Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Teacher-Child Relationships and Children's Success in the First Years of School

Teacher-Child Relationships and Children's Success in the First Years of School School Psychology Review, 2004, Volume 33, No. 3, pp. 444-458 Teacher-Child Relationships and Children’s Success in the First Years of School Robert C. Pianta and Megan W. Stuhlman University of Virginia Abstract. This work examines associations between closeness and conflict in teacher-child relationships and children’s social and academic skills in first grade in a sample of 490 children. Assessments of teacher-child relationships were ob- tained in preschool, kindergarten, and first grade. Results demonstrate moderate correlations among teachers’ ratings of conflict and slightly lower correlations among teachers’ ratings of closeness across years. Hierarchical regression analy- ses predicted children’s skills in first grade from teacher-child relationship quality. Child gender, socioeconomic status, and preschool estimates of outcomes of inter- est were controlled statistically. Although preschool assessments of social and aca- demic skills were closely associated with individual skill differences at first grade, teacher-child relationship quality also was associated with changes in skill levels. Findings generally confirm that teacher-child relationships play a role in children’s ability to acquire the skills necessary for success in school. Concerns about children lacking the tion (Entwisle & Alexander, 1999; Ladd, Birch skills necessary for success in early elemen- & Buhs, 1999; Ladd & Burgess, 1999). Both tary school classrooms have moved to the fore of these competencies show great variation in in recent years, as the number of children en- national studies (NCES, 1999). The present countering difficulties in this setting has in- study focuses on the extent to which the qual- creased (National Education Goals Panel, ity of children’s relationships with three dif- 1997). Several major research initiatives have ferent early school teachers represents a unique explored how to facilitate school success; many source of variation in their early school adjust- have concluded that key components include ment, and predicts their social and academic the development of strong pre-academic, so- functioning at first grade. cial, and behavioral skills early in children’s The quality of children’s relationships school careers (Lyon, 2002; NICHD ECCRN, with their early school teachers is increasingly 2002a; National Center for Educational Sta- recognized as a contributor to school adapta- tistics [NCES], 1999; Ramey, Ramey, & tion (Birch & Ladd, 1997; 1998; Howes, Phillips, 1996). Emerging from this research Hamilton, & Matheson, 1994; Howes & is the impression that early and subsequent Matheson, 1992; Pianta, Steinberg, & Rollins, school functioning hinges on two related sets 1995). Similar to parent-child relationships, of competencies: those pertaining to early lit- teacher-child relationships appear to serve a eracy and language development and those regulatory function with regard to children’s associated with relationships and self-regula- social and emotional development (Greenberg, Correspondence concerning this article should be addressed to Robert C. Pianta, PhD, NICHD Study of Early Child Care, University of Virginia, 853 W. Main St., Ste. 100, Republic Plaza Annex, Charlottesville, VA 22903; E-mail: [email protected] Copyright 2004 by the National Association of School Psychologists, ISSN 0279-6015 444 Teacher-Child Relationships and Children’s Success Speltz, & Deklyen, 1993; Pianta, 1999) and ample, Burchinal and colleagues found that therefore have the potential to exert a positive teacher-reported closeness with students was or negative influence on children’s ability to positively related to growth in children’s re- succeed in school. In fact, the development of ceptive vocabulary and reading abilities from children’s early competencies in several do- preschool to second grade, specifically for chil- mains has been linked to (and is perhaps fa- dren of color and children whose parents re- cilitated by) the quality of the teacher-child ported more authoritarian attitudes, respec- relationship. Specifically, kindergarten chil- tively. dren who have highly negative relationships One potential criticism of examining with their teachers have been found to demon- teacher-child relationships is that teachers’ strate higher levels of behavior problems and perceptions of the relationships they share with lower levels of behavioral competencies 2 children is often synonymous with their per- years later as compared to their peers who have ception of children’s behavioral orientation. highly positive relationships with kindergar- However, kindergarten teachers’ perceptions ten teachers (Pianta et al., 1995). The quality of their relationships with children have been of teacher-child relationships also has predicted linked to those students’ subsequent academic changes in children’s behavioral orientation and social functioning independent of the kin- across kindergarten through first grade: Con- dergarten teachers’ reports of overall adjust- flict with the kindergarten teacher predicted ment and uniquely predict both academic and declining prosocial behavior and slightly in- behavioral skills in students through eighth creasing aggressive behavior with peers (Birch grade (Hamre & Pianta, 2001). Negativity de- & Ladd, 1998). “Secure” and “improved” scribed by teachers in their relationships with teacher-child relationships in kindergarten are children, particularly boys and children who associated with competent behavior in that had early behavior problems, was an especially classroom and fewer problems in first grade robust predictor of those children’s long-term classrooms, and dependent teacher-child rela- adjustment to school (Hamre & Pianta, 2001). tionships are associated with children’s lack These results signify the salience of relational of competence (Pianta & Nimetz, 1991). Ex- processes in the early school years, and high- tant studies also have found that children with light the fact that indicators of teacher-child chronic conflict in relationships with teachers relationships may better forecast subsequent in kindergarten and first grade demonstrate less adaptation in the classroom setting than more cooperative participation in school and lower general indicators of social competence or be- levels of school liking as compared to children havior problems (Hamre & Pianta, 2001). with high levels of teacher-child closeness An additional area of interest that has yet (Ladd & Burgess, 2001). to be thoroughly studied concerns the stability Given the associations between teacher- of the teacher-child relationship across years, child relationships and children’s school lik- and how timing influences the association be- ing and behavioral adjustment, it is not sur- tween this relationship and child outcomes. prising that children who have more conflict Extant studies have found that teachers’ per- in relationships with teachers also are less en- ceptions of conflict in their relationship with a gaged in the classroom (Ladd, Birch, & Buhs, given student is marked by moderate consis- 1999) and are at increased risk for poor aca- tency across the preschool to kindergarten tran- demic achievement (Ladd & Burgess, 2001). sition (Howes, Phillipsen, & Peisner-Feinberg, Conversely, close and supportive relationships 1999) as well as across kindergarten through with teachers have demonstrated the potential second grade (Pianta et al., 1995). In contrast, to mitigate the risk of negative outcomes for teachers’ ratings of relational closeness appear students who might otherwise have difficulty more variable across 3 years in early elemen- succeeding in school (Burchinal, Peisner- tary school (Pianta et al., 1995). Furthermore, Feinberg, Pianta, & Howes, 2002; Lynch & despite moderate continuity across teachers’ Cicchetti, 1992; Pianta et al., 1995). For ex- ratings of conflict in their relationship with a 445 School Psychology Review, 2004, Volume 33, No. 3 given child, associations between relationship ments of student-teacher relationships from quality and children’s outcomes may change preschool through first grade also was descrip- over time. Evidence indicating that earlier re- tively assessed in terms of both overall mean lationships are more closely associated with levels of conflict and closeness reported and children’s language and social development correlations between different teachers’ ratings has accumulated in recent years. For example, of relationships with a given child. Burchinal et al. (2002) found that the associa- Method tion between teacher-child closeness and children’s development is stronger when chil- Participants dren are in preschool than when they are in The 490 children and their families and early elementary school. Teacher-child close- teachers who were participants in the present ness was significantly associated with concur- study were a small subset of the National In- rent receptive vocabulary scores among pre- stitute of Child Health and Human school-aged children (e.g., 51 months and 61 Development’s Study of Early Child Care. months of age), but not among kindergarten Children and their parents and teachers were through second grade students. observed in preschool, kindergarten, and first Thus, past research has built a credible grade, and parents and teachers reported on case for the salience of interpersonal processes, children’s social and academic development at particularly the teacher-child relationship, in these times. Furthermore, children’s academic children’s early school experiences. Teacher- skill/cognitive development was tested by child relationships appear to be both contribu- trained data collectors in preschool and first tors to and indicators of children’s school ad- grade. Mothers of the study children had an justment. The present study expands previous average of 14.81 years of education (SD = work by examining the extent to which pre- 2.40), and 16% of the families were below the school, kindergarten, and first grade teachers’ poverty threshold when the children were 1 perceptions of their relationship with students (in terms of both conflict and closeness) are month old. Fifty-one percent of the children associated with those students’ social and be- were male and 14% were nonwhite. Teachers havioral outcomes. Because demographic vari- of these children had, on average, 14.66 (SD = ables have been associated with both teacher- 9.5 years) years of teaching experience and child relationship quality and child outcomes ranged from first year teachers to those with (i.e., Ladd, Birch, & Buhs, 1999), these asso- 40 years of experience. Ninety-five percent of ciations are adjusted for child gender and so- teachers were Caucasian, 2.2% were African cioeconomic status. Furthermore, to evaluate American, 1.1% were Asian, and the remain- the unique associations between teacher-child der classified themselves as “other.” Ninety- relationship quality and children’s social and six percent were female. The preschool con- academic skill development above and beyond texts were predominantly child care centers the skills that children have already acquired (74%), but children in other care settings were prior to school entry, analyses also are adjusted included as well (19% were in home-based care for preschool social and academic skills. Fi- facilities and 7% were cared for by a nally, this study incorporates ratings of nonmaternal relative). children’s skills from multiple sources. Both The children who were participants in teachers’ and mothers’ reports of children’s the present study were all those with full sets behavior problems are used to determine the of the data under consideration (i.e., mother association between teacher-child relationships and teacher reports at preschool, kindergarten, and behavior problems both in and out of the and first grade; observations in preschool and school setting. In addition, children’s academic first grade classrooms; academic test results at skills were tested and their social skills in the all assessment occasions; and parent reported classroom were observed and rated by neutral income). Thus, only about one-third of the en- observers. The stability of repeated assess- tire NICHD–SECC sample was used as a re- 446 Teacher-Child Relationships and Children’s Success sult of the majority of participants having at used to provide an estimate of their achieve- least 1 missing data point at any of the time ment in preschool. periods utilized for analyses. To determine Children’s vocabulary skills–tested. whether the current analytic subsample was Children’s vocabulary development in pre- comparable to the sample of all 1,364 partici- school and first grade was assessed using the pants, the two groups were evaluated for Picture Vocabulary subtest of the Woodcock equivalence on all covariate, predictor, and Johnson Psycho-Educational Battery–Revised outcomes variables utilized in the present Tests Cognitive Ability and Tests of Achieve- study. Out of these 23 comparisons, only four ment (WJ-R; Woodcock & Johnson, 1990). statistically significant differences were found: The WJ-R is a nationally normed assessment The present subsample had a slightly higher instrument that has widely demonstrated reli- level of maternal education (average of 14.23 ability and validity. The Picture Vocabulary years for the whole sample vs. 14.81 years in subtest measures children’s ability to recog- this sample; t = 5.36, p<.001), less likelihood nize or name pictured objects and is designed of poverty (34% below poverty vs. 16%; t = - to tap children’s verbal comprehension and 10.95, p<.001), closer teacher-child relation- crystallized intelligence (Woodcock & ships in kindergarten (mean rating of 34.22 vs. Johnson, 1990). Split-half reliability of the Pic- 34.68; t = 2.01, p = .045), and higher child ture Vocabulary subtest is .815 at age 4 and .773 vocabulary scores in preschool (mean score of at age 6 (Woodcock & Mather, 1989). Children’s 101.03 vs. 103.32; t = 3.45, p = .001). Descrip- raw score on the Vocabulary subscale was con- tive statistics for all predictor variables used verted into a standard scores (X = 100, SD = 15) in this study are presented in Table 1; those for based on the age of the child and this standard outcomes are presented in Table 2. score was used in the data analysis. Measures Children’s behavior problems–rated. Children’s mothers and preschool and first Child academic achievement–rated. grade teachers rated children’s behavior prob- In the Spring of first grade, teachers completed lems by completing the Child Behavior Check- a mock report card created for the NICHD list (CBCL; Achenbach, 1991) or the Teacher Study of Early Child Care that contained 19 Report Form (TRF; Achenbach, 1991), respec- items regarding children’s school performance tively. The CBCL is a 118-item measure of a in three domains: academic performance, work wide range of behavioral problems in school- habits, and social and emotional development. aged children. The CBCL has demonstrated All items were rated in a 5-point Likert-type sound psychometric properties: Estimates of scale with higher scores indicating better de- test-retest and interrater reliability are above veloped skills (1 = child is performing be- low grade level; 5 = child is performing be- .90 in normative samples (Achenbach, 1991). yond grade level). Teachers rated children’s The TRF is a parallel measure to the CBCL achievement in six content areas and these and is one of the most widely used measures ratings were averaged to form the Current of children’s classroom behavior (Wilson & School Performance composite. Although lim- Reschly, 1996). Its validity as a measure of ited psychometric data are available for this behavior problems has been demonstrated re- measure, the items are straightforward queries peatedly (Kamphaus & Frick, 2002) and its regarding academic performance, and the raw reliability has also been well documented: The items that comprised the composite had high average internal consistency estimate (coeffi- internal reliability (Cronbach’s alpha = .93), cient alpha) for the TRF scales is .87, and the suggesting that the construct assessed was co- average test re-test reliability coefficient is .92 hesive. Ratings of children’s academic skills (Kamphaus & Frick, 2002). Standard T-scores were not obtained in preschool; rather, for rated Internalizing and Externalizing prob- children’s obtained scores on tests of vocabu- lems from both mothers’ and teachers’ reports lary development (as described below) were were used in the present study. 447 School Psychology Review, 2004, Volume 33, No. 3 Table 1 Descriptive Statistics for Predictor Variables (N = 490) Predictor Pearson Correlations Variables 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Mean SD 1. Tested vocabulary (WJ-R) _ 103 15 2. TRF Internalizing -.01 _ 50 10 3.TRF Externalizing -.08* .48** _ 50 9.4 4. CBCL Internalizing -.06 .13** .001 _ 47 9.1 5. CBCL Externalizing -.08 .05* .25** .59** _ 52 9.3 6. Rated Social Competence (CA-PS) .18** -.52** -.61** -.10* -.16** _ 106 14 7. Observed Social Competence (ORCE) .03 -.11* -.08 -.04 -.06 .20** _ 2.9 .46 8. Conflict (STRS) -.03 .40** .69** .09* .23** -.50** -.09* _ 19 6.5 9. Closeness (STRS) .01 -.36** -.10* -.10* -.06 .41** .12** -.30** _ 45 6.2 10. Kindergarten Conflict (STRS) -.10* .10* .30** .06* .20** -.23** -.11* .29** -.06 _ 10 5.3 11. Kindergarten Closeness (STRS) .07** -.14** -.001 -.02 -.003 .12** .19** -.01 .25** -.27** _ 35 5.0 12. First Grade Conflict (STRS) -.11* .11** .36** .004 .18** -.25** -.001 .34** -.04 .36** -.001 _ 11 5.0 13. First Grade Closeness (STRS) .11** -.18** .04 -.01 .01 .23** .12** -.03 .18** -.05** .25** -.23** 34 4.8 Variables are 54-month assessments unless otherwise indicated. * p < .05, ** p <.01. Teacher-Child Relationships and Children’s Success Table 2 Descriptive Statistics for Outcome Variables (N = 490) Pearson Correlations First Grade Outcome Variables 1. 2. 3. 4. 5. 6. 7. Mean SD 1. Rated Achievement _ 3.3 .90 2. Tested Vocabulary .34** _ 108 16 3. Rated Social Competence .50** .29** _ 104 13 4. Observed Social Competence .25** .05 .28** _ 9 1.5 5. TRF Internalizing (T-score) -.24** -.10* -.40** -.10* _ 49 9.2 6.TRF Externalizing (T-score) -.21** -.16** -.54** -.07 .28** _ 50 8.7 7. CBCL Internalizing (T-score) -.03 -.01 -.15** -.05 .17** .08 _ 48 8.9 8. CBCL Externalizing (T-score) -.15* -.11* -.24** -.02 .12* .34** .59** 48 9.5 * p < .05, ** p <.01. School Psychology Review, 2004, Volume 33, No. 3 Children’s social competence–rated. in classrooms for 2–3 hours and then made At preschool age, teachers rated children’s so- qualitative ratings on 4-point scales ranging cial competence using the California Preschool from 1 = not at all characteristic to 4 = highly Social Competency Scale (Levine, Elzey, & characteristic in preschool and on 7-point Lewis, 1970). The original California Pre- scales that utilized equivalent endpoints but school Social Competency Scale contains 30 spread the scale across a wider range in first items assessing a broad range of preschool- grade. The qualitative ratings of children’s self- aged children’s social competencies. In the reliance, positive affect, and attention to ob- present study, four new items also were added jects or activities were averaged to form the to tap children’s cooperative play, rule follow- “Child Social Competence” composite in pre- ing, empathy, and aggression. All 34 items school. The qualitative ratings of child self- were rated by preschool teachers on a 4-point reliance and positive affect were averaged to scale with higher scores indicating greater lev- form this composite in first grade (as the child els of social competency. Scores ranged from attention variable was no longer coded in the 46 to 135 (M = 105, SD = 14). This scale has first grade observation). Complete descriptions been demonstrated to provide a valid assess- are available in the NICHD-SECC data col- ment of social functioning in preschool chil- lection manuals (NICHD Early Child Care dren (NICHD ECCRN, 2002b), and the 34 Research Network, 1993). Pearson correlations items have high internal reliability (Cronbach’s among these child-level qualitative ratings alpha = .88). ranged from .74 to .94. In first grade, teachers completed the Teacher-child relationship–rated. school versions of the Social Skills Question- Preschool, kindergarten, and first grade teach- naire (SSQ) from the Social Skills Rating Sys- ers rated their perceptions of their relationships tem (SSRS; Gresham & Elliott, 1990). This with the study children using the short form of instrument consists of 38 items indexing four the Student-Teacher Relationship Scale (STRS; skill areas: cooperation (e.g., keeps room neat Pianta, 2001). The short form of the STRS is a and clean without being reminded), assertion self-report instrument composed of 15 items (e.g., makes friends easily), responsibility (e.g., rated on 5-point Likert-type scales that assesses asks permission before using a family a teacher’s perception of her relationship with member’s property), and self-control (e.g., a particular student. The STRS item ratings can controls temper when arguing with other chil- be summed into two groups of items that are dren). Teachers indicated on a 3-point scale referred to as the conflict and closeness how often the child exhibited each behavior. subscales (Pianta, 2001). The conflict subscale The total score is the sum of all 38 items, with assesses the degree to which a teacher feels higher scores reflecting higher levels of per- that her relationship with a particular student ceived competence. The SSRS was normed on is characterized by negativity. The closeness a diverse, national sample of children in the scale measures the extent to which a teacher 3–5-year age range and shows high levels of feels that her relationship with the student be- internal consistency (median = .90) and test- ing assessed is characterized by warmth, af- retest reliability (.75 to .88). fection, and open communication. In terms of Children’s social competence–ob- reliability, statistically significant test-retest served. Interactions between teachers and chil- correlations over a 4-week interval and high dren were observed in the classroom at pre- internal consistency for both conflict and close- school using the Observational Record of the ness subscales have been demonstrated (Pianta, Caregiving Environment (ORCE; NICHD 2001). The predictive and concurrent validity Early Child Care Research Network, 1996) and of the STRS also has been demonstrated re- at first grade using an upward extension of the peatedly: The STRS is correlated with both ORCE, the Classroom Observation System– current and future academic skills (Hamre & First Grade (COS-1; NICHD-ECCRN, 2002a). Pianta, 2001), behavioral adjustment (Birch & Trained research assistants observed children Ladd, 1998), risk of retention (Pianta et al., 450 Teacher-Child Relationships and Children’s Success 1995), disciplinary infractions (Hamre & variate correlations were utilized to examine Pianta, 2001), and peer relations (Birth & Ladd, the stability of different teachers’ perceptions 1998). Furthermore, findings from Hamre and of both conflict and closeness in their relation- Pianta (2001) demonstrate the discriminant ships with the same student over time. Results validity of closeness and conflict scores and can be found in Table 1. Overall, teachers’ rat- suggest that use of both scales to assess posi- ings of the level of conflict they perceived in tive and negative aspects of student-teacher their relationships with a student were moder- relationships is indicated. ately correlated across time/ teacher (range r = .32 between preschool and kindergarten to r = Data Analysis .40 between preschool and first grade). Teach- ers’ ratings of the closeness they felt in rela- Bivariate correlations were conducted to tionships with students were somewhat less evaluate the stability of teachers’ relationship stable between preschool and first grade (range ratings across the preschool to first grade time r = .21 from preschool to first grade to r = .31 frame. Additionally, a descriptive examination between kindergarten and first grade), but oth- of mean levels of conflict and closeness re- erwise estimates of the stability of closeness ported over time was conducted using two re- were moderate and similar to estimates of the peated measures analysis (one for conflict and stability of conflict. Furthermore, the overall one for closeness). Regression analyses evalu- levels of teacher-reported conflict and close- ating the extent to which teacher-child relation- ness over time were evaluated using repeated ships predicted first grade skills above and measures analysis. Results indicate that over- beyond preschool skills and demographic char- acteristics also were conducted. These analy- all mean levels of both conflict and closeness ses began with a model predicting child changed very slightly but were statistically sig- competencies or problems in first grade nificantly over time (for conflict Greenhouse- from demographic factors and earlier assess- Geisser = .990. p <.01 and for closeness Green- ments of analogous competencies or prob- house-Geisser = .998. p <.01). The general lems, and then determined how much addi- trend indicated decreasing levels of both con- tional variance in outcomes could be ex- flict and closeness reported by teachers over plained by teachers’ perceptions of their re- time (see Table 1 for means). lationships with students. As teacher-child Hierarchical regression analyses were relationships were assessed repeatedly from used to address the second question of the value preschool through Grade 1, examination of added by teacher-child relationships to predic- the contributions of current and previous re- tions of children’s skill levels in first grade lationships with teachers was possible by add- above and beyond assessments of those skills ing these indicators in temporal sequence. Fi- in preschool. In each regression, predictor vari- nally, interactions between time-averaged STRS ables were entered in a series of six steps. In conflict and closeness scores and earlier com- the first step, demographic covariates were petencies or problems, gender, and poverty sta- entered as a block (gender and poverty status). tus in predicting developmental outcomes in In the second step, the 54-month version of first grade were examined. These last analy- the outcome being predicted was entered. The ses examined whether, in general, children’s STRS conflict and closeness scores at 54 relational competencies with teachers moder- months, kindergarten, and first grade were en- ated the association between demographic risk tered as blocks in the third, fourth, and fifth factors and success in first grade. steps, respectively. Finally, interactions be- tween the three covariates and an average of Results the 54-month through first grade conflict and Two broad questions were addressed in closeness scores were entered in a sixth block. this study. The first question concerned the sta- However, because only 1 interaction term of bility of teacher-child relationships across the the 48 interactions that were entered into analy- preschool to first grade interval. Pearson bi- ses (6 in each of the 8 regressions) was statis- 451 School Psychology Review, 2004, Volume 33, No. 3 tically significant, it was ascertained that over- Behavior Problems all, interactions did not add meaningful infor- Both mothers’ and teachers’ ratings of mation to prediction analyses. Thus, the inter- internalizing and externalizing problems were action blocks are not reported. Results of hier- investigated to provide a more complete picture archical regression analyses can be found in of how teacher-child relationships were associ- Table 3 and are summarized below. ated with children’s behavioral development in- Overall, the demographic covariate side and outside of the school setting. Mother- block, which included poverty status and gen- rated child internalizing behavior at first grade der, accounted for significant variance in the was predicted by preschool and first grade following child outcomes: teachers’ ratings of teachers reports of closeness. Specifically, less children’s achievement, externalizing behav- closeness reported by the preschool or first ior, and social competence, and tested vocabu- grade teacher was linked to higher levels of lary development. Similarly, the second mothers’ reports of internalizing behavior. covariate block, which included the 54-month In terms of externalizing behavior rated assessment of child outcomes, accounted for by mothers, more teacher-rated relational con- significant variance in child outcomes at the flict at first grade was significantly related to time that it was entered in all analyses except mother’s reports of children’s externalizing when observed social competence was the out- behavior. Previous teachers’ ratings of close- come under consideration. When all variables ness and conflict were not associated with were included, the 54-month assessments of mothers’ ratings of child externalizing in first child outcomes were significant predictors for grade. teacher-rated achievement, tested vocabulary With regard to first grade teachers’ rat- development, mother-rated internalizing be- ings of child problem behavior, their concur- havior, teacher- and mother-rated externaliz- rent perceptions of closeness and conflict in ing behavior, and teacher-rated social compe- their relationships with students predicted their tence in the expected directions (i.e. higher perception of the child’s internalizing prob- levels of each construct at 54 months was as- lems: More conflict and less closeness were sociated with higher levels of that construct in associated with higher levels of perceived child first grade). Results for variance attributable internalizing problems. No associations with to relational conflict and closeness at 54 prior teacher-child conflict or closeness were months, kindergarten, and first grade, adjusted found for teacher-rated internalizing at first for the two covariate blocks, are summarized grade. Regarding externalizing problems, below for each outcome examined. teacher-reported relational conflict with chil- Academic Performance dren in both kindergarten and first grade were associated with first grade teachers’ ratings of When predicting teacher ratings of externalizing in the expected directions. achievement, both conflict and closeness in first grade were significant predictors. First Social Competence grade teachers rated children’s achievement more highly for those children with whom they Both teachers’ and trained observers’ reported sharing a closer relationship, and as- ratings of children’s social competence were signed lower achievement ratings to those chil- investigated. Kindergarten and first grade dren with whom they felt more conflict. No teacher’s reports of conflict in their relation- other indicators of the teacher-child relation- ships with children were inversely related to ship were statistically significant in predicting the first grade teachers’ reports of children’s rated achievement. Regarding tested vocabu- social competence: Higher conflict was asso- lary development in first grade, preschool ciated with lower social competence. In addi- through first grade teachers’ perceptions of tion, as first grade teachers reported more their relationships with students did not dem- closeness in their relationship with a given onstrate significant prediction. child, they also reported higher levels of so- 452 Teacher-Child Relationships and Children’s Success Table 3 Prediction of Social and Academic Outcomes in First Grade (N = 490) First Grade Outcomes Rated Tested TRF TRF CBCL CBCL Rated Observed Achieve- Vocab- Internal- External- Internal- External- Social Social ment ulary izing izing izing izing Comp. Comp. Model R .17 .48 .11 .61 .34 .52 .47 .57 Model F 10.8** 49.5** 6.2** 83.5** 28.8** 57.1** 48.8** 3.3** Step 1: Demographics ( R ) .04** .13** .01 .03** .001 .01 .04** .01 Gender (high = female) -.10* -.15** .06 .07* .003 .03 -.09** -.05 Poverty (high = poor) -.08 -.07 .03 .13** .06 .001 -.21** -.02 Step 2: 54-m indicator of Grade 1 .05** .34** .11* .21** .32** .48** .13** .001 outcome ( R ) .19** .61** .06 .21** .55** .67** .19** .02 Step 3: 54m STRS ( R ) .02** .001 .002 .01** .01** .004 .01* .01 Conflict -.05 .05 -.01 -.01 .01 .08 -.02 -.03 Closeness .04 -.02 .01 .05 -.10* -.02 -.004 .06 Step 4: Kindergarten STRS ( R ) .01 .01 .005 .069** .01 .006 .05** .02** Conflict -.03 .03 .03 .12** .06 .02 -.11** -.03 Closeness .03 .06 .04 .02 .05 .07 .02 .09 Step 5: First Grade STRS ( R ) .06** .001 .08** .29** .01* .02** .25** .020** Conflict -.10* -.03 .19** .65** -.05 .14** -.38** -.09 Closeness .22** .02 -.19** .03 -.07 -.04 .32** .10* Cell values are standardized β coefficients derived from the final model. *p < .05, **p < .01. School Psychology Review, 2004, Volume 33, No. 3 cial competence in that child. Furthermore, relationships and both mothers’ and neutral when entered as a block, preschool teachers’ observers’ ratings of children. reports of their relationship with children ac- Regarding the descriptive examination counted for significant variance in first grade of stability in different teachers’ perceptions teachers’ ratings of children’s social compe- of conflict and closeness in their relationships tence, but when subsequent predictors were with a given child, findings indicated moder- included in the final model, neither of the kin- ate stability in conflict and slightly less in close- dergarten relationship indicators demonstrated ness across the 54-month to first grade time a significant association with ratings of social period. This finding is consistent with previ- competence in first grade. Finally, variance in ous studies of the stability of teacher-child re- trained observers’ ratings of children’s social lationships from kindergarten to first grade competence in the first grade classroom was (Pianta et al., 1995) and from preschool to kin- predicted by first grade teachers’ perceptions dergarten (Howes et al., 2000). Previous and of closeness in their relationship with a child present findings suggest that relational conflict in the expected directions. Although at entry with a given child is more stable across teach- kindergarten teacher-child closeness was pre- ers (perhaps related to stable child character- dictive of first grade observed social compe- istics such as temperament), whereas relational tence (∆ R for the block = .02, p = .03), when closeness may depend in greater part on the all predictors were entered in the full model, goodness-of-fit between the interpersonal this link was no longer significant. styles of teachers and children. The second set of descriptive findings Discussion pertained to consistency in the mean levels of teacher-reported conflict and closeness from The overall pattern of results in the preschool through first grade. Results suggest present study provides preliminary guidance a slight trend towards teachers reporting less for future investigations of how teacher-child intense relationships with children in both relationship quality is related to children’s so- negative and positive dimensions with increas- cial and academic development in the early ing child age. This finding may reflect a de- school years. Preliminary descriptive findings creasing emphasis on teacher-child relation- indicate that different teachers’ perceptions of ships in the classroom from preschool to early their relationships with a given child are moder- elementary school. Previous studies have sug- ately stable from preschool through first grade. gested a similar dynamic across this time pe- In addition, there was a small but statistically riod (Johnson, Gallagher, Cook, & Wong, significant trend towards teachers reporting decreasing levels of both conflict and close- 1995; Rusher, McGrevin, & Lambiotte, 1992; ness from preschool through first grade. Spodek, 1988; Vartuli, 1999). With regard to associations between teacher- With regard to associations between child relationships and child outcomes, af- teacher-child relationships and first grade com- ter adjusting for gender, socioeconomic sta- petencies, estimates of prior social and aca- tus, and children’s prior skill levels, several demic skills were consistently significant cor- indicators of teacher-child relationships relates of first grade assessments of those out- made small to moderate contributions to pre- comes, with the exception of observed social diction of social and teacher-rated academic competence. In several cases (tested vocabu- skills in first grade. Although the associations lary development, externalizing behavior, and between teacher-child relationships and rated social competence), demographic vari- children’s outcomes tended to be greatest when ables also accounted for significant variance the outcomes and relationships were rated by in first grade skills. However, above and be- the same person (i.e., the teacher), not all of yond the effects of these factors, teachers’ per- the findings can be attributed solely to rater ceptions of the student-teacher relationship on effects. In fact, small but significant associa- at least one occasion between preschool and tions were found between teachers’ ratings of first grade had small but significant relations 454 Teacher-Child Relationships and Children’s Success to all child outcomes except tested vocabulary found that preschool or childcare teacher-child development. Although many assessments of relationship quality is more closely associated teacher-child relationships were not signifi- with children’s development over time than cantly associated with child outcomes, these quality of relationships with elementary school preliminary findings add to our understanding teachers (Burchinal et al., 2002). The present of how teacher-child relationships are associ- findings indicate that although preschool and ated with changes in individual differences in kindergarten teacher-child relationships dem- children’s social and academic competence onstrate small associations with some first over time. Previously, work by Birch and Ladd grade skills (rated achievement, teacher-re- (1998) demonstrated that kindergarten teach- ported externalizing behavior, mother-reported ers’ perceptions of conflict in their relationship internalizing behavior, and rated social com- with students accounted for significant vari- petence), both the number of significant effects ance in children’s prosocial behavior in first and the magnitude of those effects is greater grade. The current study extended that finding for concurrent teachers’ ratings of relational by expanding the time interval in which quality. First grade teachers’ ratings of rela- children’s behavioral development was stud- tional quality were significantly related to all ied and by using repeated assessments of the first grade outcomes except tested vocabulary teacher-child relationship. development, and in all cases accounted for a An additional pattern in the findings that greater proportion of variance in those out- merits discussion is the association between comes than earlier teachers’ reports of relation- teacher-child relationships and mother- and ships, even though those first grade relation- teacher-rated behavior problems. Regardless ship indicators were entered into the analytic of rater and controlling for prior problem be- model last. haviors, demographic variables, and prior Several important limitations to this teacher-child conflict and closeness, contem- work also merit review. First, this work clearly poraneous conflict (or lack of closeness) be- focuses on the contributions of the teacher- tween teachers and children is linked to child relationships to predicting child out- children’s externalizing (or internalizing) prob- comes. In adopting such a focus, consider- lems in first grade. The fact that changes in ation of other aspects of the teachers’ pres- mothers’ reports of child behavior problems is ence in the classroom and/or other aspects associated with quality of the teacher-child of her or his interactions with children is relationship suggests some degree of crossover omitted. As a result, it cannot be stated con- between home and school contexts. This is clusively that teacher’s perceptions of rela- consistent with previous findings linking ob- tionship quality is uniquely predictive, as served quality of classroom processes to this could be highly collinear with other as- changes in mothers’ perceptions of their pects of the teachers’ behavior in the class- children’s internalizing behavior from pre- room (including but not limited to her or his school through first grade (NICHD ECCRN, teaching style, classroom management style, 2003). or the overall emotional tone of the class- As this study incorporated three assess- room). To conclude that it is this aspect of ments of the teacher-child relationship over the teacher’s presence and not others that time, issues pertaining to the timing of this re- contributes to changes in individual differ- lationship were available for consideration. ences in children’s competencies, future Earlier work has suggested that teacher-child work may include other classroom process relationships may have the greatest effect on variables in addition to teachers’ perceptions children’s competency in that specific class- of their relationships with children so that the room, particularly among high-risk children overlap between these constructs can be fur- (i.e., that teacher-child relationships have pri- ther explored. Furthermore, the possibility ex- marily concurrent effects; see Pianta, Nimetz, ists that the teacher’s perception of the teacher- & Bennett, 1997). Conversely, others have child relationship is more of a marker of the 455 School Psychology Review, 2004, Volume 33, No. 3 child’s overall competence than an indepen- advantages of adding an assessment of the dent construct. A fruitful area for future re- teacher-child relationship to the standard test search would experimentally test whether im- batteries used by school psychologists. Add- provements in the teacher-child relationship ing this dyadic component to evaluations of result in changes in children’s patterns of be- children with school problems could poten- havior and success in school. tially contribute valuable information not con- Despite these limitations, this work dem- tained in traditional child-centered behavior onstrates that the interpersonal aspects of checklists. These results also have implications children’s experience make a difference in their for early intervention efforts and school policy ability to develop competencies in the early in that they suggest a focus on relational pro- years of school. Further work may include not cesses between teachers and students as a only the teacher-child relationship, but also mechanism for enhancing opportunities for other aspects of the social environment that success in school. children are exposed to as well to get a richer Footnotes picture of the interpersonal context of early elementary school and how that context 1 Of note, the 1 interaction term (out of 36) functions as a medium in which children that accounted for statistically significant variance develop. Examples of such initiatives in- in a child outcome was in predicting teacher-rated academic skills in first grade. In that case, the 54- clude Ladd and colleagues’ (1999) work, month closeness x 54-month vocabulary skills in- which utilized assessments of peer accep- teraction term explained significant variance in tance, children’s best friendships, and teachers ratings of academic skills in first grade. teacher-child relationship quality within the Exploration of this interaction revealed that there kindergarten year and found that social dy- was a stronger association between closeness and namics with both peers and teachers were rated achievement in first grade among children important aspects of a model predicting whose academic achievement was above the mean children’s academic achievement. Clearly, at 54 months. additional work is needed in this area to de- References velop a clear picture of the effect of rela- tional processes in the classroom. Continued Achenbach, T. (1991). Manual for the Child Behavior work in this direction will provide teachers, Checklist/4-18 and 1991 Profile. Burlington, VT: Au- school psychologists, and other school person- thor. Birch, S. H., & Ladd, G. W. (1997). The teacher-child re- nel with a better understanding of how to fa- lationship and children’s early school adjustment. Jour- cilitate positive outcomes for all children in nal of School Psychology, 35, 61-79. school. Birch, S. H., & Ladd, G. W. (1998). Children’s interper- In sum, although the associations be- sonal behaviors and the teacher-child relationship. Developmental Psychology, 34, 934-946. tween teacher-child relationships and Burchinal, M. R., Peisner-Feinberg, E., Pianta, R. C., & children’s development tended to be small and Howes, C. (2002). Development of academic skills were not significant in all cases, these re- from preschool through second grade: Family and sults do suggest a salient relational compo- classroom predictors of developmental trajectories. Journal of School Psychology, 40, 415-436. nent to children’s early school competence. Entwisle, D. R., & Alexander, K. (1999). Early schooling Despite children’s general tendencies to- and social stratification. In R. C. Pianta & M. J. Cox wards certain relational styles with teach- (Eds.), The transition to kindergarten (pp. 13-38. Bal- ers inferred from the moderate stability timore, MD: Paul H. Brookes. Greenberg, M. T., Speltz, M. L., & Deklyen, M. (1993). across different teachers’ reports of their The role of attachment in the early development of relationships with a given child across the disruptive behavior problems. Development and Psy- preschool to first grade time period, each chopathology, 5, 191-213. new relationship may represent an opportu- Gresham, F., & Elliott, S. (1990). Social Skills Rating System. Circle Pines, MN: American Guidance Ser- nity to introduce unique variance in the devel- vice. opment of social and academic skills. Research Hamre, B. K., & Pianta, R. C. (2001). Early teacher-child findings such as these suggest the possible relationships and the trajectory of children’s school 456 Teacher-Child Relationships and Children’s Success outcomes through eighth grade. Child Development, NICHD Early Child Care Research Network. (2002a). The 72, 625-638. relation of global first grade classroom environment Howes, C., Hamilton, C. E., & Matheson, C. C. (1994). to structural classroom features, teacher, and student Children’s relationships with peers: Differential asso- behaviors. Elementary School Journal, 102, 367-387. ciations with aspects of the teacher-child relationship. NICHD Early Child Care Research Network. (2002b). Child Development, 65, 253-263. Child care structure—process—outcome: Direct and Howes, C., & Matheson, C. C. (1992). Contextural con- indirect effects of child care quality on young children’s straints on the concordance of mother-child and development. Psychological Science, 13, 199-206. teacher-child relationships. In R. C. Pianta (Ed.), Be- NICHD Early Child Care Research Network. (2003). So- yond the parent: The role of other adults in children’s cial functioning in first grade: Associations with ear- lives (Vol. 57, pp. 25-40). San Francisco: Jossey-Bass. lier home and child care predictors and with current Howes, C., Phillipsen, L. C., & Peisner-Feinberg, E. classroom experiences. Child Development, 74, 1639- (2000). The consistency of perceived teacher-child re- 1662. lationships between preschool and kindergarten. Jour- Pianta, R. C. (1999). Enhancing relationships between nal of School Psychology, 38, 113-132. children and teachers. Washington, DC: American Johnson, L. J., Gallagher, R. J., Cook, M., & Wong, P. Psychological Association. (1995). Critical skills for kindergarten: Perceptions Pianta, R.C. (2001). Student-Teacher Relationship Scale: from kindergarten teachers. Journal of Early Interven- Professional manual. Odessa, FL: Psychological As- tion, 19, 315-327. sessment Resources. Kamphaus, R. W. & Frick, P .J. (2002). Clinical assess- Pianta, R. C., & Nimetz, S. (1991). Relationships between ment of children and adolescent personality and be- teachers and children: Associations with behavior at havior. Boston: Allyn & Bacon. home and in the classroom. Journal of Applied Devel- Ladd, G. W., Birch, S. H., Buhs, E. S. (1999). Children’s opmental Psychology, 12, 379-393. social and scholastic lives in kindergarten: Related Pianta, R. C., Nimetz, S. L., & Bennett, E. (1997). Mother- spheres of influence? Child Development, 70, 1373- child relationships, teacher-child relationships, and 1400. school outcomes in preschool and kindergarten. Early Ladd, G. W., & Burgess, K. B. ( 1999). Charting the rela- Childhood Research Quarterly, 12, 263-280. tionship trajectories of aggressive, withdrawn, and Pianta, R. C., Steinberg, M. S., & Rollins, K. (1995). The aggressive/withdrawn children during early grade first two years of school: Teacher-child relationships school. Child Development, 70, 910-929. and deflections in children’s classroom adjustment. Ladd, G. W., & Burgess, K. B. ( 2001). Do relational risk Development & Psychopathology, 7, 295-312. and protective factors moderate the links between Ramey, S. L., Ramey, C. T., & Phillips, M. M. (1996). childhood aggression and early psychological and Head Start children’s entry into public school: An in- school adjustment? Child Development, 72, 1579-1601. terim report on the National Head Start-Public School Levine, S., Elzey, F., & Lewis, M. (1970). California Pre- Early Childhood Transition Demonstration Study. school Social Competency Scale. Palo Alto, CA: Con- Washington, DC: U.S. Department of Health and Hu- sulting Psychologists Press. man Services. Lynch, M., & Cicchetti, D. (1992). Maltreated children’s Rusher, A. S., McGrevin, C. Z., & Lambiotte, J. G. (1992). reports of relatedness to their teachers. In R. C. Pianta Belief systems of early childhood teachers and their (Ed.), Beyond the parent: The role of other adults in principals regarding early childhood education. Early children’s lives (Vol. 57, pp. 81-108). San Francisco: Childhood Research Quarterly, 7, 227-296. Jossey-Bass. Spodek, B. (1988). The implicit theories of early child- Lyon, G. R. (2002). Reading development, reading diffi- hood teachers. Early Childhood Development and culties, and reading instruction: Educational and public Care, 29, 197-208. health issues. Journal of School Psychology, 40, 1-4 Vartuli, S. (1999). How early childhood teacher beliefs National Center for Education Statistics. (1999). America’s vary across grade level. Early Childhood Research kindergartners. Washington, DC: National Center for Quarterly, 14, 489-514. Education Statistics. Wilson, M. S. & Reschly, D. J. (1996). Assessment in National Education Goals Panel. (1997).Building a na- school psychology training and practice. School Psy- tion of learners. Washington, DC: Author. chology Review, 25, 9-23 National Institute for Child Health and Human Develop- Woodcock, R. W., & Johnson, M. B. (1990). Woodcock- ment (NICHD) Early Child Care Research Network. Johnson Psycho-Educational Battery—Revised. Allen, (1993). The NICHD study of early child care: A com- TX: DLM Teaching Resources. prehensive longitudinal study of young children’s lives. Woodcock, R. W., & Mather, N. (1989). Woodcock ERIC Clearinghouse on Elementary and Early Child Johnson—Revised Tests of Cognitive Ability standard Education, ED 353-087. battery: Examiner’s manual. In R. W. Woodcock & NICHD Early Child Care Research Network. (1996). Char- M. B. Johnson, Woodcock-Johnson Psycho-Educa- acteristics of infant child care: Factors contributing to tional Battery—Revised. Allen, TX: DLM Teaching positive caregiving. Early Childhood Research Quar- Resources. terly, 11, 269-306. 457 School Psychology Review, 2004, Volume 33, No. 3 Robert C. Pianta, PhD, is a professor in The Curry School of Education at the University of Virginia and holds the Novartis U.S. Foundation Chair in Education. He is particularly interested in how relationships with teachers and parents, as well as classroom quality, can help improve outcomes for at-risk children and youth. Dr. Pianta is a principal investiga- tor on the National Institute of Child Health and Human Development Study of Early Child Care and Youth Development, a senior investigator with the National Center for Early Development and Learning, and editor of the Journal of School Psychology. Megan W. Stuhlman, PhD, graduated from The Curry School of Education’s Programs in Clinical and School Psychology at the University of Virginia in 2004 and is currently a research fellow at the University of Miami-School of Medicine. Her research interests include examining how developmental trajectories of children and youth at-risk for diffi- culties in home and school settings are influenced by relationships between children and parental and nonparental adults. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png School Psychology Review Taylor & Francis

Teacher-Child Relationships and Children&apos;s Success in the First Years of School

Pianta, Robert C.; Stuhlman, Megan W.
School Psychology Review , Volume 33 (3): 15 – Sep 1, 2004
Download PDF
15 pages

Loading next page...
 
/lp/taylor-francis/teacher-child-relationships-and-children-apos-s-success-in-the-first-xQoBlGi291

References (36)

Publisher
Taylor & Francis
Copyright
Copyright 2004 by the National Association of School Psychologists
ISSN
0279-6015
eISSN
2372-966X
DOI
10.1080/02796015.2004.12086261
Publisher site
See Article on Publisher Site

Abstract

School Psychology Review, 2004, Volume 33, No. 3, pp. 444-458 Teacher-Child Relationships and Children’s Success in the First Years of School Robert C. Pianta and Megan W. Stuhlman University of Virginia Abstract. This work examines associations between closeness and conflict in teacher-child relationships and children’s social and academic skills in first grade in a sample of 490 children. Assessments of teacher-child relationships were ob- tained in preschool, kindergarten, and first grade. Results demonstrate moderate correlations among teachers’ ratings of conflict and slightly lower correlations among teachers’ ratings of closeness across years. Hierarchical regression analy- ses predicted children’s skills in first grade from teacher-child relationship quality. Child gender, socioeconomic status, and preschool estimates of outcomes of inter- est were controlled statistically. Although preschool assessments of social and aca- demic skills were closely associated with individual skill differences at first grade, teacher-child relationship quality also was associated with changes in skill levels. Findings generally confirm that teacher-child relationships play a role in children’s ability to acquire the skills necessary for success in school. Concerns about children lacking the tion (Entwisle & Alexander, 1999; Ladd, Birch skills necessary for success in early elemen- & Buhs, 1999; Ladd & Burgess, 1999). Both tary school classrooms have moved to the fore of these competencies show great variation in in recent years, as the number of children en- national studies (NCES, 1999). The present countering difficulties in this setting has in- study focuses on the extent to which the qual- creased (National Education Goals Panel, ity of children’s relationships with three dif- 1997). Several major research initiatives have ferent early school teachers represents a unique explored how to facilitate school success; many source of variation in their early school adjust- have concluded that key components include ment, and predicts their social and academic the development of strong pre-academic, so- functioning at first grade. cial, and behavioral skills early in children’s The quality of children’s relationships school careers (Lyon, 2002; NICHD ECCRN, with their early school teachers is increasingly 2002a; National Center for Educational Sta- recognized as a contributor to school adapta- tistics [NCES], 1999; Ramey, Ramey, & tion (Birch & Ladd, 1997; 1998; Howes, Phillips, 1996). Emerging from this research Hamilton, & Matheson, 1994; Howes & is the impression that early and subsequent Matheson, 1992; Pianta, Steinberg, & Rollins, school functioning hinges on two related sets 1995). Similar to parent-child relationships, of competencies: those pertaining to early lit- teacher-child relationships appear to serve a eracy and language development and those regulatory function with regard to children’s associated with relationships and self-regula- social and emotional development (Greenberg, Correspondence concerning this article should be addressed to Robert C. Pianta, PhD, NICHD Study of Early Child Care, University of Virginia, 853 W. Main St., Ste. 100, Republic Plaza Annex, Charlottesville, VA 22903; E-mail: [email protected] Copyright 2004 by the National Association of School Psychologists, ISSN 0279-6015 444 Teacher-Child Relationships and Children’s Success Speltz, & Deklyen, 1993; Pianta, 1999) and ample, Burchinal and colleagues found that therefore have the potential to exert a positive teacher-reported closeness with students was or negative influence on children’s ability to positively related to growth in children’s re- succeed in school. In fact, the development of ceptive vocabulary and reading abilities from children’s early competencies in several do- preschool to second grade, specifically for chil- mains has been linked to (and is perhaps fa- dren of color and children whose parents re- cilitated by) the quality of the teacher-child ported more authoritarian attitudes, respec- relationship. Specifically, kindergarten chil- tively. dren who have highly negative relationships One potential criticism of examining with their teachers have been found to demon- teacher-child relationships is that teachers’ strate higher levels of behavior problems and perceptions of the relationships they share with lower levels of behavioral competencies 2 children is often synonymous with their per- years later as compared to their peers who have ception of children’s behavioral orientation. highly positive relationships with kindergar- However, kindergarten teachers’ perceptions ten teachers (Pianta et al., 1995). The quality of their relationships with children have been of teacher-child relationships also has predicted linked to those students’ subsequent academic changes in children’s behavioral orientation and social functioning independent of the kin- across kindergarten through first grade: Con- dergarten teachers’ reports of overall adjust- flict with the kindergarten teacher predicted ment and uniquely predict both academic and declining prosocial behavior and slightly in- behavioral skills in students through eighth creasing aggressive behavior with peers (Birch grade (Hamre & Pianta, 2001). Negativity de- & Ladd, 1998). “Secure” and “improved” scribed by teachers in their relationships with teacher-child relationships in kindergarten are children, particularly boys and children who associated with competent behavior in that had early behavior problems, was an especially classroom and fewer problems in first grade robust predictor of those children’s long-term classrooms, and dependent teacher-child rela- adjustment to school (Hamre & Pianta, 2001). tionships are associated with children’s lack These results signify the salience of relational of competence (Pianta & Nimetz, 1991). Ex- processes in the early school years, and high- tant studies also have found that children with light the fact that indicators of teacher-child chronic conflict in relationships with teachers relationships may better forecast subsequent in kindergarten and first grade demonstrate less adaptation in the classroom setting than more cooperative participation in school and lower general indicators of social competence or be- levels of school liking as compared to children havior problems (Hamre & Pianta, 2001). with high levels of teacher-child closeness An additional area of interest that has yet (Ladd & Burgess, 2001). to be thoroughly studied concerns the stability Given the associations between teacher- of the teacher-child relationship across years, child relationships and children’s school lik- and how timing influences the association be- ing and behavioral adjustment, it is not sur- tween this relationship and child outcomes. prising that children who have more conflict Extant studies have found that teachers’ per- in relationships with teachers also are less en- ceptions of conflict in their relationship with a gaged in the classroom (Ladd, Birch, & Buhs, given student is marked by moderate consis- 1999) and are at increased risk for poor aca- tency across the preschool to kindergarten tran- demic achievement (Ladd & Burgess, 2001). sition (Howes, Phillipsen, & Peisner-Feinberg, Conversely, close and supportive relationships 1999) as well as across kindergarten through with teachers have demonstrated the potential second grade (Pianta et al., 1995). In contrast, to mitigate the risk of negative outcomes for teachers’ ratings of relational closeness appear students who might otherwise have difficulty more variable across 3 years in early elemen- succeeding in school (Burchinal, Peisner- tary school (Pianta et al., 1995). Furthermore, Feinberg, Pianta, & Howes, 2002; Lynch & despite moderate continuity across teachers’ Cicchetti, 1992; Pianta et al., 1995). For ex- ratings of conflict in their relationship with a 445 School Psychology Review, 2004, Volume 33, No. 3 given child, associations between relationship ments of student-teacher relationships from quality and children’s outcomes may change preschool through first grade also was descrip- over time. Evidence indicating that earlier re- tively assessed in terms of both overall mean lationships are more closely associated with levels of conflict and closeness reported and children’s language and social development correlations between different teachers’ ratings has accumulated in recent years. For example, of relationships with a given child. Burchinal et al. (2002) found that the associa- Method tion between teacher-child closeness and children’s development is stronger when chil- Participants dren are in preschool than when they are in The 490 children and their families and early elementary school. Teacher-child close- teachers who were participants in the present ness was significantly associated with concur- study were a small subset of the National In- rent receptive vocabulary scores among pre- stitute of Child Health and Human school-aged children (e.g., 51 months and 61 Development’s Study of Early Child Care. months of age), but not among kindergarten Children and their parents and teachers were through second grade students. observed in preschool, kindergarten, and first Thus, past research has built a credible grade, and parents and teachers reported on case for the salience of interpersonal processes, children’s social and academic development at particularly the teacher-child relationship, in these times. Furthermore, children’s academic children’s early school experiences. Teacher- skill/cognitive development was tested by child relationships appear to be both contribu- trained data collectors in preschool and first tors to and indicators of children’s school ad- grade. Mothers of the study children had an justment. The present study expands previous average of 14.81 years of education (SD = work by examining the extent to which pre- 2.40), and 16% of the families were below the school, kindergarten, and first grade teachers’ poverty threshold when the children were 1 perceptions of their relationship with students (in terms of both conflict and closeness) are month old. Fifty-one percent of the children associated with those students’ social and be- were male and 14% were nonwhite. Teachers havioral outcomes. Because demographic vari- of these children had, on average, 14.66 (SD = ables have been associated with both teacher- 9.5 years) years of teaching experience and child relationship quality and child outcomes ranged from first year teachers to those with (i.e., Ladd, Birch, & Buhs, 1999), these asso- 40 years of experience. Ninety-five percent of ciations are adjusted for child gender and so- teachers were Caucasian, 2.2% were African cioeconomic status. Furthermore, to evaluate American, 1.1% were Asian, and the remain- the unique associations between teacher-child der classified themselves as “other.” Ninety- relationship quality and children’s social and six percent were female. The preschool con- academic skill development above and beyond texts were predominantly child care centers the skills that children have already acquired (74%), but children in other care settings were prior to school entry, analyses also are adjusted included as well (19% were in home-based care for preschool social and academic skills. Fi- facilities and 7% were cared for by a nally, this study incorporates ratings of nonmaternal relative). children’s skills from multiple sources. Both The children who were participants in teachers’ and mothers’ reports of children’s the present study were all those with full sets behavior problems are used to determine the of the data under consideration (i.e., mother association between teacher-child relationships and teacher reports at preschool, kindergarten, and behavior problems both in and out of the and first grade; observations in preschool and school setting. In addition, children’s academic first grade classrooms; academic test results at skills were tested and their social skills in the all assessment occasions; and parent reported classroom were observed and rated by neutral income). Thus, only about one-third of the en- observers. The stability of repeated assess- tire NICHD–SECC sample was used as a re- 446 Teacher-Child Relationships and Children’s Success sult of the majority of participants having at used to provide an estimate of their achieve- least 1 missing data point at any of the time ment in preschool. periods utilized for analyses. To determine Children’s vocabulary skills–tested. whether the current analytic subsample was Children’s vocabulary development in pre- comparable to the sample of all 1,364 partici- school and first grade was assessed using the pants, the two groups were evaluated for Picture Vocabulary subtest of the Woodcock equivalence on all covariate, predictor, and Johnson Psycho-Educational Battery–Revised outcomes variables utilized in the present Tests Cognitive Ability and Tests of Achieve- study. Out of these 23 comparisons, only four ment (WJ-R; Woodcock & Johnson, 1990). statistically significant differences were found: The WJ-R is a nationally normed assessment The present subsample had a slightly higher instrument that has widely demonstrated reli- level of maternal education (average of 14.23 ability and validity. The Picture Vocabulary years for the whole sample vs. 14.81 years in subtest measures children’s ability to recog- this sample; t = 5.36, p<.001), less likelihood nize or name pictured objects and is designed of poverty (34% below poverty vs. 16%; t = - to tap children’s verbal comprehension and 10.95, p<.001), closer teacher-child relation- crystallized intelligence (Woodcock & ships in kindergarten (mean rating of 34.22 vs. Johnson, 1990). Split-half reliability of the Pic- 34.68; t = 2.01, p = .045), and higher child ture Vocabulary subtest is .815 at age 4 and .773 vocabulary scores in preschool (mean score of at age 6 (Woodcock & Mather, 1989). Children’s 101.03 vs. 103.32; t = 3.45, p = .001). Descrip- raw score on the Vocabulary subscale was con- tive statistics for all predictor variables used verted into a standard scores (X = 100, SD = 15) in this study are presented in Table 1; those for based on the age of the child and this standard outcomes are presented in Table 2. score was used in the data analysis. Measures Children’s behavior problems–rated. Children’s mothers and preschool and first Child academic achievement–rated. grade teachers rated children’s behavior prob- In the Spring of first grade, teachers completed lems by completing the Child Behavior Check- a mock report card created for the NICHD list (CBCL; Achenbach, 1991) or the Teacher Study of Early Child Care that contained 19 Report Form (TRF; Achenbach, 1991), respec- items regarding children’s school performance tively. The CBCL is a 118-item measure of a in three domains: academic performance, work wide range of behavioral problems in school- habits, and social and emotional development. aged children. The CBCL has demonstrated All items were rated in a 5-point Likert-type sound psychometric properties: Estimates of scale with higher scores indicating better de- test-retest and interrater reliability are above veloped skills (1 = child is performing be- low grade level; 5 = child is performing be- .90 in normative samples (Achenbach, 1991). yond grade level). Teachers rated children’s The TRF is a parallel measure to the CBCL achievement in six content areas and these and is one of the most widely used measures ratings were averaged to form the Current of children’s classroom behavior (Wilson & School Performance composite. Although lim- Reschly, 1996). Its validity as a measure of ited psychometric data are available for this behavior problems has been demonstrated re- measure, the items are straightforward queries peatedly (Kamphaus & Frick, 2002) and its regarding academic performance, and the raw reliability has also been well documented: The items that comprised the composite had high average internal consistency estimate (coeffi- internal reliability (Cronbach’s alpha = .93), cient alpha) for the TRF scales is .87, and the suggesting that the construct assessed was co- average test re-test reliability coefficient is .92 hesive. Ratings of children’s academic skills (Kamphaus & Frick, 2002). Standard T-scores were not obtained in preschool; rather, for rated Internalizing and Externalizing prob- children’s obtained scores on tests of vocabu- lems from both mothers’ and teachers’ reports lary development (as described below) were were used in the present study. 447 School Psychology Review, 2004, Volume 33, No. 3 Table 1 Descriptive Statistics for Predictor Variables (N = 490) Predictor Pearson Correlations Variables 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Mean SD 1. Tested vocabulary (WJ-R) _ 103 15 2. TRF Internalizing -.01 _ 50 10 3.TRF Externalizing -.08* .48** _ 50 9.4 4. CBCL Internalizing -.06 .13** .001 _ 47 9.1 5. CBCL Externalizing -.08 .05* .25** .59** _ 52 9.3 6. Rated Social Competence (CA-PS) .18** -.52** -.61** -.10* -.16** _ 106 14 7. Observed Social Competence (ORCE) .03 -.11* -.08 -.04 -.06 .20** _ 2.9 .46 8. Conflict (STRS) -.03 .40** .69** .09* .23** -.50** -.09* _ 19 6.5 9. Closeness (STRS) .01 -.36** -.10* -.10* -.06 .41** .12** -.30** _ 45 6.2 10. Kindergarten Conflict (STRS) -.10* .10* .30** .06* .20** -.23** -.11* .29** -.06 _ 10 5.3 11. Kindergarten Closeness (STRS) .07** -.14** -.001 -.02 -.003 .12** .19** -.01 .25** -.27** _ 35 5.0 12. First Grade Conflict (STRS) -.11* .11** .36** .004 .18** -.25** -.001 .34** -.04 .36** -.001 _ 11 5.0 13. First Grade Closeness (STRS) .11** -.18** .04 -.01 .01 .23** .12** -.03 .18** -.05** .25** -.23** 34 4.8 Variables are 54-month assessments unless otherwise indicated. * p < .05, ** p <.01. Teacher-Child Relationships and Children’s Success Table 2 Descriptive Statistics for Outcome Variables (N = 490) Pearson Correlations First Grade Outcome Variables 1. 2. 3. 4. 5. 6. 7. Mean SD 1. Rated Achievement _ 3.3 .90 2. Tested Vocabulary .34** _ 108 16 3. Rated Social Competence .50** .29** _ 104 13 4. Observed Social Competence .25** .05 .28** _ 9 1.5 5. TRF Internalizing (T-score) -.24** -.10* -.40** -.10* _ 49 9.2 6.TRF Externalizing (T-score) -.21** -.16** -.54** -.07 .28** _ 50 8.7 7. CBCL Internalizing (T-score) -.03 -.01 -.15** -.05 .17** .08 _ 48 8.9 8. CBCL Externalizing (T-score) -.15* -.11* -.24** -.02 .12* .34** .59** 48 9.5 * p < .05, ** p <.01. School Psychology Review, 2004, Volume 33, No. 3 Children’s social competence–rated. in classrooms for 2–3 hours and then made At preschool age, teachers rated children’s so- qualitative ratings on 4-point scales ranging cial competence using the California Preschool from 1 = not at all characteristic to 4 = highly Social Competency Scale (Levine, Elzey, & characteristic in preschool and on 7-point Lewis, 1970). The original California Pre- scales that utilized equivalent endpoints but school Social Competency Scale contains 30 spread the scale across a wider range in first items assessing a broad range of preschool- grade. The qualitative ratings of children’s self- aged children’s social competencies. In the reliance, positive affect, and attention to ob- present study, four new items also were added jects or activities were averaged to form the to tap children’s cooperative play, rule follow- “Child Social Competence” composite in pre- ing, empathy, and aggression. All 34 items school. The qualitative ratings of child self- were rated by preschool teachers on a 4-point reliance and positive affect were averaged to scale with higher scores indicating greater lev- form this composite in first grade (as the child els of social competency. Scores ranged from attention variable was no longer coded in the 46 to 135 (M = 105, SD = 14). This scale has first grade observation). Complete descriptions been demonstrated to provide a valid assess- are available in the NICHD-SECC data col- ment of social functioning in preschool chil- lection manuals (NICHD Early Child Care dren (NICHD ECCRN, 2002b), and the 34 Research Network, 1993). Pearson correlations items have high internal reliability (Cronbach’s among these child-level qualitative ratings alpha = .88). ranged from .74 to .94. In first grade, teachers completed the Teacher-child relationship–rated. school versions of the Social Skills Question- Preschool, kindergarten, and first grade teach- naire (SSQ) from the Social Skills Rating Sys- ers rated their perceptions of their relationships tem (SSRS; Gresham & Elliott, 1990). This with the study children using the short form of instrument consists of 38 items indexing four the Student-Teacher Relationship Scale (STRS; skill areas: cooperation (e.g., keeps room neat Pianta, 2001). The short form of the STRS is a and clean without being reminded), assertion self-report instrument composed of 15 items (e.g., makes friends easily), responsibility (e.g., rated on 5-point Likert-type scales that assesses asks permission before using a family a teacher’s perception of her relationship with member’s property), and self-control (e.g., a particular student. The STRS item ratings can controls temper when arguing with other chil- be summed into two groups of items that are dren). Teachers indicated on a 3-point scale referred to as the conflict and closeness how often the child exhibited each behavior. subscales (Pianta, 2001). The conflict subscale The total score is the sum of all 38 items, with assesses the degree to which a teacher feels higher scores reflecting higher levels of per- that her relationship with a particular student ceived competence. The SSRS was normed on is characterized by negativity. The closeness a diverse, national sample of children in the scale measures the extent to which a teacher 3–5-year age range and shows high levels of feels that her relationship with the student be- internal consistency (median = .90) and test- ing assessed is characterized by warmth, af- retest reliability (.75 to .88). fection, and open communication. In terms of Children’s social competence–ob- reliability, statistically significant test-retest served. Interactions between teachers and chil- correlations over a 4-week interval and high dren were observed in the classroom at pre- internal consistency for both conflict and close- school using the Observational Record of the ness subscales have been demonstrated (Pianta, Caregiving Environment (ORCE; NICHD 2001). The predictive and concurrent validity Early Child Care Research Network, 1996) and of the STRS also has been demonstrated re- at first grade using an upward extension of the peatedly: The STRS is correlated with both ORCE, the Classroom Observation System– current and future academic skills (Hamre & First Grade (COS-1; NICHD-ECCRN, 2002a). Pianta, 2001), behavioral adjustment (Birch & Trained research assistants observed children Ladd, 1998), risk of retention (Pianta et al., 450 Teacher-Child Relationships and Children’s Success 1995), disciplinary infractions (Hamre & variate correlations were utilized to examine Pianta, 2001), and peer relations (Birth & Ladd, the stability of different teachers’ perceptions 1998). Furthermore, findings from Hamre and of both conflict and closeness in their relation- Pianta (2001) demonstrate the discriminant ships with the same student over time. Results validity of closeness and conflict scores and can be found in Table 1. Overall, teachers’ rat- suggest that use of both scales to assess posi- ings of the level of conflict they perceived in tive and negative aspects of student-teacher their relationships with a student were moder- relationships is indicated. ately correlated across time/ teacher (range r = .32 between preschool and kindergarten to r = Data Analysis .40 between preschool and first grade). Teach- ers’ ratings of the closeness they felt in rela- Bivariate correlations were conducted to tionships with students were somewhat less evaluate the stability of teachers’ relationship stable between preschool and first grade (range ratings across the preschool to first grade time r = .21 from preschool to first grade to r = .31 frame. Additionally, a descriptive examination between kindergarten and first grade), but oth- of mean levels of conflict and closeness re- erwise estimates of the stability of closeness ported over time was conducted using two re- were moderate and similar to estimates of the peated measures analysis (one for conflict and stability of conflict. Furthermore, the overall one for closeness). Regression analyses evalu- levels of teacher-reported conflict and close- ating the extent to which teacher-child relation- ness over time were evaluated using repeated ships predicted first grade skills above and measures analysis. Results indicate that over- beyond preschool skills and demographic char- acteristics also were conducted. These analy- all mean levels of both conflict and closeness ses began with a model predicting child changed very slightly but were statistically sig- competencies or problems in first grade nificantly over time (for conflict Greenhouse- from demographic factors and earlier assess- Geisser = .990. p <.01 and for closeness Green- ments of analogous competencies or prob- house-Geisser = .998. p <.01). The general lems, and then determined how much addi- trend indicated decreasing levels of both con- tional variance in outcomes could be ex- flict and closeness reported by teachers over plained by teachers’ perceptions of their re- time (see Table 1 for means). lationships with students. As teacher-child Hierarchical regression analyses were relationships were assessed repeatedly from used to address the second question of the value preschool through Grade 1, examination of added by teacher-child relationships to predic- the contributions of current and previous re- tions of children’s skill levels in first grade lationships with teachers was possible by add- above and beyond assessments of those skills ing these indicators in temporal sequence. Fi- in preschool. In each regression, predictor vari- nally, interactions between time-averaged STRS ables were entered in a series of six steps. In conflict and closeness scores and earlier com- the first step, demographic covariates were petencies or problems, gender, and poverty sta- entered as a block (gender and poverty status). tus in predicting developmental outcomes in In the second step, the 54-month version of first grade were examined. These last analy- the outcome being predicted was entered. The ses examined whether, in general, children’s STRS conflict and closeness scores at 54 relational competencies with teachers moder- months, kindergarten, and first grade were en- ated the association between demographic risk tered as blocks in the third, fourth, and fifth factors and success in first grade. steps, respectively. Finally, interactions be- tween the three covariates and an average of Results the 54-month through first grade conflict and Two broad questions were addressed in closeness scores were entered in a sixth block. this study. The first question concerned the sta- However, because only 1 interaction term of bility of teacher-child relationships across the the 48 interactions that were entered into analy- preschool to first grade interval. Pearson bi- ses (6 in each of the 8 regressions) was statis- 451 School Psychology Review, 2004, Volume 33, No. 3 tically significant, it was ascertained that over- Behavior Problems all, interactions did not add meaningful infor- Both mothers’ and teachers’ ratings of mation to prediction analyses. Thus, the inter- internalizing and externalizing problems were action blocks are not reported. Results of hier- investigated to provide a more complete picture archical regression analyses can be found in of how teacher-child relationships were associ- Table 3 and are summarized below. ated with children’s behavioral development in- Overall, the demographic covariate side and outside of the school setting. Mother- block, which included poverty status and gen- rated child internalizing behavior at first grade der, accounted for significant variance in the was predicted by preschool and first grade following child outcomes: teachers’ ratings of teachers reports of closeness. Specifically, less children’s achievement, externalizing behav- closeness reported by the preschool or first ior, and social competence, and tested vocabu- grade teacher was linked to higher levels of lary development. Similarly, the second mothers’ reports of internalizing behavior. covariate block, which included the 54-month In terms of externalizing behavior rated assessment of child outcomes, accounted for by mothers, more teacher-rated relational con- significant variance in child outcomes at the flict at first grade was significantly related to time that it was entered in all analyses except mother’s reports of children’s externalizing when observed social competence was the out- behavior. Previous teachers’ ratings of close- come under consideration. When all variables ness and conflict were not associated with were included, the 54-month assessments of mothers’ ratings of child externalizing in first child outcomes were significant predictors for grade. teacher-rated achievement, tested vocabulary With regard to first grade teachers’ rat- development, mother-rated internalizing be- ings of child problem behavior, their concur- havior, teacher- and mother-rated externaliz- rent perceptions of closeness and conflict in ing behavior, and teacher-rated social compe- their relationships with students predicted their tence in the expected directions (i.e. higher perception of the child’s internalizing prob- levels of each construct at 54 months was as- lems: More conflict and less closeness were sociated with higher levels of that construct in associated with higher levels of perceived child first grade). Results for variance attributable internalizing problems. No associations with to relational conflict and closeness at 54 prior teacher-child conflict or closeness were months, kindergarten, and first grade, adjusted found for teacher-rated internalizing at first for the two covariate blocks, are summarized grade. Regarding externalizing problems, below for each outcome examined. teacher-reported relational conflict with chil- Academic Performance dren in both kindergarten and first grade were associated with first grade teachers’ ratings of When predicting teacher ratings of externalizing in the expected directions. achievement, both conflict and closeness in first grade were significant predictors. First Social Competence grade teachers rated children’s achievement more highly for those children with whom they Both teachers’ and trained observers’ reported sharing a closer relationship, and as- ratings of children’s social competence were signed lower achievement ratings to those chil- investigated. Kindergarten and first grade dren with whom they felt more conflict. No teacher’s reports of conflict in their relation- other indicators of the teacher-child relation- ships with children were inversely related to ship were statistically significant in predicting the first grade teachers’ reports of children’s rated achievement. Regarding tested vocabu- social competence: Higher conflict was asso- lary development in first grade, preschool ciated with lower social competence. In addi- through first grade teachers’ perceptions of tion, as first grade teachers reported more their relationships with students did not dem- closeness in their relationship with a given onstrate significant prediction. child, they also reported higher levels of so- 452 Teacher-Child Relationships and Children’s Success Table 3 Prediction of Social and Academic Outcomes in First Grade (N = 490) First Grade Outcomes Rated Tested TRF TRF CBCL CBCL Rated Observed Achieve- Vocab- Internal- External- Internal- External- Social Social ment ulary izing izing izing izing Comp. Comp. Model R .17 .48 .11 .61 .34 .52 .47 .57 Model F 10.8** 49.5** 6.2** 83.5** 28.8** 57.1** 48.8** 3.3** Step 1: Demographics ( R ) .04** .13** .01 .03** .001 .01 .04** .01 Gender (high = female) -.10* -.15** .06 .07* .003 .03 -.09** -.05 Poverty (high = poor) -.08 -.07 .03 .13** .06 .001 -.21** -.02 Step 2: 54-m indicator of Grade 1 .05** .34** .11* .21** .32** .48** .13** .001 outcome ( R ) .19** .61** .06 .21** .55** .67** .19** .02 Step 3: 54m STRS ( R ) .02** .001 .002 .01** .01** .004 .01* .01 Conflict -.05 .05 -.01 -.01 .01 .08 -.02 -.03 Closeness .04 -.02 .01 .05 -.10* -.02 -.004 .06 Step 4: Kindergarten STRS ( R ) .01 .01 .005 .069** .01 .006 .05** .02** Conflict -.03 .03 .03 .12** .06 .02 -.11** -.03 Closeness .03 .06 .04 .02 .05 .07 .02 .09 Step 5: First Grade STRS ( R ) .06** .001 .08** .29** .01* .02** .25** .020** Conflict -.10* -.03 .19** .65** -.05 .14** -.38** -.09 Closeness .22** .02 -.19** .03 -.07 -.04 .32** .10* Cell values are standardized β coefficients derived from the final model. *p < .05, **p < .01. School Psychology Review, 2004, Volume 33, No. 3 cial competence in that child. Furthermore, relationships and both mothers’ and neutral when entered as a block, preschool teachers’ observers’ ratings of children. reports of their relationship with children ac- Regarding the descriptive examination counted for significant variance in first grade of stability in different teachers’ perceptions teachers’ ratings of children’s social compe- of conflict and closeness in their relationships tence, but when subsequent predictors were with a given child, findings indicated moder- included in the final model, neither of the kin- ate stability in conflict and slightly less in close- dergarten relationship indicators demonstrated ness across the 54-month to first grade time a significant association with ratings of social period. This finding is consistent with previ- competence in first grade. Finally, variance in ous studies of the stability of teacher-child re- trained observers’ ratings of children’s social lationships from kindergarten to first grade competence in the first grade classroom was (Pianta et al., 1995) and from preschool to kin- predicted by first grade teachers’ perceptions dergarten (Howes et al., 2000). Previous and of closeness in their relationship with a child present findings suggest that relational conflict in the expected directions. Although at entry with a given child is more stable across teach- kindergarten teacher-child closeness was pre- ers (perhaps related to stable child character- dictive of first grade observed social compe- istics such as temperament), whereas relational tence (∆ R for the block = .02, p = .03), when closeness may depend in greater part on the all predictors were entered in the full model, goodness-of-fit between the interpersonal this link was no longer significant. styles of teachers and children. The second set of descriptive findings Discussion pertained to consistency in the mean levels of teacher-reported conflict and closeness from The overall pattern of results in the preschool through first grade. Results suggest present study provides preliminary guidance a slight trend towards teachers reporting less for future investigations of how teacher-child intense relationships with children in both relationship quality is related to children’s so- negative and positive dimensions with increas- cial and academic development in the early ing child age. This finding may reflect a de- school years. Preliminary descriptive findings creasing emphasis on teacher-child relation- indicate that different teachers’ perceptions of ships in the classroom from preschool to early their relationships with a given child are moder- elementary school. Previous studies have sug- ately stable from preschool through first grade. gested a similar dynamic across this time pe- In addition, there was a small but statistically riod (Johnson, Gallagher, Cook, & Wong, significant trend towards teachers reporting decreasing levels of both conflict and close- 1995; Rusher, McGrevin, & Lambiotte, 1992; ness from preschool through first grade. Spodek, 1988; Vartuli, 1999). With regard to associations between teacher- With regard to associations between child relationships and child outcomes, af- teacher-child relationships and first grade com- ter adjusting for gender, socioeconomic sta- petencies, estimates of prior social and aca- tus, and children’s prior skill levels, several demic skills were consistently significant cor- indicators of teacher-child relationships relates of first grade assessments of those out- made small to moderate contributions to pre- comes, with the exception of observed social diction of social and teacher-rated academic competence. In several cases (tested vocabu- skills in first grade. Although the associations lary development, externalizing behavior, and between teacher-child relationships and rated social competence), demographic vari- children’s outcomes tended to be greatest when ables also accounted for significant variance the outcomes and relationships were rated by in first grade skills. However, above and be- the same person (i.e., the teacher), not all of yond the effects of these factors, teachers’ per- the findings can be attributed solely to rater ceptions of the student-teacher relationship on effects. In fact, small but significant associa- at least one occasion between preschool and tions were found between teachers’ ratings of first grade had small but significant relations 454 Teacher-Child Relationships and Children’s Success to all child outcomes except tested vocabulary found that preschool or childcare teacher-child development. Although many assessments of relationship quality is more closely associated teacher-child relationships were not signifi- with children’s development over time than cantly associated with child outcomes, these quality of relationships with elementary school preliminary findings add to our understanding teachers (Burchinal et al., 2002). The present of how teacher-child relationships are associ- findings indicate that although preschool and ated with changes in individual differences in kindergarten teacher-child relationships dem- children’s social and academic competence onstrate small associations with some first over time. Previously, work by Birch and Ladd grade skills (rated achievement, teacher-re- (1998) demonstrated that kindergarten teach- ported externalizing behavior, mother-reported ers’ perceptions of conflict in their relationship internalizing behavior, and rated social com- with students accounted for significant vari- petence), both the number of significant effects ance in children’s prosocial behavior in first and the magnitude of those effects is greater grade. The current study extended that finding for concurrent teachers’ ratings of relational by expanding the time interval in which quality. First grade teachers’ ratings of rela- children’s behavioral development was stud- tional quality were significantly related to all ied and by using repeated assessments of the first grade outcomes except tested vocabulary teacher-child relationship. development, and in all cases accounted for a An additional pattern in the findings that greater proportion of variance in those out- merits discussion is the association between comes than earlier teachers’ reports of relation- teacher-child relationships and mother- and ships, even though those first grade relation- teacher-rated behavior problems. Regardless ship indicators were entered into the analytic of rater and controlling for prior problem be- model last. haviors, demographic variables, and prior Several important limitations to this teacher-child conflict and closeness, contem- work also merit review. First, this work clearly poraneous conflict (or lack of closeness) be- focuses on the contributions of the teacher- tween teachers and children is linked to child relationships to predicting child out- children’s externalizing (or internalizing) prob- comes. In adopting such a focus, consider- lems in first grade. The fact that changes in ation of other aspects of the teachers’ pres- mothers’ reports of child behavior problems is ence in the classroom and/or other aspects associated with quality of the teacher-child of her or his interactions with children is relationship suggests some degree of crossover omitted. As a result, it cannot be stated con- between home and school contexts. This is clusively that teacher’s perceptions of rela- consistent with previous findings linking ob- tionship quality is uniquely predictive, as served quality of classroom processes to this could be highly collinear with other as- changes in mothers’ perceptions of their pects of the teachers’ behavior in the class- children’s internalizing behavior from pre- room (including but not limited to her or his school through first grade (NICHD ECCRN, teaching style, classroom management style, 2003). or the overall emotional tone of the class- As this study incorporated three assess- room). To conclude that it is this aspect of ments of the teacher-child relationship over the teacher’s presence and not others that time, issues pertaining to the timing of this re- contributes to changes in individual differ- lationship were available for consideration. ences in children’s competencies, future Earlier work has suggested that teacher-child work may include other classroom process relationships may have the greatest effect on variables in addition to teachers’ perceptions children’s competency in that specific class- of their relationships with children so that the room, particularly among high-risk children overlap between these constructs can be fur- (i.e., that teacher-child relationships have pri- ther explored. Furthermore, the possibility ex- marily concurrent effects; see Pianta, Nimetz, ists that the teacher’s perception of the teacher- & Bennett, 1997). Conversely, others have child relationship is more of a marker of the 455 School Psychology Review, 2004, Volume 33, No. 3 child’s overall competence than an indepen- advantages of adding an assessment of the dent construct. A fruitful area for future re- teacher-child relationship to the standard test search would experimentally test whether im- batteries used by school psychologists. Add- provements in the teacher-child relationship ing this dyadic component to evaluations of result in changes in children’s patterns of be- children with school problems could poten- havior and success in school. tially contribute valuable information not con- Despite these limitations, this work dem- tained in traditional child-centered behavior onstrates that the interpersonal aspects of checklists. These results also have implications children’s experience make a difference in their for early intervention efforts and school policy ability to develop competencies in the early in that they suggest a focus on relational pro- years of school. Further work may include not cesses between teachers and students as a only the teacher-child relationship, but also mechanism for enhancing opportunities for other aspects of the social environment that success in school. children are exposed to as well to get a richer Footnotes picture of the interpersonal context of early elementary school and how that context 1 Of note, the 1 interaction term (out of 36) functions as a medium in which children that accounted for statistically significant variance develop. Examples of such initiatives in- in a child outcome was in predicting teacher-rated academic skills in first grade. In that case, the 54- clude Ladd and colleagues’ (1999) work, month closeness x 54-month vocabulary skills in- which utilized assessments of peer accep- teraction term explained significant variance in tance, children’s best friendships, and teachers ratings of academic skills in first grade. teacher-child relationship quality within the Exploration of this interaction revealed that there kindergarten year and found that social dy- was a stronger association between closeness and namics with both peers and teachers were rated achievement in first grade among children important aspects of a model predicting whose academic achievement was above the mean children’s academic achievement. Clearly, at 54 months. additional work is needed in this area to de- References velop a clear picture of the effect of rela- tional processes in the classroom. Continued Achenbach, T. (1991). Manual for the Child Behavior work in this direction will provide teachers, Checklist/4-18 and 1991 Profile. Burlington, VT: Au- school psychologists, and other school person- thor. Birch, S. H., & Ladd, G. W. (1997). The teacher-child re- nel with a better understanding of how to fa- lationship and children’s early school adjustment. Jour- cilitate positive outcomes for all children in nal of School Psychology, 35, 61-79. school. Birch, S. H., & Ladd, G. W. (1998). Children’s interper- In sum, although the associations be- sonal behaviors and the teacher-child relationship. Developmental Psychology, 34, 934-946. tween teacher-child relationships and Burchinal, M. R., Peisner-Feinberg, E., Pianta, R. C., & children’s development tended to be small and Howes, C. (2002). Development of academic skills were not significant in all cases, these re- from preschool through second grade: Family and sults do suggest a salient relational compo- classroom predictors of developmental trajectories. Journal of School Psychology, 40, 415-436. nent to children’s early school competence. Entwisle, D. R., & Alexander, K. (1999). Early schooling Despite children’s general tendencies to- and social stratification. In R. C. Pianta & M. J. Cox wards certain relational styles with teach- (Eds.), The transition to kindergarten (pp. 13-38. Bal- ers inferred from the moderate stability timore, MD: Paul H. Brookes. Greenberg, M. T., Speltz, M. L., & Deklyen, M. (1993). across different teachers’ reports of their The role of attachment in the early development of relationships with a given child across the disruptive behavior problems. Development and Psy- preschool to first grade time period, each chopathology, 5, 191-213. new relationship may represent an opportu- Gresham, F., & Elliott, S. (1990). Social Skills Rating System. Circle Pines, MN: American Guidance Ser- nity to introduce unique variance in the devel- vice. opment of social and academic skills. Research Hamre, B. K., & Pianta, R. C. (2001). Early teacher-child findings such as these suggest the possible relationships and the trajectory of children’s school 456 Teacher-Child Relationships and Children’s Success outcomes through eighth grade. Child Development, NICHD Early Child Care Research Network. (2002a). The 72, 625-638. relation of global first grade classroom environment Howes, C., Hamilton, C. E., & Matheson, C. C. (1994). to structural classroom features, teacher, and student Children’s relationships with peers: Differential asso- behaviors. Elementary School Journal, 102, 367-387. ciations with aspects of the teacher-child relationship. NICHD Early Child Care Research Network. (2002b). Child Development, 65, 253-263. Child care structure—process—outcome: Direct and Howes, C., & Matheson, C. C. (1992). Contextural con- indirect effects of child care quality on young children’s straints on the concordance of mother-child and development. Psychological Science, 13, 199-206. teacher-child relationships. In R. C. Pianta (Ed.), Be- NICHD Early Child Care Research Network. (2003). So- yond the parent: The role of other adults in children’s cial functioning in first grade: Associations with ear- lives (Vol. 57, pp. 25-40). San Francisco: Jossey-Bass. lier home and child care predictors and with current Howes, C., Phillipsen, L. C., & Peisner-Feinberg, E. classroom experiences. Child Development, 74, 1639- (2000). The consistency of perceived teacher-child re- 1662. lationships between preschool and kindergarten. Jour- Pianta, R. C. (1999). Enhancing relationships between nal of School Psychology, 38, 113-132. children and teachers. Washington, DC: American Johnson, L. J., Gallagher, R. J., Cook, M., & Wong, P. Psychological Association. (1995). Critical skills for kindergarten: Perceptions Pianta, R.C. (2001). Student-Teacher Relationship Scale: from kindergarten teachers. Journal of Early Interven- Professional manual. Odessa, FL: Psychological As- tion, 19, 315-327. sessment Resources. Kamphaus, R. W. & Frick, P .J. (2002). Clinical assess- Pianta, R. C., & Nimetz, S. (1991). Relationships between ment of children and adolescent personality and be- teachers and children: Associations with behavior at havior. Boston: Allyn & Bacon. home and in the classroom. Journal of Applied Devel- Ladd, G. W., Birch, S. H., Buhs, E. S. (1999). Children’s opmental Psychology, 12, 379-393. social and scholastic lives in kindergarten: Related Pianta, R. C., Nimetz, S. L., & Bennett, E. (1997). Mother- spheres of influence? Child Development, 70, 1373- child relationships, teacher-child relationships, and 1400. school outcomes in preschool and kindergarten. Early Ladd, G. W., & Burgess, K. B. ( 1999). Charting the rela- Childhood Research Quarterly, 12, 263-280. tionship trajectories of aggressive, withdrawn, and Pianta, R. C., Steinberg, M. S., & Rollins, K. (1995). The aggressive/withdrawn children during early grade first two years of school: Teacher-child relationships school. Child Development, 70, 910-929. and deflections in children’s classroom adjustment. Ladd, G. W., & Burgess, K. B. ( 2001). Do relational risk Development & Psychopathology, 7, 295-312. and protective factors moderate the links between Ramey, S. L., Ramey, C. T., & Phillips, M. M. (1996). childhood aggression and early psychological and Head Start children’s entry into public school: An in- school adjustment? Child Development, 72, 1579-1601. terim report on the National Head Start-Public School Levine, S., Elzey, F., & Lewis, M. (1970). California Pre- Early Childhood Transition Demonstration Study. school Social Competency Scale. Palo Alto, CA: Con- Washington, DC: U.S. Department of Health and Hu- sulting Psychologists Press. man Services. Lynch, M., & Cicchetti, D. (1992). Maltreated children’s Rusher, A. S., McGrevin, C. Z., & Lambiotte, J. G. (1992). reports of relatedness to their teachers. In R. C. Pianta Belief systems of early childhood teachers and their (Ed.), Beyond the parent: The role of other adults in principals regarding early childhood education. Early children’s lives (Vol. 57, pp. 81-108). San Francisco: Childhood Research Quarterly, 7, 227-296. Jossey-Bass. Spodek, B. (1988). The implicit theories of early child- Lyon, G. R. (2002). Reading development, reading diffi- hood teachers. Early Childhood Development and culties, and reading instruction: Educational and public Care, 29, 197-208. health issues. Journal of School Psychology, 40, 1-4 Vartuli, S. (1999). How early childhood teacher beliefs National Center for Education Statistics. (1999). America’s vary across grade level. Early Childhood Research kindergartners. Washington, DC: National Center for Quarterly, 14, 489-514. Education Statistics. Wilson, M. S. & Reschly, D. J. (1996). Assessment in National Education Goals Panel. (1997).Building a na- school psychology training and practice. School Psy- tion of learners. Washington, DC: Author. chology Review, 25, 9-23 National Institute for Child Health and Human Develop- Woodcock, R. W., & Johnson, M. B. (1990). Woodcock- ment (NICHD) Early Child Care Research Network. Johnson Psycho-Educational Battery—Revised. Allen, (1993). The NICHD study of early child care: A com- TX: DLM Teaching Resources. prehensive longitudinal study of young children’s lives. Woodcock, R. W., & Mather, N. (1989). Woodcock ERIC Clearinghouse on Elementary and Early Child Johnson—Revised Tests of Cognitive Ability standard Education, ED 353-087. battery: Examiner’s manual. In R. W. Woodcock & NICHD Early Child Care Research Network. (1996). Char- M. B. Johnson, Woodcock-Johnson Psycho-Educa- acteristics of infant child care: Factors contributing to tional Battery—Revised. Allen, TX: DLM Teaching positive caregiving. Early Childhood Research Quar- Resources. terly, 11, 269-306. 457 School Psychology Review, 2004, Volume 33, No. 3 Robert C. Pianta, PhD, is a professor in The Curry School of Education at the University of Virginia and holds the Novartis U.S. Foundation Chair in Education. He is particularly interested in how relationships with teachers and parents, as well as classroom quality, can help improve outcomes for at-risk children and youth. Dr. Pianta is a principal investiga- tor on the National Institute of Child Health and Human Development Study of Early Child Care and Youth Development, a senior investigator with the National Center for Early Development and Learning, and editor of the Journal of School Psychology. Megan W. Stuhlman, PhD, graduated from The Curry School of Education’s Programs in Clinical and School Psychology at the University of Virginia in 2004 and is currently a research fellow at the University of Miami-School of Medicine. Her research interests include examining how developmental trajectories of children and youth at-risk for diffi- culties in home and school settings are influenced by relationships between children and parental and nonparental adults.

Journal

School Psychology Review – Taylor & Francis

Published: Sep 1, 2004

There are no references for this article.