55832821

Predicting SchoolAdjustment in EarlyElementary School:Impact of Teacher-Child RelationshipQuality and RelationalClassroom Climate

Evelien BuyseKarine VerschuerenPieter VerachtertJan Van DammeKatholieke Universiteit Leuven

Abstract

This longitudinal study evaluated the impact ofdyadic and classroom-level teacher-child rela-tionship quality in first grade on children’s psy-chosocial and academic adjustment in first (N �3,784), second (N � 3,666), and third (N � 3,582)grade, controlling for several child features,namely, child demographics and children’s ini-tial levels of adjustment in kindergarten. Resultsof multilevel hierarchical regression analysesshowed that first-grade dyadic relationship vari-ables (i.e., teacher-child conflict and closeness)as well as classroom relational climate variables(i.e., the average level of teacher-child conflictand closeness in the classroom) were associatedwith children’s psychosocial adjustment in thefirst years of primary school. Associations be-tween first-grade dyadic relationship qualityand classroom relational climate, on the onehand, and academic achievement on the other,however, were negligible.

Understanding how children adapt toschool has been an important objective forresearchers interested in the promotion ofcompetence and the prevention of educa-tional and psychological maladjustment.This objective has spurred research on thedeterminants of early school adaptation(Ladd, Birch, & Buhs, 1999). Historically,researchers have defined school adjustmentin terms of children’s academic progress orachievement. The construct of school ad-justment has thus been defined rather nar-rowly, and, as a consequence, the search forits antecedents has also been limited (Birch& Ladd, 1996). Most investigators have fo-cused on internal characteristics of the childas determinants of early school adjustment,such as gender and intelligence. Some re-searchers tried to elaborate upon these ear-

The Elementary School JournalVolume 110, Number 2© 2009 by The University of Chicago. All rights reserved.0013-5984/2009/11002-0001$10.00

lier models of school adjustment by restat-ing the concept of school adjustment itselfand rethinking its hypothesized determi-nants (Birch & Ladd, 1996; Ladd, 1989,2003). According to them, the concept ofschool adjustment should, for example, beexpanded to include dimensions of psycho-social adjustment as well (e.g., Birch &Ladd, 1997; Ladd, 2003; Ladd et al., 1999;Pianta, Steinberg, & Rollins, 1995). In ourstudy, we followed this broader definition,focusing on reading and mathematics skillsas aspects of academic achievement, andincluding children’s aggressive behavior,popularity with peers, and feelings of well-being at school as aspects of psychosocialadjustment at school.

In line with the broader definition, newideas arose considering the possible determi-nants of school adjustment. One such per-spective has been termed a “person by envi-ronment model,” in which successful schooladjustment is seen as originating both in thechild and in the surrounding interpersonalenvironment, such as the child’s relation-ships with teachers (Birch & Ladd, 1996;Ladd, 2003; Ladd et al., 1999). Over the past15 years, increasing attention has been de-voted to the role of relationships betweenchildren and teachers in influencing earlyschool adjustment (e.g., Pianta, 1999, 2006;Pianta, Hamre, & Stuhlman, 2003). In thesestudies, two dimensions are often seen asoperationalizations of teacher-child relation-ship quality, namely, teacher-child closenessand teacher-child conflict (Pianta et al., 1995,2003).

Closeness in the teacher-child relation-ship may function as a support in the schoolenvironment (Birch & Ladd, 1996). It is re-flected in the degree of warmth and opencommunication between a teacher and achild. It encompasses the extent to whichchildren seem comfortable approaching theteacher, talking about their feelings and ex-periences, and using the teacher as a sourceof support and comfort when upset. Havingwarm and open communication with theteacher may facilitate positive feelings to-

ward school. From close relationships withteachers, children may derive potential re-sources, such as emotional support and secu-rity, which may enhance positive behaviorsand exclude more negative behaviors (suchas aggression) in social contexts such as class-rooms (Birch & Ladd, 1998; Ladd et al., 1999).Children may then use their teachers as re-sources for other social relationships, includ-ing their relationships with peers (Howes,2000). A supportive teacher-child relation-ship may also motivate children to becomemore engaged in the school environment. Inthis manner, closeness may encourage youngchildren’s learning and performance inschool as well (Birch & Ladd, 1996, 1997;Howes, 2000). In line with these hypotheses,findings from previous studies conducted inkindergarten and/or the early grades of ele-mentary school indeed show that closenessin the teacher-child relationship functions asa support for young children in the schoolenvironment. Greater closeness in the teach-er-child relationship is associated with betteradjustment to school, for instance, with morepositive feelings about school (e.g., Birch &Ladd, 1997), fewer behavioral problems,more behavioral competencies and socialskills (e.g., Birch & Ladd, 1998; Howes, 2000;Hughes, Cavell, & Jackson, 1999; Pianta et al.,1995; Pianta & Stuhlman, 2004; Silver, Mea-selle, Armstrong, & Essex, 2005), and higheracademic achievement (e.g., Birch & Ladd,1997; Ladd et al., 1999).

Conflict in the teacher-child relation-ship may function as a stressor for childrenand may impair successful adjustment toschool (Birch & Ladd, 1996; Ladd et al.,1999). Conflicted relationships are charac-terized by discordant interactions and alack of rapport between teachers and chil-dren. As a potential stressor in the schoolenvironment, teacher-child conflict may beemotionally upsetting to young children,yielding negative behaviors. Relationshipconflict may, for example, exacerbate ag-gressive behavior (Birch & Ladd, 1998).Friction between teacher and child mayalso foster negative feelings about school or

120 THE ELEMENTARY SCHOOL JOURNAL

DECEMBER 2009

disengagement concerning school matters,causing academic problems as well (Birch& Ladd, 1996). In line with the idea thatteacher-child conflict may function as astressor for children’s early school adjust-ment, empirical research has confirmedthat it predicts more negative feelingsabout school (e.g., Birch & Ladd, 1997),more behavioral problems, fewer behav-ioral competencies and social skills (e.g.,Birch & Ladd, 1997, 1998; Doumen et al.,2008; Howes, 2000; Hughes et al., 1999;Mantzicopoulos, 2005; Pianta et al., 1995;Pianta & Stuhlman, 2004; Silver et al., 2005),and poorer academic performance or readi-ness (e.g., Hamre & Pianta, 2001; Ladd etal., 1999) in kindergarten and/or the firstyears of formal schooling.

In sum, teacher-child closeness and con-flict are found to play a significant role inregulating the development of the childrenin these relationships. This regulatory roleof both features of teacher-child relation-ships is thought to be especially relevant inthe early years of formal schooling, becausethose years comprise a sensitive period fordevelopment in school (Alexander & En-twisle, 1988). Up until third grade, whenschool trajectories are generally well estab-lished, relationships with teachers may or-ganize and provide direction to develop-ment (Pianta et al., 1995). Therefore, in ourstudy we investigated the significance offirst-grade teacher-child closeness and con-flict in the prediction of first-, second-, andthird-grade academic and psychosocial ad-justment at school.

As described above, some research hasalready been conducted on this topic. Inmany of these studies, however, principalpredictors of the outcomes were not con-trolled for. Moreover, if control variableswere included, this control has rarely beencomprehensive, including a set of multiplerelevant predictors of outcomes. Therefore,in our study we controlled for demo-graphic factors that have been shown todetermine children’s academic and psycho-social adjustment, including children’s gen-

der, socioeconomic status (SES), and ethnicbackground. In several international stud-ies, girls, children from families withhigher SES, and children of native parentsadjusted better to school, academically aswell as psychosocially (e.g., Hamre & Pi-anta, 2001; Ladd et al., 1999). In Belgium aswell, these are consistent findings (e.g.,Groenez, Van den Brande, & Nicaise, 2003;Hirtt, Nicaise, & De Zutter, 2007). In addi-tion to controlling for these backgroundfeatures, we also controlled for initial levelsof children’s academic and psychosocialadjustment, which was measured while thechildren were in kindergarten. In this man-ner, we could evaluate whether teacher-child closeness and teacher-child conflictprovide a unique prediction of child out-comes beyond child demographics and ini-tial levels of adjustment.

In addition to child characteristics, it isalso worthwhile to control for classroom in-teractional features when investigating theassociation between teacher-child relation-ship quality and child adjustment. Dyadicteacher-child relationships are embeddedwithin the overall interactional environmentof the classroom (Birch & Ladd, 1997; Howes,2000; Pianta et al., 2003). In classroom effec-tiveness literature, various dimensions of theinteractional quality of the classroom, such asinstructional and emotional climate, haveproved important for children’s school out-comes (e.g., Burchinal et al., 2000; Peisner-Feinberg et al., 2001; Pianta, La Paro, Payne,Cox, & Bradley, 2002; Wilson, Pianta, &Stuhlman, 2007). In our study, the averagelevel of teacher-child closeness and conflictwithin each classroom was used as an indi-cator of relational classroom climate. As inthe studies of Birch and Ladd (1997) andHowes (2000), we controlled for these vari-ables when predicting school adjustment out-comes from dyadic teacher-child closenessand conflict. It has been hypothesized thatbad classroom composition (e.g., high aver-age levels of conflict or low average levels ofcloseness) places an extra burden or stress onteachers, which can result in teachers behav-

PREDICTING SCHOOL ADJUSTMENT 121

ing more negatively with children, with as-sociated consequences for children’s adjust-ment (e.g., Mantzicopoulos, 2005; Pianta etal., 2003; Rimm-Kaufman, Pianta, & Cox,2000). However, this hypothesis has seldombeen tested explicitly. In the aforementionedstudies (Birch & Ladd, 1997; Howes, 2000) inwhich classroom relational composition vari-ables were included, effects were indeed de-tected of the relational classroom climate onchildren’s school adjustment in the earlyyears of formal schooling.

Furthermore, not only concurrent butalso prospective associations have beenfound between relational variables andyoung children’s school adjustment in sub-sequent school years (e.g., Hamre & Pianta,2001; Howes, 2000; Pianta et al., 1995; Silveret al., 2005). However, associations betweenrelationship variables and children’s out-comes may change over time and tend to bethe greatest when the outcomes and rela-tionships were rated in the same schoolyear (Pianta & Stuhlman, 2004). In ourstudy, relationship quality was evaluatedin first grade, but not in second or thirdgrade. We therefore expected the associa-tions between first-grade relational vari-ables and indicators of children’s school ad-justment to be larger in first grade than insecond and third grade.

In sum, in the present study we evalu-ated the predictive value of dyadic teacher-child closeness and conflict in first gradefor children’s academic and psychosocialadjustment over the first 3 years of elemen-tary school, controlling for several childfeatures, namely, child demographics andchildren’s initial levels of adjustment inkindergarten. Concerning the control vari-ables, we expected children’s levels of ad-justment in kindergarten to be positivelyrelated to children’s levels of adjustment inearly elementary school. Furthermore, wehypothesized that boys, children from fam-ilies with lower SES, and children from im-migrant families would perform worse onacademic tests and score lower on psycho-social teacher ratings. When evaluating the

association between teacher-child relation-ship quality and school adjustment, we alsocontrolled for first-grade relational class-room features, that is, the average level ofteacher-child closeness and conflict withineach classroom. We expected a more posi-tive relational climate, characterized byhigh average class levels of closeness orlow average class levels of conflict, to beassociated with better school adjustment.With regard to the key predictors of ourstudy, we expected higher teacher-childcloseness and lower teacher-child conflictto foster children’s school adjustment overand above relational classroom featuresand child-level covariates. Finally, we stud-ied the differential effects of dyadic andclassroom relational variables on children’sschool adjustment in the same and subse-quent school years. We hypothesized thatwe would find the strongest associationsbetween relationship variables and adjust-ment in first grade because relationshipquality was evaluated in that school year.

MethodParticipantsThe study described here was part of an

ongoing large-scale longitudinal study inFlemish education. It was designed to de-scribe and explain children’s educationalcareers throughout elementary education.1To this purpose, a random stratified sampleof 122 schools (215 classrooms) was se-lected. Stratification was based on educa-tional network and school size.2 This sam-ple was found to be representative of theentire Flemish school population in termsof the applied stratification criteria (Ver-haeghe, Maes, Gombeir, & Peeters, 2002).Data collection started on September 1,2002, when children entered their kinder-garten year (Verachtert, 2008). A large sam-ple of 3,798 kindergartners was recruited.Of all children, 50.5% were boys. At thebeginning of the school year, the mean ageof the kindergartners was 5 years and 3months. To assess the SES of the children’s


DECEMBER 2009

families, a score was compiled based on theeducational level of both parents, their pro-fessional status, and household income (n �3,461; see Reynders, Nicaise, & Van Damme,2005, for more detailed information on thisvariable). Most parents who filled out thequestionnaires completed high school (38%of the mothers and 41% of the fathers) or inaddition attended higher education (43% ofthe mothers and 39% of the fathers). Parents’nationality at birth was used as an indicatorof ethnic background (n � 3,296). Childrenwere classified into one of three categories:both parents had Belgian nationality at birth(Belgian, 80%), both parents had a foreignnationality (foreign, 11%), or one parent hadBelgian nationality while the other had a for-eign nationality (mixed, 9%).

When children arrived in first grade(2003–2004), second grade (2004–2005), andthird grade (2005–2006), data were availablefor 3,784 children, 3,666 children, and 3,582children, respectively, divided over 213 class-rooms (117 schools), 202 classrooms (118schools), and 200 classrooms (116 schools),respectively. Hence, for some children, scoreson the outcome variables were missing infirst, second, or third grade because thosechildren did not progress regularly throughthe successive grades or because they did notstay in a school participating in the study.Preliminary analysis revealed that childrenfor whom outcomes were missing in secondand/or third grade differed from childrenwho had complete data throughout thestudy. In particular, dropouts differed fromnondropouts with regard to SES (t(3448) ��14.26, p � .001), immigrant status (�2(2)� 41.21, p � .001), and initial levels of ad-justment, concerning aggressive behavior(t(3786) � 9.48, p � .001), popularity withpeers (t(3788) � �13.60, p � .001), feelings ofwell-being (t(3789) � �9.31, p � .001), andlanguage achievement (t(3695) � �22.28, p �.001) in kindergarten. Higher SES, lower per-centage of immigrants, and higher levels ofinitial adjustment were reported for non-dropouts. Effect sizes are rather modest, witheta-squared ranging between 1% and 12%.

Scores on mathematics achievement in kin-dergarten did not differ between dropoutsand nondropouts (t(3685) � �1.67, ns).

Furthermore, because of additionalmissing data on some predictor variables,the sample size is sometimes smallerthroughout the analyses. The exact numberof cases on which results are based is re-ported for every analysis separately.

InstrumentsRelationship quality. To evaluate the

quality of the teacher-child relationship infirst grade, we used a short Dutch versionof the Student-Teacher Relationship Scale(STRS; Pianta, 2001). The validity of the STRShas been demonstrated in relation to a rangeof social and academic outcomes (e.g., Pianta,2001). The short version (Buyse, Verschueren,Doumen, Van Damme, & Maes, 2008; Cor-nelissen & Verschueren, 2001) consists ofeight items and was used to assess teachers’perceptions of two features of their relation-ships with their pupils. The conflict subscalecomprises four items that tap the extent towhich the teacher-child relationship is char-acterized by disharmonious interactions (e.g.,”This child and I always seem to be strug-gling with each other”). The closeness sub-scale is a four-item index of the degree ofwarmth and open communication presentin the teacher-child relationship (e.g., “Thischild openly shares his/her feelings and ex-periences with me”).

In the study of Cornelissen and Vers-chueren (2001), the correlations betweenthe scores for both shortened subscales andthe complete scales were satisfactory: .92for relational closeness and .91 for rela-tional conflict. In the same study, high val-ues of internal consistency were found forboth scales of the shorter version of theSTRS (conflict: � � .87, closeness: � � .84),which was confirmed in our study (conflict:� � .82, closeness: � � .88). Moreover, con-firmatory factor analysis showed that close-ness and conflict are two different con-structs (Cornelissen & Verschueren, 2001;


see also Koomen, Verschueren, & Pianta,2007). In line with these results, the cor-relation between both subscales in thepresent study was significant, but not sub-stantively, justifying the inclusion of aseparate conflict and closeness subscale(r(377l) � �.20, p � .001).

Relational classroom climate. First-grade relational classroom climate vari-ables were obtained by averaging scores onteacher-child closeness and teacher-childconflict for all children in each classroom.Again, the correlation between both sub-scales was significant, but not substan-tively, justifying the inclusion of separatesubscales for average classroom closenessand conflict (r(215) � �.32). Further (mul-tilevel) analyses showed classroom averagecloseness to predict dyadic closeness (� �.34, p � .001), but not conflict (� � �.03, ns),whereas classroom average conflict pre-dicted dyadic conflict (� � .49, p � .001),but not closeness (� � �.05, ns).3

Psychosocial adjustment. Teachers inkindergarten and primary school (i.e., first,second, and third grade) rated children’spsychosocial adjustment in the classroomon several short subscales. All items on thedifferent subscales were rated on a six-point Likert scale with values ranging from1 (does not apply at all) to 6 (applies com-pletely). Three subscales were selected forthis study. For each subscale, scale scoreswere computed by averaging the itemscores.

First, the aggressive behavior subscalewas derived from the Child Behavior Scale(CBS), developed by Ladd and Profilet(1996). Supportive evidence for the validityof the CBS has been obtained, includingsignificant correlations in expected direc-tions with observations and peer ratings forbehavior (Doumen et al., 2008; Ladd & Pro-filet, 1996). For use in this study, the sub-scale was reduced from seven to four items(e.g., “Threatens other children”). This re-duction was based on the results of theoriginal factor analysis, performed and re-ported by Ladd and Profilet (1996). Ex-

cluded items were the items that had thelowest loadings on the common subscalefactor. Internal consistency of this subscalewas good, with Cronbach’s alpha coeffi-cient equaling .90, .92, .94, and .92 in kin-dergarten, first, second, and third grade,respectively. Correlations across gradesranged between .48 and .51 (p � .001).

Second, the popularity-with-peers sub-scale was selected from the PRIMA study(Driessen, van Langen, & Vierke, 2000; Jung-bluth, Roede, & Roeleveld, 2001). The itemson this subscale resemble the excluded-by-peers subscale from the CBS (Ladd & Pro-filet, 1996), but because face validity of thepopularity scale from the PRIMA study ishigher than face validity of the CBS sub-scale, and because the subscale is shorter,we preferred to work with the PRIMAscale. The scale consists of four items (e.g.,“Gets along well with peers”) with suffi-cient internal consistency, as Cronbach’s al-pha coefficients equaled .79, .83, .84, and .88in kindergarten and first, second, and thirdgrade, respectively, in our study. This is inaccordance with the internal consistencyreported in the PRIMA study (� � .84).Correlations across grades ranged between.39 and .44 (p � .001).

Third, a scale measuring feelings of well-being at school was developed for this study,based on another PRIMA scale (Driessen etal., 2000; Jungbluth et al., 2001) and on theschool-liking subscale of the Teacher RatingScale of School Adjustment (TRSSA; Ladd,1992). The scale consists of four items (e.g.,“Enjoys most of the classroom activities”),with Cronbach’s alpha equaling .82, .88, .87,and .88 in kindergarten and first, second, andthird grade, respectively. Correlations acrossgrades ranged between .20 and .33 (p � .001).

Academic achievement. The word-reading test (Moelands, Kamphuis, & Ry-menans, 2003; Moelands & Rymenans,2003), a Flemish version of a well-validatedDutch test (see Evers et al., 2002, for evi-dence concerning validity), was adminis-tered in first, second, and third grade. Thetest contains three reading charts with


DECEMBER 2009

words of increasing difficulty. For eachchart, the child was asked to read as manywords as accurately as he/she could for 1minute. The number of correctly readwords was recorded. Pearson correlationcoefficients between the scores for the threereading charts were high in first grade (.90–.95), second grade (.89–.94), and thirdgrade (.86–.92). Therefore, scores on thethree separate reading charts were aver-aged to yield a composite reading score foreach grade. Correlations between the com-posite reading scores across grades rangedbetween .57 and .87.

At the end of kindergarten, children wereadministered a shortened form of the lan-guage test Taal voor Kleuters voor Vlaan-deren (Kindergartners’ Language Achieve-ment Test for Flanders; Citogroep, 2003),which in turn is an adapted version of awidely used Dutch language proficiency test,Taal voor Kleuters (van Kuijk, 1996). The ad-aptations that were made to the original testare described by Ponjaert-Kristoffersen, An-dries, Celestin-Westreich, and Samaey (2000)and Verachtert (2003). The test that was usedin this study consisted of five differentsubtests (each including eight items): listen-ing comprehension, sound and rhyme, audi-tory sequencing, literacy knowledge, andsound blending. The internal consistency ofthe test was good (� � .86; Verachtert, 2003).

Curriculum-based mathematics achieve-ment tests specifically designed for use in thisstudy were administered. Tests for eachgrade were constructed in such a way thatthey had a considerable number of items incommon with the tests for the preceding andfollowing grades. Hence it was possible toconvert the raw mathematics scores to verti-cally equated Item Response Theory (IRT)–based scale scores (Lord, 1980). This conver-sion enabled the comparison of mathematicsscores across time. The mathematics tests ad-ministered in kindergarten and first gradeconsisted of 40 items, whereas the test forsecond and third grade included 50 itemsand 60 items, respectively (Verachtert, 2008).Internal consistencies, as measured by Cron-

bach’s alpha coefficient, were .92 in kinder-garten, first, and second grade, and .94 inthird grade. The kindergarten mathematicstest assessed a number of skills that are oftenincluded in definitions of number sense(Berch, 2005; Malofeeva, Day, Saco, Young, &Ciancio, 2004), including comparing magni-tudes, counting, and understanding mathe-matics concepts. The tests in primary schoolnot only contained items on number sense,but also applied items on arithmetic wordproblems, estimation, number decomposi-tion, number series, geometry, and/or men-tal arithmetic. Correlations across gradesranged between .81 and .83.

ProcedureInformation on child demographics was

initially collected through a kindergartenparent questionnaire distributed among thechildren in February of 2003. Additionaldata-collection efforts were performed atthe end of kindergarten and during the sec-ond and third year of the study in order toretrieve the information for the childrenwhose parents did not return the initialquestionnaire.

Teachers filled out the questionnairesconcerning children’s psychosocial adjust-ment in February of each school year. In firstgrade, teachers additionally rated teacher-child relationship quality items in February.Academic achievement tests were adminis-tered by the researchers in May of eachschool year. The word-reading test was ad-ministered for each child individually. Thelanguage test in kindergarten and the math-ematics tests were all untimed and were ad-ministered collectively. The sizes of the test-ing groups, however, differed betweenkindergarten and primary school. Becausekindergartners tend to have little experiencein taking paper-and-pencil tests, the kinder-garten mathematics and language tests wereadministered to small groups of about six toeight children instead of to the entire class.This decision was taken to ensure that testadministrators could provide enough guid-


ance to these young children when they tookthe tests. From the start of primary school, allmathematics tests were administered to allchildren in the entire classroom collectively(Verachtert, 2008).

Data AnalysisAll explanatory variables were centered

around the grand mean. Among the predic-tor variables included in this study, vari-ables from at least two different levels canbe identified: child-level variables (i.e.,child demographics, initial adjustment inkindergarten, and teacher-child relation-ship quality) and class-level variables (i.e.,relational classroom climate variables). Be-cause we wanted to evaluate child andclass characteristics as possible predictorsof children’s adjustment in primary school,data were analyzed by means of multilevelmodeling techniques (Goldstein, 1995),making use of the software program ML-wiN (Rasbash, Charlton, Browne, Healy, &Cameron, 2005). These techniques are espe-cially designed to analyze variables fromdifferent levels simultaneously (Hox, 2002).

Because we had data available for threesubsequent grades on all outcome vari-ables, time can be considered an additionalpredictor at an additional level in the multi-level models (Singer & Willett, 2003).Hence, three levels of information are in-volved in this study: time (level 1), nestedwithin children (level 2), and nested withinclasses (level 3).4 An attractive feature ofmultilevel models in the analysis of longi-tudinal data is the treatment of missingdata (Hedeker & Gibbons, 1997; Singer &Willett, 2003). Specifically, subjects who aremissing at a given wave are not excludedfrom the analysis, and “the model estimatesthe subject’s trend across time on the basisof whatever data the subject has, aug-mented by the time trend that is estimatedfor the sample as a whole and effects of allcovariates in the model” (Hedeker & Gib-bons, 1997, p. 65). Therefore, if subject at-trition is related to previous performance,

in addition to other observable subject char-acteristics, as is the case in our study (seecomparison of dropouts and nondropouts inthe “Participants” section above), multilevelmodels (making use of the maximum-likelihood estimation) provide valid statisti-cal inferences for the model parameters. Anoverview of the subsequent models that havebeen tested to predict children’s school ad-justment across the early years of elementaryschool is given in Figure 1.

Preliminarily, for each outcome, a three-level model was tested involving time as theonly predictor (this became the baselinemodel). In this manner, we could evaluatethe general trajectories for all children withinall classes during the first three grades ofprimary school concerning psychosocial andacademic adjustment. In the baseline model,and across all further models evaluated, theeffect of time was allowed to vary acrossclasses and across children. As can be seen inAppendix A, four additional random param-eters (i.e., the covariance term �CONS,TIME andthe variance term �TIME

2 at the class and thechild level) were therefore estimated, com-pared to a model in which the effect of timewould have been fixed (in which case only�CONS

2 would have been estimated at eachlevel). We included these extra parametersbecause we were interested in the possibledifferential effect of classroom- and child-level variables across time.

To address our research goals, we con-ducted multiple hierarchical regressionanalyses, starting from the baseline modelfor each of the outcomes. In a first step(Model A), children’s initial level of adjust-ment in kindergarten, corresponding to theoutcome predicted, as well as child demo-graphics (i.e., gender, SES, and ethnic back-ground), were entered as (child-level) pre-dictors. In a second step (Model B), theaverage level of first-grade teacher-childconflict and closeness within classroomswere added as (class-level) predictors toModel A, already containing children’s ini-tial level of adjustment and child demo-graphics as predictors. In a third step


DECEMBER 2009

(Model C), the individual level of first-grade teacher-child conflict and closenesswithin teacher-child dyads were added as(child-level) predictors to Model B, alreadycontaining children’s initial level of adjust-ment, child demographics, and class-levelrelational climate variables as predictors.

As mentioned above, we allowed theeffect of time to vary across classes andacross children, as we were interested inthe possible differential effect of classroom-and child-level variables across time. Morespecifically, we evaluated the differentialeffect of classroom relational variables anddyadic teacher-child relationship closenessand conflict across subsequent schoolyears. Therefore, interaction terms betweenthese relational variables, on the one hand,and the time variable on the other, wereadded to Models B and C. If an interactionwas significant, simple slope analyses wereperformed with the classroom and/or dy-adic relational variable(s) predicting theoutcome for each grade separately (based

on the procedure described by Aiken &West, 1991). Note that random parametersindicating random slopes for the effect oftime across children or classes need not besignificant to justify evaluation of these in-teraction effects. Evaluation of cross-levelinteractions can be based on a theoreticalargument formulated before looking at thedata and can therefore be tested irrespec-tive of whether a random slope has beenfound (Snijders & Bosker, 1999).

ResultsAs shown in Table 1, results from the base-line model indicate that children’s feelingsof well-being at school, as rated by theirteachers, decrease over the first 3 years ofprimary school. Children’s teacher-ratedaggressive behavior also (slightly) de-creases over time. Reading and mathemat-ics competencies, on the other hand, in-crease over the years. Finally, no changewas detected over time concerning chil-

FIG. 1.—Overview of the subsequent models in the prediction of children’s school adjustment across theearly years of elementary school. The baseline model only included time as a predictor. Model A included childcharacteristics as (child-level) predictors (i.e., gender, SES, ethnic background, and initial level of adjustment),next to the time variable. Model B included relational climate variables as (class-level) predictors, next to thepredictors that were included in Model A. In Model C, dyadic teacher-child relationship variables were addedas (child-level) predictors, next to the predictor variables that were included in Model B.


TA

BL

E1.

Pred

icti

ngC

hild

ren’

sPs

ycho

soci

alan

dA

cad

emic

Ad

just

men

tin

Gra

des

1–3

byIn

itia

lL

evel

sof

Ad

just

men

tin

Kin

der

gart

en,C

hild

Dem

ogra

phic

s,R

elat

iona

lC

lass

room

Clim

ate,

and

Dya

dic

Tea

cher

-Chi

ldR

elat

ions

hip

Qua

lity

Agg

ress

ive

Beh

avio

r(n

�7,

107)

aPo

pula

rity

(n�

7,10

6)a

Wel

l-B

eing

(n�

7,10

6)a

Rea

din

g(n

�6,

983)

bM

athe

mat

ics

(n�

6,99

4)c

��

R2

��

R2

��

R2

��

R2

��

R2

Bas

elin

em

odel

:T

ime

�.0

4*�

.01

�.4

0***

.70*

**.7

1***

Mod

elA

.23/

.26/

.26

.13/

.15/

.15

.06/

.07/

.07

.12/

.08/

.05

.47/

.47/

.43

Init

ial

leve

lki

nder

gart

en.4

5***

.36*

**.1

9***

.19*

**.4

5***

Gen

der

d�

.14*

**.0

5***

.08*

**.0

0�

.10*

**SE

S�

.09*

**.1

6***

.13*

**.0

9***

.09*

**E

thni

city

—m

ixed

.01

.00

�.0

1�

.02

�.0

2E

thni

city

—B

elgi

an.0

1.0

1.0

3�

.06*

**e

�.0

4**f

Mod

elB

.09/

.03/

.00

.07/

.03/

.00

.08/

.03/

.00

.00/

.00/

.00

Cla

ssav

erag

eco

nflic

t.1

8***

�.1

3***

�.0

9***

�.0

2C

lass

aver

age

clos

enes

s�

.03

.09*

**.1

0***

.03*

Inte

ract

ions

tim

e:C

lass

aver

age

confl

ict

�.4

2***

.26*

**.3

6***

�.0

1�

grad

e1

.26*

**�

.18*

**�

.18*

**�

grad

e2

.15*

**�

.11*

**�

.08*

**�

grad

e3

.03

�.0

4.0

2C

lass

aver

age

clos

enes

s�

.01

�.4

3***

�.7

0***

�.0

1�

grad

e1

.13*

**.1

8***

�gr

ade

2.0

7***

.09*

**�

grad

e3

.02

.01

Mod

elC

.11/

.10/

.08

.09/

.08/

.07

.07/

.07/

.06

.00/

.00/

.00

Tea

cher

-chi

ldco

nflic

t.3

5***

�.2

4***

�.1

5***

�.0

3**

Tea

cher

-chi

ldcl

osen

ess

�.0

1.1

6***

.20*

**�

.01

Inte

ract

ions

tim

e:T

each

er-c

hild

confl

ict

�.3

9***

.10*

*.3

0***

�.0

1�

grad

e1

.51*

**�

.28*

**�

.27*

**�

grad

e2

.33*

**�

.24*

**�

.14*

**�

grad

e3

.16*

**�

.20*

**.0

0

128

Tea

cher

-chi

ldcl

osen

ess

.07

�.4

8***

�.5

2***

�.0

1�

grad

e1

.29*

**.3

3***

�gr

ade

2.1

6***

.19*

**�

grad

e3

.04*

**.0

5**

R2

tota

l.4

3/.3

9/.3

4.2

9/.2

6/.2

5.2

1/.1

7/.1

3.1

2/.0

8/.0

5.4

7/.4

7/.4

3

NO

TE.—

R2

and

�R

2ar

ere

port

edfo

rea

chou

tcom

ein

first

grad

e/in

seco

ndgr

ade/

inth

ird

grad

e.a I

nth

epr

edic

tion

ofag

gres

sive

beha

vior

,po

pula

rity

wit

hpe

ers,

and

feel

ings

ofw

ell-

bein

gat

scho

ol,

dat

aw

ere

avai

labl

eon

all

pred

icto

rva

riab

les

for

2,65

7ch

ildre

n,yi

eld

ing

(2,6

57

3sc

hool

year

s�

)7,

971

dat

apo

ints

ofin

form

atio

n.O

ver

the

3sc

hool

year

s,86

4,86

5,an

d86

5pi

eces

ofin

form

atio

nw

ere

mis

sing

onag

gres

sive

beha

vior

,pop

ular

ity

wit

hpe

ers,

and

feel

ings

ofw

ell-

bein

gat

scho

ol,r

espe

ctiv

ely

(lea

din

gto

n�

7,10

7,n

�7,

106,

and

n�

7,10

6fo

rth

ese

outc

omes

,re

spec

tive

ly).

b In

the

pred

icti

onof

read

ing,

dat

aw

ere

avai

labl

eon

allp

red

icto

rva

riab

les

for

2,67

2ch

ildre

n,yi

eld

ing

8,01

6d

ata

poin

tsof

info

rmat

ion.

Ove

rth

e3

scho

olye

ars,

1,03

3pi

eces

ofin

form

atio

nw

ere

mis

sing

onre

adin

g(l

ead

ing

ton

�6,

983)

.c I

nth

epr

edic

tion

ofm

athe

mat

ics,

dat

aw

ere

avai

labl

eon

all

pred

icto

rva

riab

les

for

2,66

0ch

ildre

n,yi

eld

ing

7,98

0d

ata

poin

tsof

info

rmat

ion.

Ove

rth

e3

scho

olye

ars,

986

piec

esof

info

rmat

ion

wer

em

issi

ngon

read

ing

(lea

din

gto

n�

6,99

4).

d0

�bo

ys;1

�gi

rls.

e Wit

hout

anot

her

pred

icto

rin

the

mod

elth

anet

hnic

ity

(exc

ept

for

tim

efr

omth

epr

evio

usm

odel

),�

equa

ls.0

3(p

�.0

5).

f Wit

hout

anot

her

pred

icto

rin

the

mod

elth

anet

hnic

ity

(exc

ept

for

tim

efr

omth

epr

evio

usm

odel

),�

equa

ls.1

8(p

�.0

01).

*p�

.05.

**p

�.0

1.**

*p�

.001

.

129

dren’s popularity with peers. Estimation ofthe random part of the baseline modelshowed that the variance in intercepts(�CONS

2) at the class and child level is signif-icant for every outcome (see Appendix A).

When initial levels of children’s psycho-social or academic adjustment and chil-dren’s demographics were simultaneouslyadded to the baseline model as predictors(Model A), the deviance of the total modelreduced significantly for every outcomevariable (see Appendix A). With regard tothe specific predictors (see Table 1), we firstnotice that all aspects of children’s adjust-ment are substantially predicted by previ-ous measures of adjustment. More aggres-sive behavior, popularity with peers, andfeelings of well-being at school as rated bythe kindergarten teacher predict more ag-gressive behavior, popularity with peers,and feelings of well-being at school as ratedby the teachers in primary school. Further-more, more language and mathematicscompetencies in kindergarten predict betterreading and mathematics skills in primaryschool. Second, teachers in primary schoolrate girls as more popular with their peers,as feeling better at school, and as less ag-gressive than boys. Boys perform better atmathematics tests in the first years of pri-mary education, but no gender differencesare detected concerning reading. Third,children from families with a higher SESare rated by their teachers as more popularwith their peers and as feeling better atschool. Higher SES is also related to lessteacher-rated aggressive behavior, betterreading skills, and better mathematics per-formance. Finally, when controlling forother background characteristics and theinitial level of adjustment, we see that chil-dren whose parents both had Belgian na-tionality at birth perform worse in readingand mathematics than the children whoseparents both had a foreign nationality. Yet,post hoc analyses including ethnic back-ground as the sole child predictor showthat Belgians score significantly higher onboth tests. Furthermore, differences be-

tween the three categories referring to eth-nic status were significant with regard toSES (F(2, 3244) � 114.81, p � .001) and withregard to kindergarten levels of languageachievement (F(2, 3233) � 129.45, p � .001),with eta-squared equaling 8% and 7%, re-spectively, favoring Belgian children. Nosignificant differences were detected be-tween ethnic categories concerning mathe-matics achievement in kindergarten (F(2,3225) � 2.62, ns). Taken together, these posthoc analyses suggest that the negative (uni-variate) effect of immigrant status on aca-demic development during the first 3 yearsin elementary school may mainly be due toits association with low SES and lower kin-dergarten levels of (language) achievement.Indeed, when controlling for these corre-lates, the negative effect of immigrant sta-tus disappeared and even became positive.

After taking all child characteristics intoaccount simultaneously in Model A, a sub-stantial amount of variance is explained ineach of the outcomes—in first as well as insecond and third grade (see Table 1). Yet asignificant amount of variance is still situ-ated at the class and child level for everyoutcome variable (see Appendix A).

In an attempt to explain the remainingdifferences between classes, we added thetwo first-grade relational classroom climatevariables to the models already containingchildren’s initial levels of adjustment and de-mographic features (Model B). Overall, thedeviance of the total model reduced signifi-cantly in the prediction of every outcome,except in the prediction of reading. There-fore, Model B was not reported for this out-come variable. With regard to the other out-comes, first, we see that higher average levelsof teacher-child conflict in first grade are as-sociated with lower levels of psychosocial ad-justment. Specifically, higher average levelsof conflict are associated with more teacher-rated aggressive behavior and with lessteacher-rated popularity with peers and feel-ings of well-being at school. Second, higheraverage levels of teacher-child closeness infirst grade are associated with better psycho-


DECEMBER 2009

social adjustment in the classroom (i.e.,greater popularity with peers and more feel-ings of well-being at school). First-grade re-lational classroom variables explain 7% to 9%of the variance in children’s psychosocial ad-justment in first grade and 3% in secondgrade, beyond all other predictors already inthe model (see Table 1). In third grade, noimpact was detected from first-grade rela-tional climate variables on children’s psycho-social adjustment. Simple slope analyses,which yield a better insight in the differentialeffects of relational classroom variablesacross subsequent school years, reveal thesame conclusion (see Table 1): Effects of first-grade relational classroom variables are thelargest in first grade and are absent by thetime children are in third grade. As an exam-ple, Figure 2 shows how, beyond everythingelse in the model, the relationship betweenthe average class level of conflict and aggres-sive behavior is strongest in the earliergrades, which are temporally closer to thetime at which class-level conflict was mea-sured.

Considering academic achievement, wedetected a small yet significant effect of theaverage level of classroom closeness in firstgrade on children’s mathematics achieve-ment. The higher the average level of teacher-

child closeness in first grade, the better theindividual children of the classroom achievemathematically. However, based on the pro-portion of explained variance (0%), we con-clude that classroom relational variables infirst grade do not add substantially to theprediction of children’s academic achieve-ment. Not only in the prediction of academicachievement, but also in the prediction ofpsychosocial adjustment outcomes, there isstill a significant amount of variance left atthe child and class-level to be explained byother predictors, after including several childfeatures and relational climate indicators inModel B (see Appendix A).

In a final step, dyadic teacher-child rela-tionship conflict and closeness in first gradewere added to models already containingchildren’s initial levels of adjustment, demo-graphic features, and first-grade relationalclassroom climate variables (Model C). Over-all, the deviance of the total model reducedsignificantly in the prediction of every out-come, except in the prediction of reading.Therefore, Model C was not reported for thisoutcome variable. With regard to the otheroutcomes, first, we see that more dyadicteacher-child conflict in first grade is associ-ated with lower levels of psychosocial adjust-ment. Specifically, more conflict is associatedwith more teacher-rated aggressive behaviorand with less teacher-rated popularity withpeers and feelings of well-being at school.Second, more dyadic closeness in first gradeis associated with better psychosocial adjust-ment in the classroom (i.e., greater popularitywith peers and more feelings of well-being atschool). First-grade teacher-child relationshipquality variables explain 7% to 11% of thevariance in children’s psychosocial adjust-ment in first grade, 7% to 10% in secondgrade, and 6% to 8% in third grade, beyondall other predictors already in the model (seeTable 1). Simple slope analyses yield thesame conclusion: Effects of first-grade dyadicteacher-child relationship quality are thelargest in first grade and smaller in secondand third grade, but still exist by the timechildren are in third grade. As an example,

FIG. 2.—Differential effect of class average conflicton aggressive behavior for the subsequent grades.


Figure 3 shows that beyond everything elsein the model, the relationship betweenteacher-child conflict and aggressive behav-ior is strongest in the earlier grades, whichare temporally closer to the time at whichteacher-child conflict was measured. There isone exception to this trend, which is thatmore teacher-child conflict in first grade is nolonger associated with lower teacher-ratedfeelings of well-being in third grade (see Ta-ble 1).

Considering academic achievement, wedetected a small yet significant effect of dy-adic teacher-child conflict in first grade onchildren’s mathematics achievement. Themore conflict in the dyadic teacher-childrelationship, the worse children achievemathematically. However, based on theproportion of explained variance (0%), weconcluded that first-grade dyadic teacher-child relationship quality variables do notadd substantially to the prediction of chil-dren’s academic achievement.

DiscussionIn this study, we were especially interestedin the association between teacher-child re-lationship quality and children’s school ad-justment beyond several child features and

classroom relational climate. We found thatmore teacher-child closeness, as rated byfirst-grade teachers, was associated withbetter psychosocial adjustment, whereasmore teacher-child conflict was associatedwith worse psychosocial adjustment, asrated by teachers. Our findings are compat-ible with findings from other studies inwhich associations have been detected be-tween teacher-child closeness and conflict,on the one hand, and similar aspects ofchildren’s psychosocial functioning on theother, not controlling for any variables orfor a selection of the variables we controlledfor in our study (e.g., Birch & Ladd, 1997,1998; Howes, 2000; Hughes et al., 1999;Mantzicopoulos, 2005; Pianta et al., 1995; Pi-anta & Stuhlman, 2004; Silver et al., 2005).The associations we found between first-grade teacher-rated relationship quality andchildren’s psychosocial adjustment were vis-ible until third grade, but were most pro-nounced for adjustment in first grade. This isin line with the idea raised by Pianta andStuhlman (2004) that the teacher-child rela-tionship has primarily concurrent effects, thatis, teacher-child relationships may have thegreatest effect on children’s competency inthat specific classroom or relational context.The stronger within-year compared withacross-year associations may, however, alsobe due to shared informant bias: Teacherswho rated children’s psychosocial adjust-ment in first grade also rated the quality ofthe teacher-child relationship.

Regarding children’s academic achieve-ment in primary school, a very small addi-tional effect from relationship quality wasdetected. Specifically, more conflict in theteacher-child relationship in first grade wasrelated to worse mathematics achievementover the first 3 years in primary school.However, the proportion of explained vari-ance by relational conflict was negligiblefor this outcome. This result supports pre-vious studies in which teacher-child rela-tionship quality related more strongly withpsychosocial outcomes than with children’sachievement (e.g., Baker, 2006; Pianta &

FIG. 3.—Differential effect of teacher-child conflicton aggressive behavior for the subsequent grades.


DECEMBER 2009

Stuhlman, 2004). These differential associationscan be explained in several ways. First, ac-cording to Hamre and Pianta (2001), psycho-social outcomes may be considered to bemore proximal to the predictor, which is itselfa measure of psychosocial adjustment. Sec-ond, academic achievement may dependmore upon the instructional quality of theclassroom than upon the quality of interper-sonal relationships in the classroom (Hamre& Pianta, 2001). Furthermore, for at leastsome subgroups of children, academic prob-lems may be rooted in central nervous sys-tem dysfunction and may therefore requiredirect instruction and remediation of learn-ing deficits to improve school adjustment(Baker, 2006; Lerner, 2003). Third, comparedto psychosocial adjustment, we detectedstronger stability for academic achievementacross grades, leaving less room for relationalvariables to explain additional variance. Fi-nally, here too, method bias may play a role(especially in first grade) because both rela-tionship quality and psychosocial adjustmentvariables were teacher-rated, while more ob-jective measures were used to evaluate chil-dren’s academic adjustment.

In contrast to most studies concerning therole of (dyadic) teacher-child relationshipquality, our study also evaluated the role ofclassroom-level closeness and conflict. Re-sults showed that classroom relational cli-mate is also associated with children’s schooladjustment, beyond several child characteris-tics. Specifically, we found that classroom re-lational closeness, as rated by first-gradeteachers, was associated with better psycho-social adjustment, whereas classroom rela-tional conflict was associated with worse psy-chosocial adjustment as rated by teachers.Aside from a very small effect, no effectswere detected from the classroom relationalclimate on children’s academic adjustment.Our results were in line with findings fromtwo other studies that evaluated the effectfrom classroom relational climate variableson children’s adjustment. First, Birch andLadd (1997) found that mean conflict was(negatively) associated with children’s school

affect and attitudes in kindergarten, but notwith children’s academic achievement. Sec-ond, Howes (2000) found that a preschoolclimate of closeness was (positively) associ-ated with children’s social competence withpeers in second grade. The effects we foundof first-grade classroom relational climatewere visible in first grade as well as in secondgrade, but were no longer detectable whenchildren were in third grade.

In sum, we can conclude that both dyadicteacher-child relationships and classroom re-lational climate play significant roles in thedevelopment of child psychosocial compe-tencies that support a broad spectrum of ad-justment in the classroom, as has been sug-gested in previous studies as well (e.g.,Howes, 2000). Moreover, findings from ourstudy not only resonate with these earlierinvestigations but, in light of the longitudinalmethod applied, take us beyond a piecemealunderstanding of this phenomenon as de-scribed by various (important) cross-sectional analyses of prior decades.

As in previous studies (e.g., Birch &Ladd, 1997, 1998; Howes, 2000; Ladd et al.,1999; Pianta & Stuhlman, 2004), we alsofound several child features to be associ-ated with children’s adjustment at school.Girls tend to adjust better psychosocially,as do children with a higher SES and higherinitial levels of psychosocial adjustment inkindergarten. Boys perform better on math-ematics achievement tests, and childrenwith higher kindergarten levels of aca-demic adjustment generally perform betteron achievement tests in primary school.When we controlled for other child vari-ables, children whose parents had a foreignnationality performed better academicallythan Belgian children. This may seem sur-prising given the myriad of evidence thatchildren from native parents are generallyexpected to have better academic adjust-ment (see introduction). However, whenethnic background was the only predictor,native Belgian children did perform betteracademically. Along with additional anal-yses, this suggests that the negative effect


of immigrant status on academic develop-ment may mainly be due to its associationwith other background factors, such as lowSES, as well as with initially lower levels ofacademic achievement. A similar pattern ofresults has been obtained in other large-scale studies in Belgium, suggesting thatchildren with higher SES and native chil-dren adjust better to school than childrenwith lower SES and immigrants (e.g.,Groenez et al., 2003; Hirtt et al., 2007).

Finally, we found children’s academicachievement to increase over the years.This academic progress across the firstyears of primary school mirrors a naturalprocess of cognitive maturation. In con-trast, children’s aggressive behavior andespecially their feelings of well-being atschool were found to decrease over time.Several studies with older children havealso detected a decline in children’s psycho-logical well-being at school across grades(Harter, 1996; Rhodes, Roffman, Reddy, &Fredriksen, 2004). Although psychosocial ad-justment trajectories still need further study,the present findings suggest that this declinein well-being may already start soon afterchildren enter primary school.

Strengths, Limitations, and FutureDirectionsThe present study relies on a large sam-

ple, from which data were gathered longi-tudinally in the early years of primary ed-ucation. Statistically, we took into accountthe nested structure of the available data,making use of multilevel data-analysistechniques. Furthermore, we controlled forseveral child and classroom environmentalvariables in a comprehensive way whenevaluating the impact of teacher-child rela-tionship quality on various aspects of chil-dren’s school adjustment.

The inclusion of a representative Bel-gian sample in this study has the advantageof extending existing research on teacher-child relationships to countries outside theUnited States. At the same time, however,

this calls for further research on the gener-alizability of our specific findings to U.S.and other samples.

Perhaps the most significant limitation ofthe present study is its reliance on the teach-ers’ perspective in assessing both teacher-child relationships and children’s psychoso-cial adjustment outcomes. As a result, it canbe argued that the associations found stemfrom teachers’ response biases rather thanreflecting genuine relations between relation-ships and adjustment (Pianta et al., 1995). Ahalo effect my occur, referring to the possi-bility that teachers assign the same ratings fordifferent aspects of the children’s perfor-mance (Mashburn, Hamre, Downer, & Pi-anta, 2006). However, it is unlikely that theconnections established in this study aresolely due to shared response bias, as all ofthe effects found in first grade were alsofound in second grade, using another teacheras the informant of children’s psychosocialadjustment. Furthermore, it is important tonote that the quality of the teacher-child re-lationship and children’s broader psychoso-cial adjustment, as rated by the teacher, arevaluable indicators of school adaptation intheir own right (e.g., Birch & Ladd, 1997).Teacher ratings of children’s adjustment arepredictive of the future school trajectoriesand school adjustment of young children.Teachers’ beliefs and ideas about children’sadjustment in general, and about their rela-tional functioning in particular, affect, for ex-ample, whether children are referred for spe-cial services or placed in ability groups. Theyare also used to inform others, such as par-ents and other teachers (e.g., Birch & Ladd,1997; Mashburn et al., 2006).

It would, however, be worthwhile forfuture studies to include observations onteacher-child relationship quality and/orpsychosocial outcomes because teacher rat-ings on these variables may also sufferfrom a social desirability tendency. Yet, al-though observational information does notsuffer from social desirability bias, the factthat it is based on limited episodes of obser-vations, or snapshots, can be considered a


DECEMBER 2009

weakness. Teacher ratings, on the otherhand, are based on daily classroom behaviorand interactions and hence draw from exten-sive behavior information, covering an ex-tended period of time. Therefore, teacher rat-ings also provide an insightful perspective onteacher and child competencies within edu-cational settings (e.g., Kenny & Chekaluk,1993; Mashburn et al., 2006).

Practical ImplicationsThe associations we found between

classroom and dyadic relationship quality,on the one hand, and children’s adjustmenton the other, clearly support the view thatteacher-child relationships are an impor-tant relational context for the child, andmay provide a window for identification ofrisks and for intervention or prevention (Pi-anta et al., 1995). However, a focus onteacher-child relationships is currently stillmissing in several school-based preventionand intervention efforts as these effortshave generally been too centered on theindividual child (e.g., Sameroff & Macken-zie, 2003; Zajac & Kobak, 2006). Our studystrongly suggests that school practitionersshould not only focus on trying to changechildren, through skill training, for exam-ple, to adjust better in class. Instead, theyshould also pay attention to the influence ofrelational aspects of the classroom environ-ment in which children are embedded inorder to foster their adjustment (see alsoDeMulder, Denham, Schmidt, & Mitchell,2000; Hamre & Pianta, 2005; Sutherland &Oswald, 2005). At this moment, several in-tervention programs aimed at improvingdyadic teacher-child relationship qualityand/or more general relational classroom cli-mate (e.g., Driscoll & Pianta, in press; McIn-tosh, Rizza, & Bliss, 2000; Pianta & Hamre,2001) are being developed and evaluated.

In addition to directly improving thequality of dyadic teacher-child relation-ships and/or the relational climate, it couldalso be interesting to investigate the possi-ble determinants of these relational quali-

ties. Studying this could yield more cluesfor practitioners about how to effectivelyenhance relationship quality in the class-room and, consequently, about how to fos-ter children’s school adjustment indirectly.Teachers, for example, bring with them aset of beliefs about relationships with chil-dren based on their own history of experi-ences, which influences the interpretationof the child’s behavior and their subse-quent responses (Sameroff & MacKenzie,2003). Furthermore, the quality of the rela-tionships teachers form with children alsodepends on teachers’ attitudes, such as au-thoritarian attitudes (Pianta et al., 2005),and on teachers’ mental health, such as de-pressive symptoms (Hamre & Pianta, 2004;Hamre, Pianta, Downer, & Mashburn,2008). As part of teacher consultation, itmay be useful to make teachers more awareof these beliefs, attitudes, and symptomsand their effects on behavior and relation-ship quality. Concerning the classroomcontext, it has been shown that some struc-tural features, such as the ratio of childrento adults, predict the emotional quality ofcontacts with students (NICHD ECCRN,2002; Pianta et al., 2002). Likewise, in moreemotionally supportive classrooms, dyadicteacher-child relationships were found tobe less conflictual (Hamre et al., 2008).

Continued work in this direction willprovide teachers and other school person-nel with a better understanding of how tofacilitate positive outcomes for children inschool (Pianta & Stuhlman, 2004). As sug-gested in the current study, fostering chil-dren’s academic and psychosocial adjust-ment in school may each require differentmeasures to be taken. For improving chil-dren’s psychosocial adjustment across theearly years of formal schooling, this studyreveals the impact of strengthening teach-ers’ relationships with their pupils as earlyas first grade. Early qualitative teacher-child relationships may thus set the stagefor further scholastic development (Piantaet al., 1995), at least when it comes to psy-chosocial aspects of development.


Appendix A

Random Parameters, Deviance, and Significance of the Reduction inDeviance for Nested Models Fitted in MLwiN

TABLE A1. Prediction of Aggressive Behavior (n � 7,107)

Baseline Model A Model B Model C

B (SE) B (SE) B (SE) B (SE)

Random part:Class level:

�CONS2 .11** (.02) .14** (.02) .07** (.01) .07** (.01)

�CONS,TIME �.05** (.01) �.06** (.01) �.04** (.01) �.04** (.01)�TIME

2 .06** (.01) .06** (.01) .06** (.01) .06** (.01)Child level:

�CONS2 .48** (.02) .18** (.02) .18** (.02) .02 (.01)

�CONS,TIME �.03* (.01) .01 (.01) .01 (.01) .05** (.01)�TIME

2 . . .a . . .a . . .a . . .aTime level .45** (.01) .45** (.01) .45** (.01) .44** (.01)

Deviance 18253.226 16937.046 16875.782 16209.537�difference deviance

2 �2(5) � 1316** �2(2) � 61** �2(2) � 666**

NOTE.—The baseline model only included time as a predictor. Model A included child characteristics as(child-level) predictors (i.e., gender, SES, ethnic background, and initial level of adjustment), next to the timevariable. Model B included relational climate variables as (class-level) predictors, next to the predictors that wereincluded in Model A. In Model C, dyadic teacher-child relationship variables were added as (child-level)predictors, next to the predictor variables that were included in Model B.

aThis parameter is dropped for the model to converge.*p � .01.**p � .001.

TABLE A2. Prediction of Popularity with Peers (n � 7,106)




�CONS2 .07* (.01) .10* (.01) .05* (.01) .05* (.01)

�CONS,TIME �.04* (.01) �.05* (.01) �.03* (.01) �.03* (.01)�TIME

2 .05* (.01) .04* (.01) .05* (.01) .05* (.01)Pupil level:

�CONS2 .26* (.02) .12* (.01) .12* (.01) .03* (.01)

�CONS,TIME .02* (.01) .03* (.01) .03* (.01) .06* (.01)�TIME

2 . . .a . . .a . . .a . . .aTime level .39* (.01) .39* (.01) .39* (.01) .39* (.01)


2 �2(5) � 784* �2(2) � 66* �2(2) � 511*


aThis parameter is dropped for the model to converge.*p � .001.


DECEMBER 2009

TABLE A3. Prediction of Feelings of Well-Being at School (n � 7,106)




�CONS2 .09* (.01) .10* (.01) .05* (.01) .05* (.01)

�CONS,TIME �.05* (.01) �.05* (.01) �.03* (.01) �.03* (.01)�TIME

2 .04* (.01) .04* (.01) .04* (.01) .04* (.01)Pupil level:

�CONS2 .25* (.01) .17* (.01) .19* (.01) .09* (.01)

�CONS,TIME �.07* (.01) �.05* (.00) �.06* (.00) �.03* (.01)�TIME

2 . . .a . . .a . . .a . . .aTime level .41* (.01) .40* (.01) .40* (.01) .40* (.01)


2 �2(5) � 454* �2(2) � 187* �2(2) � 141*


aThis parameter is dropped for the model to converge.*p � .001.

TABLE A4. Prediction of Reading (n � 6,983)




�CONS2 24.41* (4.32) 21.14* (3.77)

�CONS,TIME �8.76* (1.81) �7.33* (1.65)�TIME

2 8.75* (1.24) 8.61* (1.23)Pupil level:

�CONS2 176.49* (6.96) 148.23* (6.18)

�CONS,TIME 13.85* (2.31) 16.76* (2.19)�TIME

2 .20 (1.53) .25 (1.53)Time level 67.50* (2.05) 67.37* (2.05)

Deviance 55238.280 54917.170 (54993.590) (54975.880)�difference deviance

2 �2(5) � 321*

NOTE.—The baseline model only included time as a predictor. Model A included child characteristics as(child-level) predictors (i.e., gender, SES, ethnic background, and initial level of adjustment), next to the timevariable. Model B and Model C were not reported in more detail because the deviance of these models was largerthan the deviance of Model A, indicating a worse overall fit.

*p � .001.


Notes

1. The study referred to here is the SiBOstudy. SiBO is a Dutch acronym for Schoolloop-banen in het BasisOnderwijs (School Careers inPrimary Education). This study was funded byresearch grant 3H070039 for the Policy ResearchCenter’s “Study and School Careers,” Depart-ment of Education, Ministry of the FlemishCommunity (Belgium). Correspondent: [email protected].

2. In Flemish education, three educationalnetworks are usually distinguished. First, com-munity schools are public schools functioningunder the authority of the Flemish Community.Second, subsidized public schools provide edu-cation organized under the authority of munici-palities and provinces. Third, subsidized private(mostly Catholic) schools provide education un-der the authority of a private person or organiza-tion (Verachtert, 2008).

3. A simple Pearson correlation could not becalculated here because average classroom close-ness and conflict are class-level variables, whereasdyadic closeness and conflict are child-level vari-ables.

4. The variance at the school level did notreach significance for any of the outcomes(equaling 1%–2% of the total variance). There-fore, the school level was excluded as a fourthlevel.

References

Aiken, L., & West, S. (1991). Multiple regression:Testing and interpreting interactions. NewburyPark, CA: Sage.

Alexander, K. L., & Entwisle, D. R. (1988).Achievement in the first two years of school:Patterns and processes. Monographs of the Soci-ety for Research in Child Development, 53(2, Se-rial No. 218).

Baker, J. A. (2006). Contributions of teacher-child relationships to positive school adjust-ment during elementary school. Journal ofSchool Psychology, 44, 211–229.

Berch, D. B. (2005). Making sense of numbersense: Implications for children with mathe-matical disabilities. Journal of Learning Dis-abilities, 38, 333–339.

Birch, S. H., & Ladd, G. W. (1996). Interpersonalrelationships in the school environment andchildren’s early school adjustment: The roleof teachers and peers. In J. Juvonen & K. R.Wentzel (Eds.), Social motivation: Understand-ing children’s school adjustment (pp. 199–225).New York: Cambridge University Press.

Birch, S. H., & Ladd, G. W. (1997). The teacher-child relationship and children’s earlyschool adjustment. Journal of School Psychol-ogy, 35, 61–79.

Birch, S. H., & Ladd, G. W. (1998). Children’sinterpersonal behaviors and the teacher-

TABLE A5. Prediction of Mathematics Achievement (n � 6,994)




�CONS2 11.34** (1.61) 5.85** (.83) 5.74** (.82) 5.74** (.82)

�CONS,TIME �3.48** (.62) �2.32** (.44) �2.32** (.44) �2.32** (.44)�TIME

2 2.79** (.37) 2.74** (.36) 2.74** (.36) 2.74** (.36)Pupil level:

�CONS2 44.01** (1.47) 15.64** (.66) 15.62** (.66) 7.42** (.87)

�CONS,TIME . . .a . . .a . . .a . . .a�TIME

2 . . .a . . .a . . .a . . .aTime level 16.94** (.37) 16.81** (.37) 16.81** (.37) 16.81** (.37)


2 �2(5) � 2156** �2(2) � 6.26* �2(2) � 159**

NOTE.—The baseline model only included time as a predictor. Model A included child characteristics as(child-level) predictors (i.e., gender, SES, ethnic background, and initial level of adjustment), next to the timevariable. Model B included relational climate variables as (class level) predictors, next to the predictors that wereincluded in Model A. In Model C, dyadic teacher-child relationship variables were added as (child-level)predictors, next to the predictor variables that were included in Model B.

aThis parameter is dropped for the model to converge.*p � .05.**p � .001.


DECEMBER 2009

child relationship. Developmental Psychology,34, 934–946.

Burchinal, M. R., Roberts, J. E., Riggins R., Jr.,Zeisel, S. A., Neebe, E., & Bryant, D. (2000).Relating quality of center-based child care toearly cognitive and language developmentlongitudinally. Child Development, 71, 338–357.

Buyse, E., Verschueren, K., Doumen, S., VanDamme, J., & Maes, F. (2008). Classroomproblem behavior and teacher-child rela-tionships in kindergarten: The moderatingrole of classroom climate. Journal of SchoolPsychology, 46, 367–391.

Citogroep (2003). Taal voor Kleuters voor Vlaan-deren [Kindergartners’ Language Achieve-ment Test for Flanders]. Arnhem, the Neth-erlands: Citogroep.

Cornelissen, G., & Verschueren, K. (2001).Nederlandse vertaling van de Student-TeacherRelationship Scale (STRS) (Pianta, 1996):Toetsing van de factorstructuur en constructievan een verkorte versie [Dutch translation ofthe Student-Teacher Relationship Scale(Pianta, 1996): Testing of factor structureand construction of a shortened version].Leuven, Belgium: Center for School Psy-chology, Katholieke Universiteit Leuven.Retrieved June 16, 2008, from http://www.steunpuntloopbanen.be/rapporten/LOA-rapport_03.pdf

DeMulder, E. K., Denham, S., Schmidt, M., &Mitchell, J. (2000). Q-sort assessment of at-tachment security during the preschoolyears: Links from home to school. Develop-mental Psychology, 36, 274–282.

Doumen, S., Verschueren, K., Buyse, E., Ger-meijs, V., Luyckx, K., & Soenens, B. (2008).Reciprocal relations between teacher-childconflict and aggressive behavior in kinder-garten: A three-wave longitudinal study.Journal of Clinical Child and Adolescent Psy-chology, 37, 588–599.

Driessen, G., van Langen, A., & Vierke, H.(2000). Basisonderwijs: Veldwerkverslag, leer-linggegevens en oudervragenlijsten. Basisrap-portage PRIMA-cohortonderzoek. Derde meting1998–1999 [Elementary school: Report on pu-pil data and parent questionnaires. PRIMAcohort study. Third wave 1998–1999]. Nijme-gen, the Netherlands: ITS.

Driscoll, K. C., & Pianta, R. C. (in press). Bank-ing Time in Head Start: Early efficacy of anintervention designed to promote support-ive teacher-child relationships. Early Educa-tion and Development.

Evers, A., van Vliet-Mulder, J. C., Resing,W. C. M., Starren, J. C. M. G., van Alphen deVeer, R. J., & van Boxtel, H. (2002). COTAN

Testboek voor het onderwijs [COTAN testbookfor education]. Amsterdam: Boom.

Goldstein, H. (1995). Multilevel statistical models.London: Arnold.

Groenez, S., Van den Brande, I., & Nicaise, I.(2003). Cijferboek sociale ongelijkheid in hetVlaamse onderwijs: Een verkennend onderzoekop de panelstudie van Belgische huishoudens[Report on social inequality in Flemish edu-cation: Exploratory research on the panelstudy on Belgian households]. Leuven, Bel-gium: HIVA, Steunpunt LOA, Unit Onder-wijsloopbanen.

Hamre, B. K., & Pianta, R. C. (2001). Earlyteacher-child relationships and the trajectoryof children’s school outcomes througheighth grade. Child Development, 72, 625–638.

Hamre, B. K., & Pianta, R. C. (2004). Self-reported depression in nonfamilial caregiv-ers: Prevalence and associations with care-giver behavior in child-care settings. EarlyChildhood Research Quarterly, 19, 297–318.

Hamre, K. H., & Pianta, R. C. (2005). Can in-structional and emotional support in thefirst-grade classroom make a difference forchildren at risk of school failure? Child De-velopment, 76, 949–967.

Hamre, K. H., & Pianta, R. C., Downer, J. T., &Mashburn, A. J. (2008). Teachers’ percep-tions of conflict with young students: Look-ing beyond problem behaviors. Social Devel-opment, 17, 115–136.

Harter, S. (1996). Teacher and classmate influ-ences on scholastic motivation, self-esteem,and level of voice. In J. Juvonen & K. R.Wentzel (Eds.), Social motivation: Understand-ing children’s school adjustment (pp. 11–42).New York: Cambridge University Press.

Hedeker, D., & Gibbons, R. D. (1997). Applica-tion of random-effects pattern-mixture mod-els for missing data in longitudinal studies.Psychological Methods, 2, 64–78.

Hirtt, N., Nicaise, I., & De Zutter, D. (2007). Deschool van de ongelijkheid [The school of in-equality]. Berchem, Belgium: EPO.

Howes, C. (2000). Social-emotional classroomclimate in child care, child-teacher relation-ships and children’s second grade peer rela-tions. Social Development, 9, 191–204.

Hox, J. (2002). Multilevel analysis: Techniques andapplications. Mahwah, NJ: Erlbaum.

Hughes, J. N., Cavell, T. A., & Jackson, T. (1999).Influence of the teacher-child relationship onchildhood conduct problems: A prospectivestudy. Journal of Clinical Child Psychology, 28,173–184.

Jungbluth, P., Roede, E., & Roeleveld, J. (2001).Validering van het PRIMA-leerlingprofiel: Se-


cundaire analyse op het PRIMA-cohort bestand(SCO-rapport nr. 608) [Validation of thePRIMA student profile: Secondary analysis onthe PRIMA cohort data (SCO report no. 608)].Amsterdam: SCO-Kohnstamn lnstituut van deFaculteit der Maatschappij- en Gedragsweten-schappen, Universiteit van Amsterdam.

Kenny, D. T., & Chekaluk, E. (1993). Early read-ing performance: A comparison of teacher-based and test-based assessments. Journal ofLearning Disabilities, 26, 227–236.

Koomen, H., Verschueren, K., & Pianta, R.(2007). Leerling Leerkracht Relatie Vragenlijst[Student Teacher Relationship Scale].Houten, the Netherlands: Bohn Stafleu vanLoghum.

Ladd, G. W. (1989). Children’s social compe-tence and social supports: Precursors ofearly school adjustment? In B. H. Schneider,G. Attili, J. Naldel, & R. P. Weisberg (Eds.),Social competence in developmental perspective(pp. 277–291). Amsterdam: Kluwer.

Ladd, G. W. (1992). The Teacher Rating Scale ofSchool Adjustment. Urbana-Champaign: Uni-versity of Illinois.

Ladd, G. W. (2003). Probing the adaptive signif-icance of children’s behavior and relation-ships in the school context: A child by envi-ronment perspective. In R. V. Kail (Ed.),Advances in child development and behavior (Vol.31, pp. 43–104). San Diego: Academic Press.

Ladd, G. W., Birch, S. H., & Buhs, E. S. (1999).Children’s social and scholastic lives in kin-dergarten: Related spheres of influence?Child Development, 70, 1373–1400.

Ladd, G. W., & Profilet, S. M. (1996). The ChildBehavior Scale: A teacher-report measure ofyoung children’s aggressive, withdrawn,and prosocial behaviors. Developmental Psy-chology, 32, 1008–1024.

Lerner, J. W. (2003). Learning disabilities: Theories,diagnosis, and teaching strategies (9th ed.). Bos-ton: Houghton Mifflin.

Lord, F. M. (1980). Applications of item responsetheory to practical testing problems. Hillsdale,NJ: Erlbaum.

Malofeeva, E., Day, J., Saco, X., Young, L., & Cian-cio, D. (2004). Construction and evaluation of anumber sense test with Head Start children.Journal of Educational Psychology, 96, 648–659.

Mantzicopoulos, P. (2005). Conflictual relation-ships between kindergarten children andtheir teachers: Associations with child andclassroom context variables. Journal of SchoolPsychology, 43, 425–442.

Mashburn, A. J., Hamre, B. K., Downer, J. T., &Pianta, R. C. (2006). Teacher and classroomcharacteristics associated with teachers’ rat-

ings of prekindergartners’ relationships andbehaviors. Journal of Psychoeducational Assess-ment, 24, 367–380.

McIntosh, D. E., Rizza, M. G., & Bliss, L. (2000).Implementing empirically supported inter-ventions: Teacher-child interaction therapy.Psychology in the Schools, 37, 453–462.

Moelands, F., Kamphuis, F., & Rymenans, R.(2003). Drie-Minuten-Toets voor Vlaanderen(DMT-V): Wetenschappelijke verantwoording[Three-Minute-Test for Flanders (DMT-V):Scientific justification]. Arnhem, the Nether-lands: Citogroep.

Moelands, F., & Rymenans, R. (2003). Drie-Minuten-Toets voor Vlaanderen: Handleiding[Three-Minute-Test for Flanders: Manual].Arnhem, the Netherlands: Citogroep.

NICHD Early Child Care Research Network.(2002). The relation of global first-gradeclassroom environment to structural class-room features, teacher, and student behav-iors. Elementary School Journal, 102, 367–387.

Peisner-Feinberg, E. S., Burchinal, M. R., Clif-ford, R. M., Culkin, M. L., Howes, C., Kagan,S. L., et al. (2001). The relation of preschoolchild-care quality to children’s cognitive andsocial developmental trajectories throughsecond grade. Child Development, 72, 1534–1553.

Pianta, R. C. (1999). Enhancing relationships be-tween children and teachers. Washington, DC:American Psychological Association.

Pianta, R. C. (2001). Student-Teacher RelationshipScale: Professional manual. Lutz, FL: Psycho-logical Assessment Resources.

Pianta, R. C. (2006). Schools, schooling and de-velopmental psychopathology. In D. Cic-chetti & D. J. Cohen (Eds.), Developmentalpsychopathology (2nd ed., pp. 494–529). Ho-boken, NJ: Wiley.

Pianta, R. C., & Hamre, B. K. (2001). Student,teachers and relationship support (STARS).Odessa, FL: Psychological Assessment Re-sources.

Pianta, R. C., Hamre, B. K., & Stuhlman, M.(2003). Relationships between teachers andchildren. In W. M. Reynolds, G. E. Miller, &I. B. Weiner (Eds.), Handbook of psychology:Vol. 7. Educational psychology (pp. 199–234).Hoboken, NJ: Wiley.

Pianta, R. C., Howes, C., Burchinal, M., Bryant, D.,Clifford, R., Early, C., et al. (2005). Features ofpre-kindergarten programs, classrooms, andteachers: Do they predict observed classroomquality and child-teacher interactions? AppliedDevelopmental Science, 9, 144–159.

Pianta, R. C., La Paro, K. M., Payne, C., Cox,M. J., & Bradley, R. (2002). The relation of


DECEMBER 2009

kindergarten classroom environment toteacher, family and school characteristicsand child outcomes. Elementary School Jour-nal, 102, 225–238.

Pianta, R. C., Steinberg, M., & Rollins, K. B.(1995). The first two years of school:Teacher-child relationships and deflectionsin children’s classroom adjustment. Develop-mental Psychopathology, 7, 295–312.

Pianta, R. C., & Stuhlman, M. W. (2004).Teacher-child relationship and children’ssuccess in the first years of school. SchoolPsychology Review, 33, 444–458.

Ponjaert-Kristoffersen, I., Andries, C., Celestin-Westreich, S., & Samaey, C. (2000). Onder-zoek naar het gebruik van het Leerlingvolgsys-teem voor jonge kinderen in Vlaanderen(OBPWO-project 96.08): Eindrapport [Investi-gation of the use of the achievement moni-toring system for young children in Flanders(OBPWO project 96.08): Final report]. Brus-sels: Vrije Universiteit Brussel, VakgroepOntwikkelings- en Levenslooppsychologie.

Rasbash, J., Charlton, C., Browne, W. J., Healy,M., & Cameron, B. (2005). MLwiNVersion2.02. Bristol: Centre for Multilevel Model-ling, University of Bristol.

Reynders, T., Nicaise, I., & Van Damme, J.(2005). Longitudinal onderzoek in het ba-sisonderwijs: De constructie van een SES-variabele voor het SiBO-onderzoek (LOA-rapport nr. 31). [A longitudinal study inprimary education: Construction of an SESvariable for use in the SiBO study (LOA-report no. 31)]. Leuven, Belgium: Steunpunt‘Loopbanen doorheen Onderwijs naar Ar-beidsmarkt’, Cel ‘Schoolloopbanen in het ba-sisonderwijs’ (SiBO). Retrieved June 16, 2008,from http://www.steunpuntloopbanen.be/rapporten/LOA-rapport_31.pdf

Rhodes, J., Roffman, J., Reddy, R., & Fredriksen,K. (2004). Changes in self-esteem during themiddle school years: A latent growth curvestudy of individual and contextual influences.Journal of School Psychology, 42, 243–261.

Rimm-Kaufman, S. E., Pianta, R. C., & Cox, M. J.(2000). Teachers’ judgments of problems inthe transition to kindergarten. Early Child-hood Research Quarterly, 15, 147–166.

Sameroff, A. J., & Mackenzie, M. J. (2003). Re-search strategies for capturing transactionalmodels of development: The limits of thepossible. Development and Psychopathology,15, 613–640.

Silver, R. B., Measelle, J. R., Armstrong, J. M., &Essex, M. J. (2005). Trajectories of classroomexternalizing behavior: Contributions of childcharacteristics, family characteristics, and the

teacher-child relationship during the schooltransition. Journal of School Psychology, 43, 39–60.

Singer, J. D., & Willett, J. B. (2003). Applied lon-gitudinal data analysis. New York: OxfordUniversity Press.

Snijders, T. A. B., & Bosker, R. J. (1999). Multi-level analysis: An introduction to basic and ad-vanced multilevel modeling. London: Sage.

Sutherland, K. S., & Oswald, D. P. (2005). Therelationship between teacher and studentbehavior in classrooms for students withemotional and behavioral disorders: Trans-actional processes. Journal of Child and FamilyStudies, 14, 1–14.

van Kuijk, J. J. (1996). Taal voor Kleuters [Kinder-gartners’ Language Achievement Test]. Arn-hem, the Netherlands: Centraal lnstituutvoor Toetsontwikkeling.

Verachtert, P. (2003). Longitudinaal onderzoek in hetbasisonderwijs: Toetsen schooljaar 2002–2003(LOA-rapport nr. 14) [A longitudinal study inprimary education: Achievement tests admin-istered in school year 2002–2003 (LOA-reportno. 14)]. Leuven, Belgium: Steunpunt ‘Loop-banen doorheen Onderwijs naar Arbeids-markt’, Cel ‘Schoolloopbanen in het ba-sisonderwijs’ (SiBO). Retrieved June 16, 2008,from http://www.steunpuntloopbanen.be/rapporten/LOA-rapport_14.pdf

Verachtert, P. (2008). Kindergarten risk factors forearly mathematics difficulties: Evidence from thelongitudinal SiBO study. Unpublished doc-toral dissertation, Centrum voor Gezins-enOrthopedagogiek, Centrum voor Method-ologie van het Pedagogisch Onderzoek,Katholieke Universiteit Leuven, Belgium.

Verhaeghe, J. P., Maes, F., Gombeir, D., &Peeters, E. (2002). Longitudinaal onderzoek inhet basisonderwijs: Steekproeftrekking (LOA-rapport nr. 5) [A longitudinal study in pri-mary education: Sample procedure (LOA-report no. 5)]. Leuven, Belgium: Steunpunt‘Loopbanen doorheen Onderwijs naar Ar-beidsmarkt’, Cel ‘Schoolloopbanen in het ba-sisonderwijs’ (SiBO). Retrieved June 16, 2008,from http://www.steunpuntloopbanen.be/rapporten/LOA-rapport_05.pdf

Wilson, H. K., Pianta, R. C., & Stuhlman, M.(2007). Typical classroom experiences in firstgrade: The role of classroom climate and func-tional risk in the development of social com-petencies. Elementary School Journal, 108, 81–96.

Zajac, K., & Kobak, R. (2006). Attachment. InG. G. Bear & K. M. Minke (Eds.), Children’sneeds III: Development, prevention and inter-vention (pp. 379–389). Washington, DC: Na-tional Association of School Psychologists.


Copyright of Elementary School Journal is the property of University of Chicago Press and its content may not

be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.

55832821

Documents

taal voor kleuters

toets voor vlaanderen

van den brande

simple slope analyses

classroom relational climate

relational classroom climate

katholieke universiteit leuven

psychological assessment resources