The Reading Teacher Vol. 56, No. 6 March 2003 ©2003 International Reading Association (pp. 554–564)

eisha is a quiet little girl with tawnyskin, long black hair, and large browneyes. Her teacher has observed that

she is shy with other children, reluctant to par-ticipate in groups, and often the last to join in ac-tivities, but the teacher never has had toreprimand Keisha. Only when the teacher satdown to fill out reports on the language devel-opment of each of her students did she realizeKeisha has a problem.

CJ is an energetic almost 5-year-old AfricanAmerican boy who attends childcare from 7:00a.m. to 4:00 p.m. five days a week. His teacherobserved that “CJ has demonstrated strong useof his oral language skills since entering my pro-gram. CJ speaks clearly and is easily understoodby adults.” The teacher was enrolled in a col-lege course that focused on enriching children’slanguage and literacy skills. She noticed aftercompleting the TROLL assessment of CJ that hedid not always choose to attend to stories inlarge group and appeared most interested in sto-ries when reading one-on-one or in a small

This rating system measures skillscritical to the New Standards forSpeaking and Listening. TROLLcan be used to track children’s

progress in language and literacydevelopment, to inform

curriculum, and to stimulatefocused communication between

parents and teachers.

group of no more than three other childrenwhere he was able to ask questions as a way toclarify the meaning of a story. She suggested tohis mother that his older brothers be encouragedto read to him at home. She also noted that hisscores on the language and reading subtotals ofthe TROLL were high—higher than his writingsubtotal—and she was knowledgeable about hisexact writing skills and limitations.

Reading, writing, and orallanguage: Roots of literacy

Early reading and writing abilities are bynow well-known dimensions of early literacy.Through their preschool years, children progres-sively construct understandings of writing (e.g.,Bissex, 1980) and reading (e.g., Sulzby & Teale,1991). Similarly, the contribution of children’sphonological awareness has often been exploredand is also widely recognized (Bryant,MacLean, & Bradley, 1990; Cronin & Carver,1998; Speece, Roth, Cooper, & de la Paz, 1999;Stanovich, 1992; Vellutino & Scanlon, 2001;

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 554 (Black plate)

554

K

Teacher Rating of OralLanguage andLiteracy (TROLL):Individualizingearly literacyinstruction with astandards-basedrating tool

David K. DickinsonAllyssa McCabeKim Sprague

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 554 (PANTONE 300 CV plate)

Wagner, Torgesen, Laughon, Simmons, &Rashotte, 1993; Wagner et al., 1997).

However, there are other lesser known orallanguage skills relevant to literacy that include thedevelopment of narrative ability (Dickinson &Tabors, 2001; see McCabe & Rollins, 1991, forreview), use of talk while pretending (Dickinson,2001; see Pelligrini & Galda, 1993, for review),and varied vocabulary usage (Tabors, Beals, &Weizman, 2001). To be able to read and write ef-fectively, children must develop strong oral lan-guage skills (Dickinson & McCabe, 1991;Dickinson & Tabors, 2001; Snow, 1983; Snow,Burns, & Griffin, 1998). Oral language skillsblossom during the preschool years, but they arealso very vulnerable and in need of stimulationduring this time, as a number of major organiza-tions involved in the education of young childrenhave recognized.

A call for developmentallyappropriate assessment

The International Reading Association(IRA) and the National Association for theEducation of Young Children (NAEYC) jointlyformulated a position statement regarding earlyliteracy development (1998). The statement ac-knowledges the difficulty that teachers face, forexample, in kindergarten classrooms where afive-year range in children’s literacy skills is notuncommon (Riley, 1996). Estimating whereeach child is in terms of the acquisition of speak-ing, listening, reading, and writing skills is criti-cal to providing developmentally appropriateinstruction to all children in this wide range. Theposition statement is quite clear that

throughout these critical years accurate assessment of chil-dren’s knowledge, skills, and dispositions in reading and writ-ing will help teachers better match instruction with how andwhat children are learning. However, early reading and writ-ing cannot simply be measured as a set of narrowly definedskills on standardized tests. These measures often are not re-liable or valid indicators of what children can do in typicalpractice, nor are they sensitive to language variation, culture,or the experiences of young children. (International ReadingAssociation & National Association for the Education ofYoung Children, 1998, p. 38; emphasis added)

As if these difficulties were not enough,preschool teachers face time constraints and typ-ically have not been trained to evaluate chil-dren’s language development as it relates to the

acquisition of literacy or their emergent readingand writing skills. In response to this need, wepresent an accessible means of evaluating eachchild in a classroom for literacy-related abili-ties. One way to help teachers track children’sdevelopment is by periodic reflection on demon-strations of early literacy. Even though teachersmay lack prior formal training regarding assess-ment of language and literacy development, wehave found that they can recognize critical as-pects of this development. Using the TeacherRating of Oral Language and Literacy (TROLL)to evaluate children can help teachers assess theeffectiveness of an educational program.

New Standards: Speaking andlistening for preschool through third grade

Because speaking and listening are so criti-cal for literacy development in early childhood,the New Standards project, a program of theNational Center on Education and the Economy(Tucker & Codding, 1998), has developedresearch-based standards for speaking and lis-tening in preschool through third grade (NewStandards Speaking and Listening Committee,2001). These standards complement those al-ready developed for reading and writing. Thespeaking and listening standards include specif-ic recommendations for teachers regarding ben-eficial habits of conversation, useful kinds oftalk such as narratives and explanations, and lan-guage conventions relevant to early childhood.Many programs are concerned with ensuringthat their students meet such national standards.

Development of TROLLTo guide teachers’observations of children’s

individual language and literacy skills and in-terests, we created an instrument that focuses onkey abilities. In constructing this tool, we drewon the same body of theory and research that wasthe foundation for the speaking and listeningstandards. Indeed, we contributed to both efforts.TROLL is a tool developed by the first authorto provide teachers with a way to track the lan-guage and literacy skills and interests of childrenin their classrooms and is represented in its en-tirety in Table 1. Although it was developed forresearch purposes separate from the NewStandards initiative, TROLL addresses all the

Teacher Rating of Oral Language and Literacy (TROLL)

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 555 (Black plate)

555

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 555 (PANTONE 300 CV plate)

central speaking and listening skills in the NewStandards, as well as many of the early readingand writing skills covered by the companionreading and writing standards. Of special interestis the fact that it allows teachers to track chil-dren’s interests in various language and literacyactivities—something that no direct assessmenttool can capture.

Using TROLL to inform instructionNo formal training is required to use the

TROLL instrument; however, it is most effectiveif teachers know about language and literacy de-velopment. In other words, TROLL can makeknowledgeable teachers better. This tool is de-signed for classroom teachers to easily track thelanguage and literacy development of all their stu-dents. The TROLL requires only 5 to 10 minutesfor each child and need not disrupt classroom ac-tivities (it can be completed during naptime).

Teachers can use the information to informtheir teaching by identifying (a) children who aredisplaying evidence of serious delay and who mayneed formal assessment by audiologists or speech-language pathologists and (b) children who areshowing high levels of literacy development andtherefore need special additional challenges in thisarea. Furthermore, by completing TROLL severaltimes over the course of a year, teachers can trackthe progress of all their students.

Second, teachers could combine results forall the children in their class to determine whichareas need more systematic instruction. For ex-ample, if all children in a class score relativelylow on rhymes, their teacher might want to beginproviding numerous opportunities to listen toand produce rhyming chants, songs, and poems.

Teachers should consider using TROLL rat-ings of children as a basis for discussions withparents. In fact, as we saw in the case of CJ at theoutset of this article, teachers are likely to initiatesuch discussions without anyone urging them todo so. (Recall that the teacher recommended thatCJ’s mother get his older brothers to read withhim one-on-one because that was an effectivesetting for the child.)

Parents can also serve as a source for rat-ings using the TROLL. In particular, teachers ofbilingual children often have a difficult time rat-ing the language competence of children whospeak English as a second language (ESL). Of

course, if a teacher can rate a child’s competencein a language other than English, it would be ofgreat interest to rate the child’s skill in bothEnglish and his or her first language. Wheneverpossible, educators need to involve the parents ofESL students. In fact, maternal reports ofpreschoolers’ literacy (when children were 3 or 4years old) significantly predicted much of thevariation in kindergarten tests, grade 1 teacherassessments, and direct assessments of decod-ing given near the end of first grade (Dickinson& DeTemple, 1998). Thus, if TROLL were usedcollaboratively with parents, it could provide apowerful way to organize a multifaceted conver-sation about a child’s full range of language andliteracy development.

TROLL has been used extensivelyOver the last several years, TROLL has been

used with 973 children in the context of researchexamining early literacy development. Over 100teachers have been involved in this process.

One measure of a good test is that all itemson the test tap related abilities. We analyzed re-sponses for 534 of these preschool children andfound strong indications that TROLL meetsstandards expected of research tools in this re-gard. Specifically, Cronbach’s alpha estimates ofinternal consistency ranged from .77 to .92 forseparate subscales, indicating strong internalconsistency. For the total TROLL scores, alphasexceeded .89 for each age.

Another way of determining the value of atool is the extent to which a child’s performanceon that tool compares to performance on othermeasures. After all, TROLL relies on a teacher’sprofessional judgement or perception of a child’sdevelopment rather than formal testing of actualdevelopment. It is therefore reassuring to findthat, for this sample, the ratings teachers provid-ed using TROLL compared favorably to formalassessments by researchers. These measures in-cluded the well-established Peabody PictureVocabulary Test (PPVT-III), which is a measureof receptive vocabulary, as well as measures ofemergent literacy and early phonological aware-ness. Teacher ratings of children’s language andliteracy development on the TROLL show mod-erate associations with children’s scores on allthree of those direct assessments despite the factthat teachers never saw those test results. So in

The Reading Teacher Vol. 56, No. 6 March 2003

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 556 (Black plate)

556

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 556 (PANTONE 300 CV plate)

about five minutes, and with no special trainingon the TROLL, teachers themselves can indexwhat specially trained researchers would spend25–30 minutes per child assessing.

Of course, the TROLL teacher ratings donot agree completely with the researchers’ tests.

This difference partly reflects the fact thatTROLL assesses other factors that teachers takeinto consideration as they rate individualchildren—factors that are not captured in thedirect, formal assessments. TROLL captures thekind of information the position statement by

Teacher Rating of Oral Language and Literacy (TROLL)

Table 1 Teacher Rating of Oral Language and Literacy (TROLL)

Language use1. How would you describe this child’s willingness to start a conversation with adults and peers and

continue trying to communicate when he or she is not understood on the first attempt? Select the statementthat best describes how hard the child works to be understood by others.

2. How well does the child communicate personal experiences in a clear and logical way? Assign thescore that best describes this child when he or she is attempting to tell an adult about events that hap-pened at home or some other place where you were not present.

3. How would you describe this child’s pattern of asking questions about topics that interest him or her(e.g., why things happen, why people act the way they do)? Assign the score that best describes thechild’s approach to displaying curiosity by asking adults questions.

Child almost neverbegins a conversa-tion with peers or theteacher and neverkeeps trying if un-successful at first.

Child sometimes be-gins conversation witheither peers or theteacher. If initial ef-forts fail he or she of-ten gives up quickly.

Child begins conver-sations with both peersand teachers on occa-sion. If initial efforts fail,he or she will some-times keep trying.

Child begins conver-sations with bothpeers and teachers.If initial efforts fail, heor she will work hardto be understood.

1 2 3 4

Child is very tenta-tive, only offers a fewwords, requires youto ask questions, hasdifficulty respondingto questions you ask.

Child offers some in-formation, but infor-mation needed toreally understand theevent is missing (e.g.,where or when it hap-pened, who was pres-ent, the sequence ofwhat happened).

Child offers informa-tion and sometimesincludes the neces-sary information tounderstand the eventfully.

Child freely offers in-formation and tellsexperiences in a waythat is nearly alwayscomplete, well se-quenced, and com-prehensible.

1 2 3 4

To your knowledge,the child has neverasked an adult aquestion reflectingcuriosity about whythings happen orwhy people dothings.

On a few occasionsthe child has askedadults some ques-tions. The discussionthat resulted wasbrief and limited indepth.

On several occasionsthe child has askedinteresting questions.On occasion thesehave lead to an inter-esting conversation.

Child often asksadults questions re-flecting curiosity.These often lead tointeresting, extend-ed conversations.

1 2 3 4

(continued)

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 557 (Black plate)

557

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 557 (PANTONE 300 CV plate)

The Reading Teacher Vol. 56, No. 6 March 2003

4. How would you describe this child’s use of talk while pretending in the house area or when playing withblocks? Consider the child’s use of talk with peers to start pretending and to carry it out. Assign the scorethat best applies.

5. How would you describe the child’s ability to recognize and produce rhymes?

6. How often does child use a varied vocabulary or try out new words (e.g., heard in stories or from teacher)?

7. When child speaks to adults other than you or the teaching assistant, is he or she understandable?

8. How often does child express curiosity about how and why things happen?

Language subtotal ___________

Reading9. How often does child like to hear books read in the full group?

Child cannot eversay if two wordsrhyme and cannotproduce a rhymewhen given exam-ples (e.g., rat, cat).

Child occasionallyproduces or identi-fies rhymes whengiven help.

Child spontaneouslyproduces rhymes andcan sometimes tellwhen word pairsrhyme.

Child spontaneouslyrhymes words ofmore than one sylla-ble and always iden-tifies whether wordsrhyme.

1 2 3 4

Never Rarely Sometimes Often1 2 3 4

Never Rarely Sometimes Often1 2 3 4

Never Rarely Sometimes Often1 2 3 4

Never Rarely Sometimes Often1 2 3 4

Child rarely or neverengages in pretendplay or else nevertalks while pretend-ing.

On occasion thechild engages in pre-tending that includessome talk. Talk isbrief, may only beused when startingthe play, and is oflimited importance tothe ongoing play ac-tivity.

Child engages in pre-tending often andconversations aresometimes importantto the play. On occa-sion child engages insome back-and-forthpretend dialogue withanother child.

Child often talks inelaborate ways whilepretending.Conversations thatare carried out “inrole” are common andare an important partof the play. Childsometimes steps outof pretend play to givedirections to another.

1 2 3 4

Table 1 (continued)Teacher Rating of Oral Language and Literacy (TROLL)

(continued)

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 558 (Black plate)

558

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 558 (PANTONE 300 CV plate)

Teacher Rating of Oral Language and Literacy (TROLL)

10. How often does child attend to stories read in the full group or small groups and react in a way thatindicates comprehension?

11. Is child able to read storybooks on his or her own?

12. How often does child remember the story line or characters in books that he or she heard before eitherat home or in class?

13. How often does child look at or read books alone or with friends?

14. Can child recognize letters? (choose one answer)None of the letters of the alphabet ....................................1Some of them (up to 10)....................................................2Most of them (up to 20) ....................................................3All of them..........................................................................4

15. Does child recognize his or her own first name in print?

16. Does child recognize other names?

17. Can child read any other words?

18. Does child have a beginning understanding of the relationship between sounds and letters (e.g., theletter B makes a “buh” sound)?

Never Rarely Sometimes Often1 2 3 4

Does not pretend toread books Pretends to read Pretends to read and

reads some wordsReads the written

words1 2 3 4

Never Rarely Sometimes Often1 2 3 4

Never Rarely Sometimes Often1 2 3 4

No One or two A few (up to four or five)

Several(six or more)

1 2 3 4

No Yes1 2

No One or two A few (up to four or five)

Several(six or more)

1 2 3 4

No One or two A few (up to four or five)

Several(six or more)

1 2 3 4

Table 1 (continued)Teacher Rating of Oral Language and Literacy (TROLL)

(continued)

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 559 (Black plate)

559

The Reading Teacher Vol. 56, No. 6 March 2003

19. Can child sound out words that he or she has not read before?

Reading subtotal ___________Writing

20. What does child’s writing look like?

21. How often does child like to write or pretend to write?

22. Can child write his or her first name, even if some of the letters are backward?

23. Does child write other names or real words?

24. How often does child write signs or labels?

25. Does child write stories, songs, poems, or lists?

Writing subtotal __________ (out of 24 possible)Oral language subtotal __________ (out of 32 possible)Reading subtotal __________ (out of 42 possible)Total TROLL score __________ (out of 98 possible)

Note. Copyright ©1997 Education Development Center. Reproduced with permission.

No Once or twice One syllable wordsoften

Many words

1 2 3 4

Only draws or scribbles

Some letter-likemarks

Many conventionalletters

Conventional lettersand words

1 2 3 4

Never Rarely Sometimes Often1 2 3 4

Never Rarely Sometimes Often1 2 3 4

No One or two A few (up to four or five) Several (six or more)1 2 3 4

Never Rarely Sometimes Often1 2 3 4

Never Rarely Sometimes Often1 2 3 4

Table 1 (continued)Teacher Rating of Oral Language and Literacy (TROLL)

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 560 (Black plate)

560

IRA and NAEYC recommended be captured inassessment. Formal tests measure how well achild does at only one point in time; childrenmay be tired or sick on the day of the PPVT-IIIassessment and receive a dismal score for theirreceptive vocabulary, whereas their teacherknows that on most days they are quick to pickup on the vocabulary of classroom units and ar-ticulate when sharing stories of personal experi-ence. The TROLL score is not as vulnerable tofluctuations in a child’s performance as are theformal tests.

Furthermore, as noted earlier, the TROLLincludes information about the child’s engage-ment in literacy activities and patterns of use oforal language. Formal assessments do not tapsuch interests and inclinations to use languageand print in any way. And yet, a child’s initia-tive in this area could be an important determi-nant of the child’s future success.

Overall, children’s scores improve from fallto spring. However, the correlations betweenTROLL scores and direct assessment measuresof literacy are generally less reliable in the springthan in the fall, which is just the opposite of whatone would predict. After all, teachers have hadfar more interaction with children—and shouldtherefore be more sensitive to their reading andwriting skills—in the spring than in the fall. Infact, however, there is no firm correlation be-tween teacher TROLL scores in the spring andformal assessments of children’s emergent liter-acy conducted at that time, although correlationswith vocabulary (PPVT) and phonologicalawareness (EPAP) remain moderately strong.Unfortunately, this may reflect the fact thatteachers do not revisit their assessment of chil-dren’s literacy skills as much as they should. Inthe fall, teachers may arrive at judgments abouta child’s accomplishments that they fail to up-date. Children’s progress in language and liter-acy may go undetected by their teachers. Thisfinding is sobering.

One danger of any judgment is that it can be-come a self-fulfilling prophecy (Rosenthal &Jacobsen, 1968). Teachers’opinions of children atthe very outset of their education can predict chil-dren’s success just by virtue of the teacher givingextra attention, motivation, or instruction to thosechildren they expect to become the most accom-plished. Conversely, teacher expectations—never revisited—can predict other children’s

failure by virtue of overlooking children expect-ed to fail.

This lack of revision of judgments about lit-eracy skills and interests points to the value ofteachers carrying out periodic informal assess-ments of children to provide concrete evidenceof children’s growth. For example, teachers canask children to write their own or others’ namesor to identify letters in them.

In general, then, rather than contribute toself-fulfilling prophecies of children’s success orfailure, we hope that TROLL will predict possi-ble failure to learn to read in order to preventsuch failure. That is, we hope that teachers willuse this instrument to give struggling childrenthe help they need to succeed so that they neverhave to experience failure at a later point.Specifically, we (Dickinson, 2001) recommendthat teachers make sure that one or more adultsread with small groups of children every day andensure that all children have this experience oncea week. Teachers can also set aside time whenchildren tell stories. Finally, teachers need to in-troduce varied and challenging vocabulary as aroutine part of the curriculum.

What TROLL scores meanTable 2 displays what different scores on

TROLL indicate about a child’s overall develop-mental level. For example, a score of 66 in thespring indicates that the child is making progressthat is average for 4-year-olds in this sample.The sample consisted only of low-income chil-dren, so these scores should be regarded as pro-visional. However, we argue that the well-knownacademic disadvantages of low-socioeconomicstatus (SES) preschool children (e.g., Stipek &Ryan, 1997) make this sample important in itsown right. If a child from a low-SES familyscores at the 10th percentile, for example, this re-sult cannot be dismissed as a result of economicdisadvantage; such a child is scoring very poorlyrelative to his or her economic peers. Scores atthe 75th and, especially, the 90th percentile,however, should prompt a teacher to provide op-portunities for children to read more advancedbooks, engage in writing frequently, and talk atlength about challenging and interesting topics.

Teacher Rating of Oral Language and Literacy (TROLL)

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 561 (Black plate)

561

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 561 (PANTONE 300 CV plate)

Program evaluation potential:TROLL measures appropriatelanguage and literacy instruction

As we mentioned, children’s scores onTROLL might well serve as an impetus to plansystematic language and literacy instruction for aclass. In fact, such instruction has been imple-mented by a number of Head Start programs inthe Boston, Massachusetts, area, where TROLLdetected changes that occurred as a result of pro-gram improvement efforts. Head Start teachersand their supervisors volunteered to participatein a professional development program calledLEEP (Literacy Environment EnrichmentProgram). They received academic credit forparticipating in two intensive three-day blocksseparated by three months. TROLL scores forchildren whose teachers participated in LEEPwere significantly higher than for childrenwhose teachers did not. Specifically, children inLEEP classrooms gained more overall from fall

to spring on average in comparison to a controlgroup.

Furthermore, the classrooms that supportedsuch advances had improved classroom languageand literacy practices. Teachers who participatedin LEEP made greater efforts to engage childrenin conversations and to provide opportunities forchildren to write and to use books. The changethat appeared to reflect the most major shift wasthe extent to which teachers planned activitieswith the intention of having children practiceliteracy-related skills. Enriching the literacy en-vironment had one additional effect that mightcome as a surprise. Children whose teachers par-ticipated in LEEP displayed significantly moregrowth in social skills than their peers, as as-sessed by the Social Skills Rating System(Gresham & Elliott, 1990). Children who arebusy talking, reading, and writing—activitiesregistered by TROLL—were more likely to beviewed by their teachers as developing strongerskills in collaborating with others.

The Reading Teacher Vol. 56, No. 6 March 2003

Table 2 What TROLL scores mean

Note. Our data come from a low-income sample. In national studies children from such homes tend to receive less support for early lan-guage and literacy development than children from economically advantaged homes. Children from more advantaged backgroundswould be expected to receive somewhat higher ratings than those reported above (roughly 5–6 points higher on average).

3-year-olds’TROLL scores

Falln = 115

Falln = 83

Springn=229

Springn = 55

Springn = 234

Falln = 336

Recommendations/meaning

4-year-olds’TROLL scores

5-year-olds’TROLL scores

Relative standingon the TROLL

40 44 43 46 51 55 10th percentile Assessment by child of audi-ologist, speech-languagepathologist. Discuss concernswith parents.

44 49 52 55 59 65 25th percentile Assessment of child byspeech-language pathologist,extra involvement in extendedconversations, and other literacy activities.

51 56 61 66 68 76 50th percentile Child is performing at an average level.

61 62 71 74 75 85 75th percentile Child is performing above average.

68 69 80 84 85 91 90th percentile Child should be encouragedto read and write at advancedlevels in school and at home.

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 562 (Black plate)

562

Case studies We conclude by returning to Keisha and CJ.

Keisha scored the lowest on the TROLL of any-one in her class of 4-year olds—a total of 44points. Her teacher realized that she spent farmore time talking to Keisha’s high-scoring class-mates than she did to Keisha. Children who al-ready were the most advanced talkers were theones who asked questions, participated in groupdiscussions, and took many opportunities to ex-plain activities to other children or tell storiesabout themselves. The teacher found that in herclassroom, as in the classrooms of otherpreschool teachers (see Dickinson & Tabors,2001), the old adage of “the rich get richer” ap-plied to language and literacy development. Theteacher also realized that several of the childrenwho misbehaved also received low scores, andshe made an effort to involve them in more con-versations. Two such boys, in particular, seemedto thrive on this extra attention for desirable be-havior and were noticeably better behaved by theend of the year.

In the coming weeks, Keisha’s teacher madea concentrated effort to involve her in conversa-tion every day. The teacher also shared herTROLL assessment of Keisha with the child’smother, who acknowledged that she had beenstruggling with a number of issues and had nothad much time to talk with Keisha, let alone readwith her. The teacher recommended regular tripsto the library and setting aside time at meals justto talk about the day. By the end of the year,Keisha was far more talkative and began to ini-tiate looking at books on her own. She did nothave to fail at reading in order to get the helpshe needed to succeed. Keisha benefited fromthe kind of early intervention strongly recom-mended by Snow et al. (1998, pp. 318–319).

CJ’s teacher responded to his advancedskills (he scored 71, which placed him above the90th percentile) by making sure that his broth-ers read to him frequently and by involving hima couple of times a week in small-group bookreading, when she encouraged his conversationsand explained a number of terms in books thatwere unfamiliar to him.

Tracking children’s language and literacydevelopment is a critical yet challenging task.The TROLL provides one means for teachers toaccomplish this and can provide a starting point

for productive conversations with colleagues andparents.

ReferencesBissex, G.L. (1980). GNYS at wrk: A child learns to write and

read. Cambridge, MA: Harvard.Bryant, P., MacLean, M., & Bradley, L. (1990). Rhyme, lan-

guage, and children’s reading. Applied Psycholinguistics,11(3), 237–252.

Clay, M.M. (1979). Stones—The concepts about print test.Exeter, NH: Heinemann.

Cronin, V., & Carver, P. (1998). Phonological sensitivity, rapidnaming, and beginning reading. Applied Psycholinguistics,19(3), 447–462.

Dickinson, D.K. (2001). Large-group and free-play times:Conversational settings supporting language and literacy de-velopment. In D.K. Dickinson & P.O. Tabors (Eds.),Beginning literacy with language, pp. 223–256. Baltimore,MD: Brookes.

Dickinson, D.K., & DeTemple, J. (1998). Putting parents in thepicture: Maternal reports of preschoolers’ literacy as a pre-dictor of early reading. Early Childhood Research Quarterly,13(2), 241–263.

Dickinson, D.K., & McCabe, A. (1991). The acquisition and de-velopment of language: A social interactionist account of lan-guage and literacy development. In J.F. Kavanagh (Ed.), Thelanguage continuum from infancy to literacy (pp. 1–40).Parkton, MD: York.

Dickinson, D.K., & Tabors, P.O. (2001). Beginning literacy withlanguage: Young children learning at home and school.Baltimore, MD: Brookes.

Gresham, F.M., & Elliott, S.N. (1990). Social skills rating sys-tem: Ages 3–5. Circle Pines, MN: American GuidanceService.

International Reading Association & National Association for theEducation of Young Children. (1998). Learning to read andwrite: Developmentally appropriate practices for young chil-dren. Young Children, 53(4), 30–46.

McCabe, A., & Dickinson, D.K. (2001, Summer). Good talk,close listening: Laying solid foundations for literacy. Childrenand Families, pp. 21–22.

McCabe, A., & Rollins, P.R. (1991). Assessment of preschoolnarrative skills: Prerequisite for literacy. Miniseminar pre-sented at the Annual Convention of the American Speech-Language-Hearing Association, Atlanta, GA.

New Standards for Speaking and Listening Committee. (2001)New standards for speaking and listening for preschoolthrough third grade. Philadelphia: National Center onEducation and the University of Pennsylvania.

Pellegrini, A.D., & Galda, L. (1993). Ten years after: A reexam-ination of symbolic play and literacy research. ReadingResearch Quarterly, 28, 162–175.

Riley, J. (1996). The teaching of reading. London: PaulChapman.

Teacher Rating of Oral Language and Literacy (TROLL)

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 563 (Black plate)

Dickinson teaches at the Lynch School of Education (BostonCollege, Chestnut Hill, MA 02467-3813, USA). McCabeteaches at the University of Massachusetts Lowell, andSprague works at the Education Development Center inNewton, Massachusetts.

563

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 563 (PANTONE 300 CV plate)

Rosenthal, R., & Jacobsen, L. (1968). Pygmalion in the class-room: Teacher expectation and pupils’ intellectual develop-ment. New York: Holt, Rinehart and Winston.

Snow, C.E. (1983). Literacy and language: Relationships duringthe preschool years. Harvard Educational Review, 53,165–189.

Snow, C.E., Burns, M.S., & Griffin, P. (Eds.). (1998). Preventingreading difficulties in young children. Washington, DC:National Research Council, National Academy Press.

Speece, D.L., Roth, F.P., Cooper, D.H., & de la Paz, S. (1999).The relevance of oral language skills to early literacy: A mul-tivariate analysis. Applied Psycholinguistics, 20(2), 167–190.

Stanovich, K.E. (1992). Speculations on the causes and conse-quences of individual differences in early reading acquisi-tion. In P.B. Gough, L.C. Ehri, & R. Treiman (Eds.), Readingacquisition (pp. 307–342). Hillsdale, NJ: Erlbaum.

Stipek, D.J., & Ryan, R.H. (1997). Economically disadvantagedpreschoolers: Ready to learn but further to go. DevelopmentalPsychology, 33(4), 711–723.

Sulzby, E., & Teale, W. (1991). Emergent literacy. In R. Barr, M.Kamil, P. Mosenthal, & P.D. Pearson (Eds.), Handbook ofreading research (Vol. II, pp. 727–758). New York:Longman.

Tabors, P.O., Beals, D.E., & Weizman, Z.O. (2001). “You knowwhat oxygen is?” Learning new words at home. In D.K.Dickinson & P.O. Tabors (Eds.), Beginning literacy with lan-guage: Young children learning at home and school (pp.93–110). Baltimore, MD: Brookes.

Tucker, M.S., & Codding, J.B. (1998). Standards for ourschools: How to set them, measure them, and reach them. SanFrancisco, CA: Jossey-Bass.

Vellutino, F.R., & Scanlon, D.M. (2001). Emergent literacyskills, early instruction, and individual differences as deter-minants of difficulties in learning to read: The case for earlyintervention. In S. Neuman & D. Dickinson (Eds.), Handbookof early literacy research (pp. 295–321). New York: Guilford.

Wagner, R.K., Torgesen, J.K., Laughon, P., Simmons, K., &Rashotte, C.A. (1993). Development of young readers’ phono-logical processing abilities. Journal of EducationalPsychology, 85, 83–103.

Wagner, R.K., Torgesen, J.K., Rashotte, C.A., Hecht, S.A.,Barker, T.A., Burgess, S.R., et al. (1997). Changing relationsbetween phonological processing abilities and word-levelreading as children develop from beginning to skilled read-ers: A 5-year longitudinal study. Developmental Psychology,33, 468–479.

The Reading Teacher Vol. 56, No. 6 March 2003

Place your ad inTHE READING TEACHER

and/orJOURNAL OF ADOLESCENT

& ADULT LITERACYand reach top reading educatorswith news about your products,

services, position openings,conferences/seminars, etc.at a very reasonable cost.

CALL NOW:Linda Hunter, Advertising Manager

(302) 731-1600, ext. 261FAX 302-369-8476

98-27 A&M 6/00

Do you have aposition opening inthe reading field?

We can help you find thehighly qualified person youare looking for. Contact IRAAdvertising Manager, LindaHunter, at 302-731-1600, ext.261 or lhunter@reading.org.We offer special discountedrates for position opening ads.

01-128 A&M 11/01

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 564 (Black plate)

564

3/03 RT #31 Dickinson 2/12/03 1:44 PM Page 564 (PANTONE 300 CV plate)