The importance of peer-student and teacher-student relationships: Evidence from the Longitudinal Study of Australian Children

Abstract

There is ample research demonstrating the importance of peer-student relationships for children’s development. For teacher-student relationships as well, evidence for its influence on children’s development is accumulating. However, little is known about the relative influence of teacher-student relationships on children’s outcomes in comparison to peer-student relationships. To date, studies including both teacher-student relationships and peer-student relationships are scarce (e.g., Leflot et al., 2011; Mercer & DeRosier, 2008). There are even less, if any, studies that have examined the relative impact of teacher-student relationships and peer-student relationships on different domains of development. Considering the instructive roles of teachers, it is possible that teacher-student relationships have a larger influence on children’s cognitive development than on behavioral development in comparison to peer-student relationships. To obtain a more complete understanding of the impact of teacher-student relationships relative to peer-student relationships, the current study examined developmental associations between teacher-student relationships, peer-student relationships, conduct problems, and language competencies in a nationally-representative Australian sample. Methods. The sample included the Kindergarten cohort (N=4,983; age in months M=56.9, SD=2.6) of the ‘Longitudinal Study of Australian Children’. Teachers completed the Closeness (CL) subscale (α=.83-.85) of the Student-Teacher Relationship Scale (Pianta, 2001). Averaged scores of parent- and teacher-reports on the Peer problems scale (α=.66-.73) and the Conduct problems scale (α=.74-.75) of the Strength and Difficulties Questionnaire (Goodman, 1997) were used as indicators of Peer problems (PP) and Conduct problems (CP), respectively. The Peabody Picture Vocabulary Test-Revised (Dunn & Dunn, 1997) measure of receptive vocabulary served as an indicator of language achievement (ACH). Measures were collected biennially at age 4/5 (Wave 1), age 6/7 (Wave 2), and age 8/9 (Wave 3). Socioeconomic status, child sex, main language spoken at home, and indigenous status were included as covariates. Autoregressive cross-lagged analyses were performed using the MLR estimator to account for non-normality. Because of the large sample size, alpha was set at .01 (two tailed) to reduce the likelihood of a Type-1 error. Parameter estimates were considered significant at the 99% confidence interval. Results. A series of nested models was tested: an autoregressive model (autoregressive paths and concurrent associations), a disorder-driven model (cross-time paths from CP and ACH to CL and PP added), a child x environment model (cross-time paths from CL and PP to CP and ACH added), and a full transactional model (cross-time paths between CL and PP added). Based on Satorra-Bentler scaled χ2difference tests for nested models (Satorra & Bentler, 2001), the child x environment model was selected as the best model (Figure 1). The fit of this model was satisfactory (χ2(20)=459.640; RMSEA=.07; SRMR=.02; CFI=.96). Conclusions. Different effects were found for peer-student and teacher-student relationships. Peer-student relationships (but not teacher-student relationships) had a significant effect on children’s behavioral development and vice versa. Importantly, teacher-student relationships (but not peer-student relationships) had a consistent and positive impact on language development. When teachers and children share a close relationship, children may be more engaged in personal conversations with the teacher, thereby advancing the child’s understanding of more sophisticated language.