for public goods. Halfpenny(1999) suggests that unconstraineddonations create wealth: the value asa charitable giving to a donee exceedsthe value kept for the donors ownuse. Winston (1999) argues thatuniversities are nonprofit; therefore,they must acquire alternative fundingfrom donations.

Donative revenues can be consideredan indicator of the institutions

education performance: alumni whodonate are recognizing the role thatthe institution played in theireducation. Burt (1989) indicates thatdonation might be understood as ameasure of the quality of education.Alumni giving also depends on the


6/26




312

alumni associations effectiveness insolicitations. Clotfelter (2003) foundthat donations to private universities

are higher than those to the publicuniversities.

Empirical Studies onAlumni Giving

Demographic variablesA body of literature of particularinterest to college presidents andalumni fundraisers focuses on thecharacteristics of alumni. Olsen et al.(1989) tested a lifecycle hypothesis of

alumni giving at a small liberal artscollege. They found that the growthrate of donations is related to the ageincome profile of donors. In a similaragedonation profile of more than4,000 alumni of a large publicuniversity, Okunade et al. (1994)estimated that growth rates of alumnidonations decline after age 52, short ofretirement age. They also found largerdonations from business schoolgraduates, members of non-Greek

social organizations. Grimes andChressanthis (1994) studied the effectof intercollegiate athletics on alumnidonation at a Division I NCAA schooland found that a winning season andtelevision appearances are related tohigher alumni giving.

Other research on thedemographics of alumni donorsTaylor and Martin (1995) investigated

selected attitudinal, demographic,involvement, and philanthropiccharacteristics of alumni donors andnondonors from a Research I, publicuniversity. The authors used a 32-itemself-reporting survey instrument fordata collection.

Six variables were found to bediscriminators of donors/nondonors:family income, need for financial

support, reading alumni publications,enrollment for graduate work, specialinterest group, and involvement withuniversity as an alumnus/a.

Bruggink and Siddiqui (1995) built amodel based on the characteristics ofthe alumni at an independent liberalarts college. With this model, theyintended to identify factors thatinfluence giving. The dependent variablewas the amount of alumni givingduring the fiscal year. The independent

variables included income, age,marriage, Greek, year, and employmentstatus. Bruggink and Siddiqui (1995)reported that income, Greek status,alumni activity, distance from the almamater, years after graduation, maritalstatus, and alumnis major werestatistically significant.

Hueston (1992) used a logisticregression framework to analyze 34,938alumni case files at New Mexico StateUniversity. Roughly 16 percent of the

high alumni donor group graduatedfrom the business school.

Lindahl and Winship (1992) built amodel for fund-raising analysis atNorthwestern University based on2,803 alumni data sets and logit modelregression estimates. They found pastgiving history, year of donation,religious beliefs, and salary levels tobe significant predictors of alumnidonation.

Okunade et al. (1994) used

covariance regression and a 1975/761989/90 data sample to analyze thedonation of undergraduate alumni ofMemphis State University whograduated in the 1926/271975/76period. They found that business schoolalumni, alumni who later earned a


7/26




313

graduate degree at the same university,and alumni members of non-Greeksocial clubs made more donations than

other alumni. Giving was also found tovary over business cycles.Okunade and Berl (1997) used a

logit regression model to analyze thesurvey response data from the1955/561990/91 alumni of a public,Research I university. The time sincegraduation, major, willingness torecommend the university, householdattributes, family ties to the almamater, and the availability of matchinggift accounts were found to be

significant factors of alumni donations.With data sets from two publicuniversities in the United Kingdom,Belfield and Beney (2000) examinedthe predictors of alumni giving. Usingordinary least square techniques, theyestimated two models for both theprobability of giving and the totalamount given. They found gendereffects to be statistically significant.Females were found to have a higherpropensity to give than males. Belfield

and Beney (2000) also found thatthose who are married have a lowerprobability to give. The elasticity ofamount given with respect to incomeis estimated at 0.20850.5534. Theirfinding supported Okunade andBerls (1997) study that alumnigiving increased with age but at adecreasing rate.

Economic variables

In Harrisons (1995) study, use of theparameters was based on the particularsample and historical period ofobservations. He emphasized that itwould not be useful to use thesecoefficients to predict a particularschools outcome from future alumni

relations expenditures. Each schoolneeds to build its own model.

Other researchLeslie et al. (1983) studied thecharitable donations to highereducation as an aggregate from 1932to 1974. Charitable contributions tohigher education as an aggregate werefound to be significantly related tobusiness environments. Leslie alsofound opposite effects of economicconditions on personal and corporatedonations. Leslie et al. (1983)concluded that personal giving will rise

when economic conditions are poorand higher education institutions aremost in need of donation, whilecorporate giving will increase withgood economic conditions.

Motivation variablesSome studies have explored theinfluence of motivation on alumnidonations. Miracle (1977) indicatedthat alumni who recognize thefinancial needs of the alma mater

would be more motivated to give thanthose without similar perceptions.

Alumni experienceNumerous studies focused on therelationship between postgraduatealumni activities and alumni donations.Shadoian (1989) used several variablesrelated to alumni experience in hisalumni donation model. Number ofpostgraduate campus visits, reading

alumni publications, and contactswith faculty members were foundto be significant predictors of futuredonation.

Miracles (1977) study results weresomewhat different from those ofShadoians. He did not find number


8/26




314

of campus visits after graduation tobe an important predictor; however,postgraduate involvement with the

university was a significant factor indetermining alumni donation.Oglesbys (1991) results confirmedthe significance of postgraduateinvolvement with the university asa predictor of alumni giving.

Oglesby (1991) also examined theinfluence of spouse donation. Heconcluded that spouse contribution toan alma mater was not a significantvariable. He suggested that anothervariable, supporting other charitable

causes, was significant.

Methodology and Procedures

Research designFor this study, a nonexperimental,applied research design and data from atwo-year alumni satisfaction survey wereutilized to investigate the impact ofstudent experience, alumni experience,alumni motivation, and demographicvariables on alumni giving.

HypothesesTo address the purpose of this study, thefollowing hypotheses were proposed:

Hypothesis 1: Student experience


alumni donors from

nondonors.

Hypothesis 2: Alumni experience


alumni donors fromnondonors.

Hypothesis 3: Alumni motivation


alumni donors from

nondonors.

Hypothesis 4: Demographic variables

(graduation year, gender,

ethnicity, type of degree, in

or out of state, membershipstatus) significantly

distinguish alumni donors

from nondonors.

SiteThe studys site was a Midwest,public, Research Extensive university.Officially chartered in 1869, thisuniversity was one of the nationsleading state universities. At the timeof the study, the university offered 149

undergraduate majors and 116graduate degree programs and had23,000 students enrolled in classes.The university had 175,000 alumniresiding in the United States andaround the world. Among them,25,000 alumni email addresses werelisted in the database.

InstrumentThe instrument used in this study was a

proprietary survey designedcooperatively by PerformanceEnhancement Group and alumniassociation officials from 12 public andprivate universities. They worked in aseries of focus group studiesto discuss the items to be included inthe survey. The survey sought responsesto questions about four areas of thealumni experience: overall experience,loyalty, student experience, and alumniexperience. A fifth area asked for

specific demographic information.The instrument, administered

annually since 2001, retains andimplements most items to enhance thelongitudinal integrity of theinstrument. In any year, additionalitems may be added to reflect new


9/26




315

inputs from participating alumniassociations. All variables included inthis study were present in both 2001

and 2002 surveys.

Population and sampleThe population for this study was175,000 alumni of the Midwestuniversity. The alumni association firstused simple random sampling to selectrespondents. The PerformanceEnhancement Group then adjusted thesample to expand its representation.The final set of selected respondentsincluded 5,960 alumni in 2001 and

5,499 alumni in 2002. In the end,1,030 alumni completed the 2001survey and 724 alumni completed the2002 survey. The response rate was 24and 18 percent for 2001 and 2002,respectively.

Variable selection

Criterion variableResponses to one survey question

served as the criterion variable:Which of the following best describesyour financial support of theuniversity? Five choices wereavailable to the respondents:

1. Have never financially supported theuniversity and do not plan to in thefuture (never/do not plan to).

2. Have financially supported theuniversity but do not plan tocontinue (donated/wont

continue).3. Have never financially supported the

university but plan to in the future(never/but plan to).

4. Currently financially support theuniversity and plan to continue(donated/plan to continue).

5. Currently financially support theuniversity and plan to increase inthe future (donated/plan to

increase).

According to their responses, alumniwere divided into five groups: never/do not plan to, donated/wontcontinue, never/but plan to,donated/will continue, anddonated/will increase. The intent ofthe research was to determine whatpredictors, if any, can explainmembership in the five groups.

Predictor variablesAccording to Stevens (2002), thenumber of predictors is a crucial factorthat determines how well a givenequation will generalize. Generally,an n/k ratio (n= sample size andk= predictors) of 15 is needed for areliable equation. In the originalsurvey, there were 71 questions. Thisnumber raised concerns that too manypredictors might decrease thepredictive power of the model.

Furthermore, several questionsreferred to the same or similarconstructs. This is viewed as amulticollinearity problem, that is,there are moderate to high inter-correlations among the predictors.

A preliminary correlation matrixwas constructed in order to examinecorrelations and determine whethermulticollinearity existed. When a largenumber of explanatory variables presentsuch a multicollinearity problem, the

relative contributions of some variablesare clouded. The data did, in fact,exhibit a strong linear relationshipamong a number of the survey items.Because of the presence of themulticollinearity problem, data reductionthrough factor analysis was used.


10/26




316

Table 1: Factor loadings

Factor 1 Factor 2 Factor 3 Factor 4 Factor 5

Factor 1Q17BP_1 Quality of alumni association website 0.708aQ17HP_1 Quality of communication regarding

your services/benefits0.702a

Q17DP_1 Quality of electronic newsletter 0.678aQ17AP_1 Quality of monthly bulletins 0.638aQ17CP_1 Quality of university website 0.633aQ17 IP_1 Quality of invitations to university activities 0.624aQ17GP_1 Quality of email 0.622aQ17EP_1 Quality of reunion mailings 0.592aQ17 JP_1 Quality of alumni magazine 0.591aQ17FP_1 Quality of alumni staff presentations

at meetings0.528a

Factor 2Q08 II_1 How important is it for alumni to provideleadership by serving on boards.

0.752a

Q08HI_1 How important it is for alumni to volunteerfor the university

0.738a

Q08 JI_1 How important it is for alumni to attendevents

0.683a

Q08 EI_1 How important it is for alumni to serve asambassadors

0.643a

Q08GI_1 How important it is for alumni to networkwith other alumni

0.631a

Q08DI_1 How important it is for alumni to recruitstudents

0.617a

Q08FI_1 How important it is for alumni to providefinancial support for the university

0.584a

Q08CI_1 How important it is to provide feedbackabout community perceptions 0.552a

Q08AI_1 How important it is for alumni to mentorstudents?

0.457a

Factor 3Q07B_1 Commitment to continuous learning 0.804aQ07C_1 Responding to new career opportunities 0.787aQ07E_1 Deepening my understanding and

commitment to personal development0.777a

Q07F_1 Further graduate education 0.718aQ07A_1 Current work status 0.707aQ07D_1 Contributing to my community 0.692a

Factor 4

Q10MP_1 What I learned about life 0.721a

Q10NP_1 Exposure to new things 0.717aQ10BP_1 Relationships with other students 0.651aQ10OP_1 Traditions or values learned on campus 0.626aQ10DP_1 Relationships with faculty 0.613aQ10CP_1 Academics classes 0.584aQ10JP _1 Relationships with administrators and staff 0.488a


11/26




317

Factor analysisIn the factor analysis, five factors wereextracted based on the followingrationale.

The list of items included in the

factor analysis is displayed in Table 1.The number and content of higherloading survey items included in eachfactor are also presented.

Factor 1named alumni experiencecontained ten survey items. Five itemsthat come with the highest loadings were:

Please tell us the quality/usefulnessof alumni association monthlybulletins.

Please tell us the quality/usefulnessof alumni association communicationregarding your membership.Please tell us the quality/usefulnessof alumni association electronicnewsletters.Please tell us the quality/usefulnessof university website.Please tell us the quality/usefulnessof invitations to university activities.

Most of the survey items were related

to alumni satisfaction with the servicesand information provided by thealumni association. This inclusion isconsistent with previous research thatconcluded that alumni experience is asignificant predictor of alumni donation(Oglesby, 1991).

Factor 2named alumni motivationincluded nine survey items. Items withthe highest loading were:

In your opinion, how important iseach of the following alumniactivities to the university: providingleadership by serving on boards,committees, etc?In your opinion, how important iseach of the following alumniactivities to the university:volunteering for the university?In your opinion, how important iseach of the following alumni activitiesto the university: attending events?

In your opinion, how important iseach of the following alumniactivities to the university: serving asan ambassador for the university?In your opinion, how important iseach of the following alumniactivities to the university:networking with other alumni?

These survey items were linked withalumni motivations to support theuniversity, which corresponds to the

alumni motivation in the decisionmodel.

Factor 3named student experienceimpact on career because these surveyitems were closely linked with studentexperience and how alumni appreciate

Table 1: Continued.

Factor 1 Factor 2 Factor 3 Factor 4 Factor 5

Factor 5Q10KP_1 Student leadership opportunities 0.707aQ10HP_1 Participation in fraternity/sorority 0.641aQ10FP_1 Attending cultural events 0.464aQ10EP_1 Attending athletic events 0.424a

aItem with the highest factor loadings.


12/26




318

the contribution of college educationto their future careersincluded sixsurvey items. Items with the highest

loading were:

How well did the highest degreefrom the university prepare you forcommitment to continuous learning?How well did the highest degreefrom the university prepare you forresponding to new careeropportunities?How well did the highest degreefrom the university prepare you fordeepening understanding and

commitment to personaldevelopment?How well did the highest degreefrom the university prepare you forfurther graduate education?

Factor 4named student experience-relationshipsincluded seven surveyitems. The survey items of studentexperience are closely related to thealumni experience, especiallyrelationships with others while they

were students. Items with the highestloading were:

How did what you learned aboutlife (on campus) affect your studentexperience?How did exposure to new things(on campus) affect your studentexperience?How did attending athletic eventsaffect your student experience?How did relationships with other

students affect your studentexperience?How did relationships with facultyaffect your student experience?

Factor 5named student experience-extracurricular activitiesis closely

related to six survey items. These itemsare clearly connected with the alumniexperience of extracurricular activities

while they were students. Items withthe highest loading were:

How did student leadershipopportunities affect your studentexperience?How did participation in fraternity/sorority affect your studentexperience?How did attending culture eventsaffect your student experience?How did attending athletic events

affect your student experience?How did orientation for newstudents affect your studentexperience?

A pictorial summary of factor andfactor names is displayed in Table 2.

Statistical techniqueDiscriminant analysis was used toidentify the important predictors of

alumni giving and to assess theusefulness of the predictive model.Discriminant analysis is used for twoproposes: (a) to statistically revealdifferences among different groups ofthe criterion variable and (b) to

Table 2: Factors and factor names

Factor Factor name

F1 Factor 1 Alumni experienceF2 Factor 2 Alumni motivationF3 Factor 3 Student experienceimpact on

careerF4 Factor 4 Student experience

relationshipsF5 Factor 5 Student experience

extracurricular activities


13/26




319

classify respondents into differentgroups based on predictor variables(Stevens, 2002).

Discriminant analysis has twofeatures: (a) parsimony of descriptionand (b) clarity of interpretations(Stevens, 2002, p. 286). Thediscriminant analysis is parsimoniousbecause predictor variables can belinearly combined to form a lessdiscriminant function. In this study,there were 11 predictor variables, andthe criterion variable had five differentgroups (never/do not plan to,donated/wont continue, never/but

plan to, donated/will continue, anddonated/will increase). In the end,one may only find that these fivegroups differ mainly on twodimensions. That is, only twosignificant discriminant functionsexplain the difference among differentgroups. Discriminant analysis is alsostraightforward to interpret in thatthese functions are uncorrelated. Eachdiscriminant function only reveals onedimension, that is, one reason, to

explain the difference among differentgroups of the criterion variable.

Data analysisThe purpose of this study was todetermine whether or not quantitativesupport existed for the multivariatealumni decision model, that is, whetherthe predictors specified in the modelsignificantly distinguished donors fromnondonors. This question was

addressed by considering whether thevariables measured in the surveys werejointly related to distinguishing donorsfrom nondonors.

In order to address the collinearityproblem, that is, too many predictorsthat were highly correlated, a factor

analysis was applied to produce datareduction. As a result of the factoranalysis, 33 items were excluded from

the factor analysis because of lowfactor loadings. The remaining 38items generated five factors describedearlier. These five factors alsocorresponded to the alumni-givingdecision model.

In the second step, after predictorvariables were identified and classifiedas factors, the discriminant analysiswas used as the primary methodto assess the significance of eachpredictor variable in the model.

Survey item 21 was appointed thecriterion variable and had five levelsof response. Table 3 displays thecriterion variable and predictorvariables included in the discriminantanalysis.

The discriminant analysis equationassessed the significance of thepredictors to correct classification ofalumni into degrees of donor ornondonor groups. The researcher canconclude whether or not the collection

of predictors correctly classifiedindividuals into groups and which ofthe variables contribute significantly tothe prediction of group membership(Stevens, 2002, p. 286).

Analysis Results

Introduction of the data analysisSPSS for Windows 13.0 software wasused for the analysis. The data analysis

involved three steps:First, factor analysis generated five

factors. These five factors, togetherwith other demographic variablessupported by previous research, wereused as predictors in the follow-updiscriminant analysis.


14/26




320

Secondly, discriminant analysis wasused to analyze the differences amongfive alumni groups and to identifyimportant predictors distinguishing

these groups. Discriminant analysisgenerated the linear combination ofthe predictors that maximallyseparated five alumni groups.

Thirdly, the predictive ability of thediscriminant function model, or thesuccess of this model in correctlyclassifying alumni into their specificgroups, was tested. The results of thesethree steps are briefly described below.

Factor analysis resultsIn the factor analysis, 33 survey itemswere excluded because of their lowfactor loadings. The remaining 38survey items were reduced to fivefactors. These five factors have beendescribed in the methodology section.

Discriminant analysis resultsAmong the four discriminant functionsgenerated by discriminant analysis,only the first two functions were

statistically significant. The firstdiscriminant function was identifiedas the dimension that maximallyseparates Group 3 never/but planto and Group 4 donated/willcontinue. The second significantfunction was interpreted as thedimension that maximally separatesGroup 1 never/do not plan toand Group 5 donated/willincrease.

In identifying significant predictors,

graduation year andgender wereidentified as significant predictors ofthe first discriminant function thatmaximally separates Group 3 never/but plan to and Group 4 donated/will continue. That is, the differencesbetween these two groups can be

Table 3: List of predictors in the discriminant analysis

Variables Origination of variable

Criterion variableDonation status Survey item #21Choice 1 never/do not plan to

Survey item #21Choice 2 donated/wont continueSurvey item #21Choice 3 never/but plan toSurvey item #21choice 4 donated/plan to continueSurvey item #21Choice 5 donated/plan to increase

Predictor variablesF1 Alumni experience Generated from factor analysisF2 Alumni motivation Generated from factor analysisF3 Impact on career Generated from factor analysisF4 Relationship Generated from factor analysisF5 Extracurricular Generated from factor analysis

Type of degree Survey item: Degree obtained from the alma mater: undergraduate/graduate

Graduation year Survey item: The year that the alumnus/a graduatedGender Survey item: Gender of the alumnus/aEthnicity Survey item: Ethnicity of the alumnus/aMembership status byrespondent

Survey item: Whether the alumnus/a is a member of the alumni association

In or out of state Survey item: Whether the alumnus/a is residing in state or out of state


15/26




321

mainly explained by the graduationyear and gender variables.

Alumni motivation, alumni

experience, student experience-relationships, and student experience-extracurricular activities were identifiedas significant predictors of the seconddiscriminant function, whichmaximally separates Group 1 never/do not plan to and Group 5donated/will increase.

Classification of alumni resultsThe other function of discriminantanalysis is to classify alumni into different

groups. For example, if an alumnus/aanswered in the survey that he/she neverdonated before and will not donate in thefuture, he belongs to Group 1 never/donot plan to in this study. Discriminantanalysis, based on the prediction of 11predictor variables, will also assign thealumnus/a to one of the five alumnigroups. If discriminant analysis predictedthat the alumnus/a was in Group 1 andhis/her actual response matches thisplacement, this would be called a correct

classification. The percent of alumnicorrectly classified are often called hitrate. The percent of alumni correctlyclassified in each of the five alumnigroups is presented in Table 4.

Overall, 56.3 percent of the alumniwere correctly classified. This modelperformed well at classifying alumniinto Group 4 donated/will continue(81 percent correctly classified) andGroup 5 donated/will increase (80.2percent correctly classified).

Description of output statisticsand diagnostics

Assumption testAs stated earlier, caution should beused when the assumptions of

discriminant analysis, especially theassumption of equal variance, areviolated. Boxs M tested the

assumption of equality of variancecovariance across groups. If this testwas significant, the equal varianceassumption of discriminant analysiswas violated.

The Boxs M test (see Table 5) wassignificant and indicated that thevariancescovariances were not equalamong the five alumni groups. Theinequality of variance indicated thateach alumni group had a differentstandard deviation on the predictor

variables. The inequality of covariancesuggested that the correlation betweenevery two alumni groups was differentamong these five alumni groups. Sincethis assumption is violated, a separatevariancecovariance matrix for eachalumni group was used.

Significant test of discriminantfunctionsIn this study, there were 11 predictor

variables, and the criterion variable hadfive groups (never/do not plan to,donated/wont continue, never/butplan to, donated/will continue, anddonated/will increase). According tothe rule stated earlier, discriminantanalysis can generate up to fourdiscriminant functions. Eigenvalue Table(Table 6) and Wilks Lambda Table(Table 7) assessed the significance andcorrelations of these four discriminantfunctions. The discriminant analysis

procedure provided the eigenvalue,displayed in Table 6, and WilksLambda, displayed in Table 7, forassessing how well the discriminantmodel as a whole fit the data.

The Eigenvalue Table providedinformation about the usefulness of


16/26




322

each discriminant function. Theeigenvalue is equivalent to a Pearsonscorrelation between the discriminantscores and the groups. The information

in Table 6 makes it clear that,although four discriminant functionswas calculated by computerautomatically, nearly all of thevariance explained by the model weredue to the first two discriminantfunctions (96.6 percent of the variance

explained). SPSS still calculatedFunction 3 and Function 4, but due totheir low eigenvalues, these twofunctions were ignored.

Wilks Lambda Table (Table 7)assessed the significance of the fourdiscriminant functions and the result

Table 4: Classification results

Predicted group membership

1 2 3 4 5

Best describe your financialparticipation

Never/donot plan to

Have/donot plan to

Never/butplan to

Currently/plan tocontinue

Currently/plan toincrease

Total

Original count1. Never/do not plan to 21 0 69 93 0 1832. Have/do not plan to 8 0 24 108 0 1403. Never/but plan to 12 0 241 175 2 4304. Currently/plan to continue 18 0 121 723 1 8635. Currently/plan to increase 0 0 25 111 2 138

Original %1. Never/do not plan to 11.5 0 37.7 50.8 0 1002. Have/do not plan to 5.7 0 17.1 77.1 0 100

3. Never/but plan to 2.8 0 56 40.7 0.5 1004. Currently/plan to continue 2.1 0 14 83.8* 0.1 1005. Currently/plan to increase 0 0 18.1 80.4* 1.4 100

Cross-validated count1. Never/do not plan to 20 0 70 93 0 1832. Have/do not plan to 8 0 24 108 0 1403. Never/but plan to 14 0 231 183 2 4304. Currently/plan to continue 18 0 127 717 1 8635. Currently/plan to increase 0 0 25 112 1 138

Cross-validated %1. Never/do not plan to 10.9 0 387.3 50.8 0 1002. Have/do not plan to 5.7 0 17.1 77.1 0 1003. Never/but plan to 3.3 0 53.7 42.6 0.5 100

4. Currently/plan to continue 2.1 0 14.7 83.1* 0.1 1005. Currently/plan to increase 0 0 18.1 81.2* 0.7 100

*Alumni groups with highest rate of correct clarification.

Table 5: Boxs M test results

Boxs M 919.706

F Approx. 3.399

d.f.1 264d.f.2 734811.552Sig. 0.000a

aAlpha < 0.05.


17/26




323

corroborated the finding from theEigenvalue Table. Wilks Lambda is ameasure of how well each functionseparates individual alumnus/alumnaeinto different alumni groups. It is equalto the proportion of the total variance in

the discriminant scores not explained bydifferences among the groups. Smallervalues of Wilks Lambda indicate greaterdiscriminatory ability of the function.Wilks Lambda agreed that only the firsttwo functions were useful. The test ofFunction 3 and Function 4 had asignificance value greater than 0.10, andso these functions contributed little tothe model.

The associated chi-square statistic(Table 7) tested the null hypothesis

that the means of the functions listedwere equal across groups. The smallsignificance value indicated that thediscriminant function did better thanchance at separating the groups. Onlythe first two discriminant functionswere significant at the p< 0.05 level.

These data further confirmed that onlythe first two functions are significantfunctions that separate five alumnigroups.

Interpretation of significantdiscriminant functionsSince the first two discriminant functionswere significant, the next question is: howdo we interpret these two discriminantfunctions, that is, linear combinations ofpredictor variables? Table of GroupCentroids (Table 8) helps in understandingthe dimension that is associated with eachdiscriminant function. A centroid is themean of discriminant scores generated bythe discriminant function in each alumni

group.Table of Group Centroids (Table 8)

displays the mean of discriminantscores by each alumni group. Forexample, in the first discriminantfunction, the mean discriminant score(centroid) for the never/but plan togroup was 0.619 and the meandiscriminant score (centroid) for thedonated/will continue groupwas 0.371. The distance between thecentroids of these two alumni groups

was the greatest. Therefore, the firstdiscriminant function mainly separatesGroup 3 and Group 4. That is, alumniin Group 3, never/but plan to, andGroup 4, donate/will continue,mainly differ on the first discriminantfunction (dimension).

Table 6: Eigenvalue

Function Eigenvalue % of varianceexplained

Cumulative % varianceexplained

Canonical correlation

1 0.199 65.8 65.8 0.4072 0.093 30.8 96.6a 0.2923 0.007 2.2 98.8 0.0824 0.004 1.2 100 0.059

aFirst two functions explained 96.6% of the total variance.

Table 7: Wilks Lambda

Test offunction(s)

WilksLambda

Chi-square d.f. Sig.

1 through 4 0.755 448.431 44 0.000*2 through 4 0.906 158.658 30 0.000*3 through 4 0.99 16.406 18 0.5644 0.996 5.631 8 0.688

*Significant when alpha was set at < 0.05.


18/26




324

In the second discriminant function,the mean discriminant score (centroid)

for Group 1, never/do not plan to,was 0.583 and the meandiscriminant score (centroid) for Group5, donated/will increase, was 0.598.The distance between these twocentroids was the greatest, and thesecond discriminant function was thedimension that explains the differencebetween Group 1 and Group 5. Sincethe third and fourth discriminantfunctions were not statisticallysignificant, they were not analyzed for

centroid effect.

Assessment of the significance of eachpredictor variableThe interest was in identifyingsignificant predictor variables ofalumni donation. The identification ofsignificant predictors was a two-stepprocess. In the first step, the analysisof Variance (ANOVA) technique wasused to explore the significance of eachpredictor variable at differentiating

different alumni groups. The ANOVAtechnique identified seven predictorvariables as significant predictors ofalumni donation.

In the second step, seven significantpredictor variables identified in thefirst step were analyzed by

discriminant analysis to determinewhich predictor variables had the

highest correlation with the significantdiscriminant functions. Discriminantanalysis already identified the first twodiscriminant functions as significantfunctions. Among the seven significantpredictor variables identified in theANOVA tests, only those predictorsthat had the highest correlations withthe first or the second discriminantfunctions were identified as significantpredictors.

ANOVA testThe result of ANOVA test wasrevealed in the Table of Equality ofGroup Means (Table 9). This tableassessed a variables potential indifferentiating alumni groups. Each testdisplayed the result of a one-wayANOVA for the predictor variableusing the criterion variable as thefactor. If the significance value wasgreater than ap-value of 0.05, the

variable probably did not contributeto the model.

Graduation year, gender, alumniexperience, alumni motivation, studentexperience-impact on career, andstudent experience-relationships weresignificant at the 0.00 level. Student

Table 8: Function at group centroids

Best describe your financialparticipation

Function

1 2 3 4

1. Never/do not plan to 0.534 0.583a 0.086 0.0952. Donated/wont continue 0.032 0.422 0.095 0.1773. Never/but plan to 0.619b 0.247 0.049 0.0184. Donated/will continue 0.371b 0.034 0.043 0.0125. Donated/will increase 0.274 0.598a 0.218 0.027

aGroup 1 and Group 5 maximally separated at Function 2.bGroup 3 and Group 4 maximally separated at Function 1.


19/26




325

experience-extracurricular activitiesand ethnicity were significant at the

p= 0.05 level. Types of degree,membership status, and in or out ofstate status are not significant at the

p= 0.05 level.Wilks Lambda for predictors

(Table 7) is another measure of apredictors potential. Smaller values

indicate that the variable is better atdiscriminating different groups. Theanalysis suggests that graduation yearis the best predictor of the model(= 0.856), followed by alumniexperience, alumni motivation, studentexperience-impact on career, studentexperience-relationships, and gender.

Discriminant analysis of significantpredictor variablesIn this step, seven significant predictor

variables identified by ANOVA werefurther analyzed by discriminantanalysis. The structure matrix(Table 10) provided the correlation ofeach predictor variable with each ofthe discriminant functions. When morethan one discriminant function exists,

letter a in the table marks eachvariables largest absolute correlationwith one of the discriminant functions.Within each function, these markedvariables are then ordered by the sizeof the correlation.

Among the seven significantpredictors identified by ANOVA,

graduation year is most strongly

correlated with the first function,followed by thegender variable.These two variables are the only twosignificant variables with the firstfunction. Thus, the differencesbetween Group 3 never/but plan toand Group 4 donated/will continueare mainly accounted for by twovariables: graduation year andgender.

Among the seven significantpredictor variables, alumni motivation,alumni experience, student experience-

relationships, and student experience-extracurricular activities have thehighest correlations with the seconddiscriminant function. As statedpreviously, the second discriminantfunction mainly explained thedifferences between Group 1

Table 9: Tests of equality of group means

WilksLambda

F d.f.1 d.f.2 Sig.

F1 Alumni experience 0.969 12.878 4 1603 0.000**F2 Alumni motivation 0.965 14.326 4 1603 0.000**F3 Student experience-impact on career 0.976 10.037 4 1603 0.000**F4 Student experience-relationships 0.986 5.813 4 1603 0.000**F5 Student experience-extracurricular activities 0.994 2.598 4 1603 0.035*Graduation year 0.856*** 67.206 4 1603 0.000**Type of degree 0.996 1.568 4 1603 0.180Gender 0.984 6.454 4 1603 0.000**Ethnicity 0.992 3.221 4 1603 0.012*Membership status by respondent 0.998 0.977 4 1603 0.419In or out of state from respondent filing 0.997 1.372 4 1603 0.241

*Significant at 0.05 level.

**Significant at 0.00 level.***Lowest value.


20/26




326

never/do not plan to and Group 5donated/will increase. Thedifferences between these twogroups were mostly explained byalumni experience, student experience-relationships, student experience-extracurricular activities, and alumnimotivation.

Student experience-impact on careerand ethnicity were closely related to

the third discriminant function, whiletypes of degree and in/out of statestatus are closely related to the fourthfunction. But these two are almostuseless functions since they do notexplain much of the differences amongthe five alumni groups. Therefore, atthe end of discriminant analysis, sixpredictor variables were identified assignificant predictors of alumnidonation: Graduation year, gender,alumni experience, alumni motivation,

student experience-extracurricularactivities, and student experience-relationships.

Prediction using discriminant analysisDiscriminant analysis is also used forprediction and classification. The

classification table (Table 4) shows thepractical results of using thediscriminant model. Discriminantanalysis used this model to classifyalumni into different groups.

As the Original% section in Table4 indicated, of the cases used to createthe discriminant analysis, 723 out ofthe 863 alumni (83.8 percent) whocurrently donate and plan to continue

were classified correctly. More than 80percent (80.4 percent) of the alumniwho currently donate and plan toincrease were classified correctly.Although overall only 56.3 percent ofthe cases were classified correctly, it ismostly because this model did notclassify nondonors very well. Foralumni donors who plan to continuedonation or who plan to increasedonation, 83.8 and 80.4 percent wereclassified correctly.

Classifications based on the casesused to create the model tend to betoo optimistic in that theirclassification rate is inflated(Mickelson, 2002). The cross-validatedsection of Table 4 attempted to correctthis. In cross-validation, each case was

Table 10: Structure matrix

Predictors Function

1 2 3 4

Graduation year 0.863a 0.461 0.106 0.039Gender 0.278a 0.072 0.195 0.022Alumni motivation 0.275 0.466a 0.047 0.357Student experience-relationships 0.008 0.386a 0.258 0.211Student experience-extracurricular activities 0.002 0.261a 0.148 0.005Ethnicity 0.173 0.017 0.533 0.204Student experience-impact on career 0.261 0.327 0.428a 0.289Type of degree 0.071 0.003 0.457 0.653aAlumni experience 0.296 0.388a 0.177 0.389aIn or out of state 0.12 0.044 0.023 0.332Membership status 0.069 0.107 0.105 0.315

aLargest absolute correlation between each variable and any discriminant function.


21/26




327

classified by the discriminant functionderived from all cases other than thatcase. That is, each case was taken out

of the discriminant analysis and therest of the cases were used to classifythis case into alumni groups. As theCross-validated% section suggested,717 out of the 863 alumni (83.1percent) who currently donate andplan to continue were classifiedcorrectly. More than 80 percent (81.2percent) of the alumni who currentlydonate and plan to increase wereclassified correctly. The classificationtable and cross-validation tables show

the overall reliability of thisdiscriminant analysis in predictingalumni donors.

FindingsThe studys results supported three ofthe four proposed hypotheses. Thediscriminant analysis supportedhypotheses 1, 2, and 3; however, theanalysis only partially supportedhypothesis 4. That is, while graduation

year andgender are significantpredictor variables of alumni donation,not all demographic variables weredetermined to be significant predictors.

The results confirmed the firstresearch hypothesis: Student experiencesignificantly influences alumnidonations. The results indicated thatsatisfaction was greater among alumniwho believed that the universitycontributed to their education, thestudent experience-impact on career

factor. Satisfaction was also greater forthose alumni who had developedrelationships with university facultyand staff during their educationalexperiences. If alumni were satisfiedwith their previous studentexperiences, they were more inclined

to give. This finding was consistentwith Shadoians (1989) and Oglesbys(1991) results in which they found that

a significant difference existed betweendonors and nondonors on thepredictor student experience.

The positive effect of alumniexperience on alumni donationconfirmed the second researchhypothesis: alumni experiencesignificantly distinguishes alumnidonors from nondonors. Since alumniexperiences are closely related toalumni marketing efforts such asparties, reunions, newsletters, and

solicitations, the results confirm thatthese efforts do engage alumni, andthat alumni may be more likely todonate than those less or not engaged.Miracles (1977) and Grills (1988)earlier results confirm the studysfindings. Taylor and Martin (1995)also reached similar findings. Alumniinvolvement with the university was asignificant predictor in the discriminantanalysis models applied.

The studys findings supported the

third hypothesis: alumni motivationsignificantly distinguishes alumnidonors from nondonors. Theimportant contribution of alumnimotivation, or internal desires rooteddeeply enough to induce a commitmentto the alma mater, was confirmed bythe study. The findings also suggestedthat alumni who were more informedabout the university had more positiveperceptions of it, were more awareof and linked with perceived

institutional needs, and, therefore,were more likely to give than thosenot well informed.

The studys findings only partiallysupported the fourth hypothesis:Demographic variables (graduationyear, gender, ethnicity, type of degree,


22/26




328

in or out of state, and membershipstatus) significantly distinguish alumnidonors from nondonors. The

demographic variables of graduationyear and gender (p= 0.00) were themost significant distinctions betweendonors and nondonors. The ethnicitydemographic variable (p= >0.05) was amore significant factor than the degreetypes, in or out of state, andmembership status demographicvariables (p= 0.05) distinguishingdonors from nondonors.

The findings suggest that thedemographic variablegraduation year,

a proxy for alumni age, was the mostsignificant predictor of alumnidonation. Previous researchers (Beeler,1982; Lindahl and Winship, 1992;Okunade et al., 1994) also foundsupport for similar variables. Olsenet al. (1989) found that the growthrate of donation coincided with theageincome profile of donors. Theannual fund-raising model at theNorthwestern University (Okunadeand Berl, 1997) found that the year

after graduation was a significantpredictor of alumni giving.

Critical to this analysis was theWilks Lambda coefficient. Graduationyear had the lowest Wilks Lambdacoefficient (= 0.856) in the test ofequality of group means, whichsuggested that the graduation yearvariable had the greatest predictivepower in the studys model. Thisfinding also suggested a moreprominent role for age as a proxy to

graduation year than prior researchrevealed. Moreover, it supported thegeneralization that older alumni havehigher net worth and larger capacityfor charitable giving. Younger alumni,on the other hand, with less incomeand possible student loan debt, may be

less generous to their alma maters thanthey otherwise might be.

ImplicationsBased on the studys findings, highereducation administrators and alumnifundraisers can begin taking stepsindividually and collaboratively toimprove alumni fund-raising results.The key to fund-raising success,however, will be contingent onincreasing collaborative, not isolated,efforts.

Comprehensive communicationstrategyA comprehensive communicationstrategy targeting past, current, andfuture students as potential donors isparamount in increasing alumnidonations. Administrators should leadthe way of keeping the horizon insight while focusing on institutionalneeds today. The studys results suggestthat the institution would benefit fromthis approach (Pearson, 1999).

In order to implement a multiplefocus vision of fund raising,administrators will benefit frominviting to the table important internalstakeholders, especially alumnifundraisers, in designing and executingtargeted communication efforts. Awell-articulated communication planwill create more focused, interlocked,and cost-effective endeavors to increasealumni gift giving.

One outcome of successfully

implementing a comprehensivecommunication strategy to expandalumni giving is budget increases forthe alumni association and other units.Administrators should be open to thisnecessary change. The studys resultssuggest a need for higher education


23/26




329

institutions to evaluate the currentexpenditure priorities. Historical trendsindicate that institutions have increased

expenditures in areas that do not havea significant impact on studentexperience and alumni satisfaction(Clotfelter, 1999). The benefits ofincreased future donations will faroutweigh the cost of todays redirectedand refocused expenditures.

Current future fundersAdministrators may want to considerconcentrating on the current

undergraduate students as futurefunders for the institution. Providingthese students with quality, well-rounded educational and extra-curricular experiences will benefit theinstitutions in the end since themajority of alumni donations comefrom donors who give to theirundergraduate alma mater (Council forAid to Education, 2007).

In developing quality studentexperiences, administrators need to

include alumni fundraisers indiscussions on how best to enhanceacademic improvement, student extra-curricular activities, and careercounseling. Prioritized expenditures inthese improvements will ensureenhanced student experiences whileselected strategies will have the biggestimpact not only on student experiencesbut also on cultivating donations forthe future. The findings of this studysuggest, and supported by similar

recommendations from Oglesby(1991), that fund-raising efforts can bemore successful when student affairs,the alumni association, universitycommunications, and foundations arecollaborating and working closely withone another.

Relationship-building effortsAdministrators can create a culture ofvaluing relationships between faculty

and students during their time at theinstitution and after. Provideopportunities for faculty to mentorstudents beyond graduation. Supportfaculty and departments in reachingout to graduates by inviting their inputon academic program improvements.Encourage departments to keep intouch with graduates through printand electronic communications. Offertechnological assistance in maintainingand updating department websites

listing graduates professionalaccomplishments beyond graduation.Most importantly, provide departmentsthe additional resources to implementthese endeavors. Administrators whoestablish relationship building withstudents as a priority throughout theinstitution may see future financialbenefits of increased donations.

Enhanced alumni services

The findings of the study confirmprevious research. Satisfied alumni aremore likely to remain engaged with theuniversity than those not satisfied.Engaged alumni are more likely toremain informed about the universitythan those not engaged. Informedalumni are more likely to understandthe needs of the university and may bemore likely to give than those lessinformed (Pearson, 1999). Alumnifundraisers who concentrate on

improving the alumni experience,which most likely will influence alumnimotivation, are more successful.

It may be fruitful for alumnifundraisers to evaluate current alumniprograms to determine costeffectiveness and return on efforts.


24/26




330

Secondly, fundraisers might considerhosting focus groups, conductingsurveys, and soliciting input from

alumni stakeholders on identifyingtheir needs and satisfaction. If alumniare satisfied with alumni services, theyare more inclined to give than thosenot satisfied (Pearson, 1999).

Demographic variables to pursueThe most significant predictor ofalumni donations, according to thestudys findings, is graduation year,with age as a proxy. This may signal

an increased focus on older alumni.The baby-boom generation is cominginto their own acquired wealth frominvestments or through inheritances.Typical of that generation, they arestill looking for ways to impact society.It may seem counter-intuitive not tospend as much effort on newlygraduated alumni; however, olderalumni are established professionallyand more financially secure. Alumnifundraisers would likely realize

increased donations if they expandedand redirected solicitation budgets totap into older alumni. Based on thisstudy, they are more likely to donatethan any other alumni group.

Another significant demographicvariable to predict alumni donations isgender, which is noteworthy for futurefund-raising strategies. Women tend todonate more often than men do,according to the studys findings.Coupled with the graduation year

factor, the largest potential source forsignificant giving are older women.Women live longer than men live andinherit 70 percent of all estates. Theyare expected to own half the wealth inthe United States by 2010 (Strout,2007b). Alumni fundraisers should

take heed of these statistics andredirect their alumni focus. Iowa StateUniversity did and the results look

promising. Since 2000, the number ofIowa State female donors has increasedby 37 percent because of Iowa Statesredirected efforts and the amount ofmoney donated increased 138 percent(Strout, 2007b). Alumni associationswith limited budgets could focuson alumni age and gender, coupledwith enhanced alumni services, andsee significantly increased donations asa result.

ConclusionThe studys findings may interest fund-raising officials in the efficient choiceof prospective donors. Alumni-specificdata collected through the survey canbe used with this model to predictindividuals propensity to give. Alumnifundraisers could use this model topredict the likelihood of giving foreach alumnus/a. For this study, themodel predicted donors correctly 81

percent of the time. Fundraisers maywant to use this model to predict thepropensity to give and spend moreresources on those alumni more likelyto give. A major factor is that thesurvey instrument is reliable andreadily available.

More research to fully test thealumni donation model proposed andthe contribution of each significantpredictor is needed. The criterionvariable in this study was classified as

a categorical variable, and as such, canonly predict the membership of donorand nondonor status. If donationamount was available as a continuousvariable, it could serve as a predictorvariable in a multiple-regressionanalysis. Such an analysis would


25/26




331

directly assess the contribution of eachpredictor to that contribution, whichwould provide more comprehensive

information than can the analysisrequired in this study.The model introduced in this study

may provide an alumni association abasis for investigating more complexlinkages between alumni and theirdonation behaviors. Such researchmay require greater integration of alarger database, including alumniincome level and alumni donationamount. These issues, alongwith accompanying research and

methodological issues, pose a challengeto improving a deeper understandingof the relationship between alumniand their donations.

Note1This manuscript is based on X. Sun(2005), A multivariate causal modelof alumni giving at a Midwest publicuniversity, unpublished doctoraldissertation, University of Nebraska-Lincoln.

ReferencesAndreoni, J. (1989), Giving with impure

altruism: Applications to charity and Ricardianequivalence, Journal of Political Economy, 97,pp. 14471458.

Andreoni, J. (1990), Impure altruism anddonations to public goods: A theory of warm-glow giving, The Economic Journal, 100,pp. 464467.

Bar-Tal, D. (1976), Prosocial Behavior: Theory andResearch, Halsted Press, New York.

Becker, G.S. (1974), A theory of social

interaction,Journal of Political Economy,82, pp. 10631093.

Beeler, K.J. (1982), A study of predictors onalumni philanthropy in private universities,unpublished doctoral dissertation, University ofConnecticut.

Belfield, C.R. and Beney, A.P. (2000), Whatdetermines alumni generosity? Evidence for theUK, Education Economics, 18, pp. 6580.

Blumer, H. (1969), Symbolic Interactionism:Perspective and Method, Prentice-Hall,Englewood Cliffs, NJ.

Bristol Jr., R. (1990), The life cycle of alumni

donations, The Review of Higher Education,13, pp. 503518.

Bruggink, T.H. and Siddiqui, K. (1995), Aneconometric model of alumni giving: A casestudy for a liberal arts college, The AmericanEconomist, 39, pp. 5361.

Burt Jr., W.W. (1989), Alumni donors and non-donors of Southeast Missouri State Universityaccording to institutional involvement, educationalsatisfaction and demographic characteristics,unpublished doctoral dissertation, University ofCalifornia, Los Angeles.

Clotfelter, C.T. (1999), The familiar but curiouseconomics of higher education: Introduction to asymposium,Journal of Economics Perspective,

13, pp. 312.Clotfelter, C.T. (2003), Alumni giving to elite

private colleges and universities, Economics ofEducation Review, 22, pp. 109120.

Coelho, P.R.P. (1985), An examination intothe causes of economic growth: Status asan economic good, Research in Law andEconomics, 7, pp. 89116.

Council for Aid to Education (2007), Voluntarysupport for education. [Internet], availablefrom www.cae.org (Accessed October 20, 2007).

Foa, U.G. and Foa, E.B. (1974), Societal Structuresof the Minds, Thomas, Springfield, IL.

Gardner, P.M. (1975), A study of the attitudes ofHarding College alumni with an emphasis on

donor and non-donor characteristics, unpublisheddoctoral dissertation, Ohio University.

Graham, A.E. and Husted, T.A. (1993),Understanding state variations in SAT scores,Economics of Education Review, 12,pp. 197202.

Grill, A.J. (1988), An analysis of the relationshipsof selected variables to financial support providedby alumni of a public university, unpublisheddoctoral dissertation, Ohio University.

Grimes, P.W. and Chressanthis, G.A. (1994),Alumni contributions to academics: The roleof intercollegiate sports and NCAA sanctions,American Journal of Economics and Sociology,53, pp. 2740.

Halfpenny, P. (1999), Economic and sociologicaltheories of individual charitable giving:Complementary or contradictory, Voluntas:International Journal of Voluntary andNonprofit Organizations, 10, pp. 197215.

Harrison, W.B. (1995), College relations and fund-raising expenditures: Influencing the probabilityof alumni giving to higher education, Economicsof Education Review, 14, pp. 7384.


26/26


INTERNATIONAL JOURNAL OF EDUCATIONAL ADVANCEMENT VOL 7 NO 4 307332332

Hueston, F.R. (1992), Predicting alumni giving: Adonor analysis test, Fund Raising Management,

July, pp. 1922.

Kleinginna, P. and Kleinginna, A. (1981),A categorized list of motivation definitions,with suggestions for a consensual definition,Motivation and Emotion, 5, pp. 263291.

Leslie, L.L., Drachman, S.S., Ramey, G.W. andConrad, C.F. (1983), Factors accountingfor variations over time in voluntary supportfor colleges and universities, Journal ofEducational Finance, 9, pp. 213215.

Lindahl, W.E. and Winship, C. (1992), Predictivemodels for annual fundraising and major giftfundraising, Nonprofit Management andLeadership, 3, pp. 364.

Mael, F. and Ashforth, B.E. (1992), Alumniand their alma mater: A partial test ofthe reformulated model of organizationalidentification,Journal of Organizational

Behavior, 13, pp. 103123.Mickelson, J. (2002), Multivariate statistics,

unpublished manuscript, University of Nebraska-Lincoln.

Miracle, W.D. (1977), Differences between giversand non-givers to the University of Georgiaannual fund, unpublished doctoral dissertation,University of Georgia.

Oglesby, R.A. (1991), Age, student involvement,and other characteristics of alumni donorsand alumni non-donors of Southwest BaptistUniversity, unpublished doctoral dissertation,University of Missouri.

Olsen, K., Smith, A.L. and Wunnava, P.V. (1989),An empirical study of the life-cycle hypothesis

with respect to alumni donations, AmericanEconomists, 33, pp. 6063.

Okunade, A.A. and Berl, R.L. (1997),Determinants of charitable giving of businessschool alumni, Research in Higher Education,38, pp. 201214.

Okunade, A.A., Wunnava, P.V. and Walsh, J.R.(1994), Charitable giving of alumni: Microdataevidence from a large public university,American Journal of Economics and Sociology,53, pp. 7384.

Pearson, J. (1999), Comprehensive research onalumni relationships: Four years of market

research at Stanford University, New Directionsfor Institutional Research, 101, pp. 521.

Shadoian, H.L. (1989), A study of predictors ofalumni in public colleges, unpublished doctoraldissertation, University of Connecticut.

Steinberg, R. (1997), Overall evaluation ofeconomic theories, Voluntas: International

Journal of Voluntary and NonprofitOrganizations, 8, pp. 179204.

Stevens, J.P. (2002), Applied Multivariate Statisticsfor the Social Sciences, Lawrence ErlbaumAssociates, Mahwah, NJ.

Strout, E. (2007a), Donations increase for 3rdyear in a row, The Chronicle in HigherEducation, [Internet]. March 2, p. A1, availablefrom: www.chronicle.com (accessed October 23,2007).

Strout, E. (2007b), Courting female doners, TheChronicle of Higher Education, [Internet], July 6,pp. A21, available from www.chronicle.com

(accessed October 23, 2007).Sugden, R. (1984), Reciprocity: The supply of

public goods through voluntary contribution,Economic Journal, 94, pp. 772787.

Taylor, A.L. and Martin Jr., J.C. (1995),Characteristics of alumni donors andnondonors at a research I, pubic university,Research in Higher Education, 36,pp. 283298.

Thibaut, J.W. and Kelley, H.H. (1959), The SocialPsychology of Groups, John Wiley & Sons,New York.

Walster, E. and Walster, G.W. (1978), Equity:Theory and Research, Allyn and Bacon, Boston.

Weintraub, E.R. (1985), General Equilibrium Theory,Cambridge University Press, Cambridge, MA.

Winston, G. (1999), For-profit education: Godzillaor Chicken Little, Change Magazine, 31,pp. 1319.

Wolverton, B. (2007), Growth in sports gifts maymean fewer academic donations, The Chronicleof Higher Education [Internet], October 5,p. A1, available from: www.chronicle.com(accessed October 23, 2007).

Yoo, Y.H. and Harrison, W.B. (1989), Altruismin the market for giving and receiving: A caseof higher education, Economics of EducationReview, 8, pp. 367379.

Alumni Giving Implications

Documents

Transcript of Alumni Giving Implications