Teacher-Student Relationship and Students’ Social-Emotional Outcomes
Abstract
Teacher-student relationships (TSRs) are expected to be a promotive factor for students’ social-emotional outcomes. This multilevel meta-analysis focuses on the association between teacher-student relationships (TSRs) and students’ social-emotional outcomes. Possible moderating factors were examined as well. In total, 17 studies, reporting on 98 effect sizes, were included. The results showed a significant positive association between TSRs and students’ social-emotional outcomes (r = .302) indicating that higher quality of TSRs were associated with more positive social-emotional outcomes. Measurement characteristics, dimension of social-emotional outcome (smaller effect size for sociability than for social skills), student’s age (smaller effect sizes with increasing age), school type (smaller effect size for secondary than for elementary school) and school population (larger effect sizes for school samples of children showing typical and atypical development) moderated the relation between TSRs and students’ social-emotional outcomes. Implications for theory and practice concerning the role of TSRs in students’ social-emotional development are discussed.
Introduction
The school is a social environment in which students can develop a personal identity, which allows them to take their place in the community (Boocock, 1973; Wentzel, 1993). The school focuses on students’ achievements, but can also be seen as a social context in which students learn to develop social-emotional knowledge and skills (Boocock, 1973; Durlack, Weissberg, Dymnicki, Taylor, & Schellinger, 2011). The social relationships students establish within the school context are thought to be important for their social-emotional development and their school learning and achievement (Aviles, Anderson, & Davila, 2006; Pianta, Nimetz, & Bennett, 1997). These social relationships consist of two types: student-peer relationships and teacher-student relationships (TSRs). In this meta-analysis, we focus on the second type of relationships and their association to social-emotional outcomes in students.
There has been much research on the association between TSRs and students’ social-emotional and school-based outcomes (Davis, 2001; Pianta, 1997). Roorda and colleagues (2011) integrated the theories and knowledge on the association between TSRs and school-outcomes by conducting a meta-analysis. They found empirical evidence for the impact of both positive and negative aspects of TSRs on the students’ school engagement and achievement. Nurmi (2012) conducted a first (selective and standard) meta-analysis accounting for only between study differences in effect sizes, and showing that teachers reported positive TSRs for students exhibiting high levels of motivation and engagement. Although the number of studies examining the relation between the TSR and students’ social-emotional outcomes has been steadily growing, this knowledge has not yet been sufficiently integrated (Decker, Dona, & Christenson, 2007; Luckner & Pianta, 2011; Poulou, 2015). Therefore, it is necessary to conduct a systematic review and multilevel meta-analysis in
which we focus on the association between the TSR and students’ social-emotional outcomes, accounting for both within and between study differences in effect sizes.
Teacher-student Relationships
Over the past two decades, there has been an increasing interest in the impact of the quality of relationships between students and teachers on student outcomes (Davis, 2003). TSRs can be described as a product of the reciprocal relation between individual teacher and student characteristics (Pianta, Hamre, & Stuhlman, 2003). Roorda et al. (2011) examined the relation between TSRs and students’ academic outcomes from two theoretical perspectives, that is, attachment theory and self-determination theory (Ryan & Deci, 2002).
From extended attachment theory, TSRs may be considered as a form of child-caregiver attachment relationships (Cornelius-White, 2007; Spilt, Koomen, & Thijs, 2011). The main idea of attachment theory is that the relationship between a child and attachment figure (caregiver) creates emotional security in the child if the attachment figure provides for a secure base and safe haven (Bowlby, 1988). Emotional security is in turn considered to be a necessary precondition for exploration of the environment and ability to effectively appeal to the caregiver for comfort in stressful situations. According to extended attachment theory, teachers’ nurturing and responsiveness to students’ needs may serve as a foundation from which students can socially and academically learn and develop (Davis, 2003). High quality TSRs are thought to support students’ motivation to explore their social and academic environment as well as foster their development of social, emotional and cognitive skills. Low quality TSRs on the other hand, reflect lack of security, and are believed to interfere with students’ attempts to cope with the demands of their environment. Through teachers’ ability to help students to accurately label, control and express the emotions they experience, teachers may become increasingly important in the process of emotion regulation (Pianta, 1999).
Another perspective to conceptualize TSRs is Hirschi’s (1969) attachment-based elaboration of social control theory (Hoeve et al., 2012; Murray & Greenberg, 2000), which assumes that the feelings of attachment, commitment, involvement and belief with and in the school are important for students. These feelings allow students to develop their prosocial skills and behaviors, and increase their involvement in prosocial groups. They are less likely to engage in deviant behaviors because these behaviors can threaten their relationship with significant others within the school context.
Because TSRs may be seen as some sort of child-caregiver attachment, the affective quality of TSRs are often assessed by using three concepts derived from attachment theory (Roorda et al., 2011), namely, closeness, conflict and dependency. Closeness refers to the degree of warmth and proximity to the teacher, conflict refers to non-attuned and coercive relationships, and dependency refers to overly dependent and sticky behaviors of the student (Pianta, 2001). TSRs can be conceptualized in terms of a positive and negative relationship as well, in which closeness can be seen as typical of a positive relationship, whereas conflict and dependency refer to a more negative relationship (Roorda et al., 2011).
According to self-determination theory (Ryan & Deci, 2017), three basic psychological needs must be fulfilled in order for students to become motivated (see Connell & Wellborn, 1991): the need for relatedness, competence and autonomy. Teachers can fulfill these needs by showing involvement, providing structure and supporting autonomy. Showing involvement refers to caring for and expressing interest in the student, providing structure refers to setting clear rules and being consequent, and supporting autonomy refers to giving students freedom to make their own decisions. Teachers’ involvement is connected to the affective dimension of the TSR derived from attachment theory, while relatedness is connected to the concept of emotional security (Connell & Wellborn, 1991). According to this theory teacher support, leadership and student freedom are important aspects of TSRs.
The TSR as a Promotive Factor for Social-emotional Outcomes
Social-emotional development refers to students’ growing ability to experience, regulate and express emotions, their ability to form interpersonal relationships and to explore their environment (CDE, 2016). Important aspects of social-emotional development are: interactions with adults, relationships with adults, interactions with peers, relationships with peers, identity of self in relation to others, recognition of ability, expression of emotion, empathy, emotion regulation, impulse control and social understanding. The TSR seems to be an important promotive factor for students’ social and emotional skills (Baker, 2006; Skalická, Stenseng, & Wichstrøm, 2015). A supportive and warm relationship with the teacher is thought to have a positive impact on students’ social-emotional development, whereas a conflictual teacher-student relationship is assumed to have a negative impact (Murray, 2002). Research by Murray and Greenberg (2000) showed that students who had a positive relationship with their teacher had higher scores on self- and teacher-ratings of social and emotional outcomes than students who had a negative relationship with their teacher. Research by Oberle, Schonert-Reichl, Guhn, Zumbo and Hertzman (2014) substantiates this by showing the importance of high quality school-based relationships on social and emotional wellbeing of children in a social context.
According to attachment theory, the TSR as a promotive factor for students’ social-emotional outcomes can be explained by students’ feeling of comfort and safety within the school-context. Having supportive relationships with teachers can influence students’ level of comfort and their confidence to explore new situations within the school setting (Hawkins & Catelano, 1992). From a social control perspective, the association between high quality TSRs and students’ positive social-emotional outcomes can be explained by feelings of belonging and connectedness with the school-context (Hirschi, 1969). From this perspective, TSRs can inhibit inappropriate social behaviors if such behaviors endanger continued
teacher-support and their school membership (Hawkins & Catelano, 1992). In contrast, students may show more appropriate behavior to maintain support of their teacher and their school membership. Furthermore, teachers can function as a model in a way that students may be more willing to adopt teachers’ prosocial behaviors when they experience a high quality TSR (Pianta, Hamre, & Stuhlman, 2003).
Impact of TSRs: The Role of Student, Teacher and Study Characteristics
The strength of the association between TSRs and students’ social-emotional
outcomes may be influenced by several factors, such as characteristics related to TSRs, different aspects of social-emotional functioning, and sample and study characteristics. Several researchers found evidence for the influence of different aspects of TSRs on social-emotional outcomes (Koomen & Jellesma, 2015; Leflot, Onghena, & Colpin, 2010; Luckner & Pianta, 2011), as well as the type of assessment of TSRs. For example, higher rates of conflict are associated with less positive student outcomes, whereas scores of closeness may be more strongly associated with positive outcomes (Birch & Ladd, 1997; Hamre & Pianta, 2001). In addition, the association between TSRs and students’ social-emotional functioning may be different per aspect of social-emotional development. For example, Luckner and Pianta (2011) found in their study that overall TSR quality had greater effects on students’ cooperative peer behavior than social withdrawal. According to them, classroom interactions may enable students to further develop their social skills, which are then carried over into interactions with their peers. These interactions between teachers and students in the classroom provide opportunities for more positive peer interactions, and decreased opportunities for negative peer interactions. The type of assessment of social-emotional outcomes may moderate the association as well. Parents may report different social-emotional behavior from their child than the teacher, as they experience the child in different environments (Skalicka, Stenseng, & Wichstrøm, 2015).
Different student characteristics may moderate the association between TSRs and students’ social-emotional outcomes. Students’ gender as well as their age and ethnicity may moderate this association. Girls seem to be more likely to have a positive relationship with their teacher, defined by more closeness and less conflict, than boys (Koomen & Jellesma, 2015). Overall, boys may display more conflict producing behaviors than girls, which negatively affects the relationship with their teacher (Ladd et al., 1999). Therefore, boys may be more likely to have a conflictual relationship with their teacher, affecting also their social-emotional behavior. Jerome, Hamre and Pianta (2009) found similar effects of students’ gender, as well as for students’ ethnicity and age. They found black children to experience greater conflictual relationships with their teachers. Children with an ethnic minority background are more likely to come from families of lower SES and maternal education (Pianta, La Paro, Cox, & Bradley, 2002). They are therefore more likely to be placed in classrooms that are more teacher-directed and less positive than their peers from higher SES backgrounds. In addition, they found a change in teacher reports on closeness and conflict dimensions of TSRs over a period of seven years: increases in teacher-reported conflict and decreases of closeness (Jerome, Hamre, & Pianta, 2009). These results indicate a change in TSR quality as students become older and enter a different school type. As children grow older, several developmental changes occur, such as increasing autonomy, decreasing dependency from adults and focus on their peers. As children become less dependent on adults when they age, the nature of the relationship with their teacher changes as well (Ang, Chong, Huan, Quek, & Yeo, 2008).
Teacher characteristics, such as gender and years of experience may affect TSRs as well and by that, students’ social-emotional outcomes. Female teachers may have better relationships, reported as more close, less conflictual and less dependent, than male teachers (Spilt, Koomen, & Jak, 2012). A possible explanation for this finding may be that females are
more socialized to develop nurturing relationships with others and are more accepting towards students’ negative behavior and comfort-seeking behavior. This may result in better TSRs. In addition, teachers with fewer years of experience may be warmer and more responsive when interacting with their students than teacher with more years of experience (McDonald-Connor, Son, Hindman, & Morrison, 2005). Therefore, TSR quality between teachers with fewer years of experience and their students may be higher, resulting in more positive social-emotional functioning of students.
Considering school characteristics, we mentioned school type to possibly moderate the association between TSRs and social-emotional outcomes. The quality of TSRs was found to fluctuate over kindergarten and elementary school years, which may affect students’ social-emotional outcomes (Jerome, Hamre, & Pianta, 2009). This is in line with the influence of students’ age and can be explained by the change in quality and quantity in TSRs as students age. Their relationship with their teacher becomes less frequent and less intensive (Lynch & Cicchetti, 1997). In addition, the school population may influence this association as well. Research on differences in TSRs of students with and without disabilities showed that students with disabilities tend to have lower quality TSRs (Blacher, Baker, & Eisenhower, 2009; Murray & Greenberg, 2001). These students experienced less closeness and more conflict and dependency than their typically developing peers over a period of two years.
Present Study
The present study is a multilevel meta-analysis on the association between TSRs and social-emotional outcomes of students attending elementary and secondary school. We expect to find a positive association between the quality of TSRs and students’ social emotional outcomes, indicating that higher quality TSRs are related to more positive social emotional outcomes. Second, this study examines moderators of the association between
TSRs and students’ social-emotional outcomes. For this part of the research, we included the following moderators: TSR dimension (e.g., support, non-attuned relationship, dependency or total quality), TSR assessment (student-report, teacher-report or observation), reliability of TSR, social-emotional aspects (e.g. prosocial behavior, peer problems, social skills or sociability), social-emotional outcomes assessment (observation, student, teacher- or parent-report), reliability of social-emotional outcomes, publication year, study design (cross-sectional or longitudinal), impact factor of the journal, continent (USA, Europe or Asia), ethnicity (minority, majority or mixed), SES (high, low or mixed), students’ mean age and gender (male, female or mixed), teachers’ mean age, gender (male, female or mixed) and experience and school population (regular or mixed) and type of school (elementary or secondary).
Method Inclusion and Exclusion Criteria
Multiple inclusion criteria were formulated to select the studies for this meta-analysis. First, students’ social-emotional outcomes had to be operationalized in terms of interactions or relationships with peers, identity of self in relation to others, recognition of ability, expression of emotion, empathy, emotion regulation, impulse control and social understanding. We excluded students’ behavior problems, as more negative types of social-emotional outcomes. Second, the study had to report on the association between TSRs and students’ social-emotional outcomes in a way that made it possible to calculate an effect size. We only included studies that reported correlations between (aspects of) TSRs and (aspects of) students’ social-emotional outcomes. Third, we only included studies with students from elementary school, middle school and high school. We excluded studies with a sample that consisted of children in Preschool or Kindergarten and college students, except for one study
with a sample that contained 15% children in Kindergarten. Fourth, we only included studies that reported results in English and were published to assure quality of the studies.
Sample of studies
All studies addressing the relation between TSRs and students’ social-emotional outcomes which were published before February 2017 were included in the current meta-analysis. We used the PsycINFO, the Educational Resources Information Center (ERIC) databases and Google Scholar to retrieve relevant studies. The search string included several combined variables for the TSR element: relation* or interaction* and student* or pupil* and teacher*. For the students’ social-emotional outcomes, the following terms were used: social*, prosocial*, behavior*, behaviour*, empath*, impuls*, aware*, self-efficacy, ability*, competenc*, express*, regulat*, understand*, or interact*. With this search string we used both American and British English terms to include all studies on this subject. We also selected studies of the meta-analysis of Nurmi (2012).
The studies had to describe original data and should include factors that are associated with TSRs and students’ social-emotional outcomes. Our literature search strategy yielded a total of 285 studies. To determine whether the retrieved studies could be included in our meta-analysis, we read titles, abstracts and full article texts. After thoroughly screening these studies, we finally found 9 studies that met our inclusion criteria. In addition, the studies that Nurmi (2012) used in his meta-analysis on the association between the TSR and student outcomes were searched for qualifying studies. This yielded a total of 4 studies that met our inclusion criteria. Finally, the reference sections of review studies were searched for qualifying studies, and yielded a total of 4 studies that met our inclusion criteria.
Coding and Moderators
In developing a coding form, guidelines proposed by Lipsey and Wilson (2001) were followed. The variables of most interest were the dimensions of TSRs and aspects of the
social-emotional outcomes. The dimensions of the TSR were coded as they were reported in the studies. We coded the following dimensions of the TSR: conflict, closeness, dependency, emotional support, negative expectations, total quality and other. After reviewing all of the reported dimensions, we came to the following broad band categorization of TSR dimensions: support (e.g., closeness, emotional support, helping, understanding, psychological proximity seeking, emotional quality, TSR-positivity, total positive relationship, student attachment security and student overall appraisal of their teacher as an attachment figure), non-attuned (e.g., conflict, uncertain, dissatisfied, admonishing, strict and TSR-negativity), dependency, student freedom, leadership and total quality. Effect sizes on TSR dimensions were coded in the expected direction. We expected more positive dimensions (support, student freedom, leadership and total quality) to be positively associated with social emotional outcomes and more negative dimensions (non-attuned and dependency) to be negatively associated with social-emotional outcomes. Further, we coded how the TSR was assessed, and the reliability of these measurements (reported alpha’s). We distinguished between three types of assessment: student-report, parent-report, teacher-report and observation. For coding the social-emotional outcomes, we used the conceptualization of the California Department of Education (CDE, 2016). After reviewing the coded aspects of the emotional outcomes, we came to the following broad categorization of the social-emotional outcomes: social skills (e.g., social skills and interpersonal problem solving skills), peer problems, prosocial behavior, self-efficacy, sociability (e.g., social initiative, social engagement and sociable/cooperative behavior with peers) and inappropriate assertiveness. Effect sizes on social-emotional outcomes were coded in the expected directions as well. We expected more positive outcomes (social skills, prosocial behavior, self-efficacy and sociability) to be positively associated and more negative outcomes (peer problems and
inappropriate assertiveness) to be negatively associated with TSRs. The type of assessment and the reliability of the measurements of social-emotional outcomes were coded as well.
Finally, various study and sample characteristics with a potential effect on the association between TSRs and students’ social-emotional outcomes were identified. Study characteristics were publication year, the impact factor of the journal, whether the study had a cross-sectional or a longitudinal design and the location of the study. Sample characteristics were the proportion of ethnic minority subjects, the ethnicity of the sample (minority, majority or mixed), the proportion of Caucasian white subjects and the social economic status (SES; low, middle or mixed). Besides these characteristics, we coded student, teacher and school characteristics. The student characteristics were proportion male, mean age and gender (male, female or mixed). As none of the studies used an exclusively male or female sample, only a mixed sample was coded. Therefore, we were not able to perform analyses on this moderator, which consisted of only one category. The teacher characteristics were proportion of male teachers, gender of the teachers (female or mixed) and years of experience. The school characteristics were school population (regular or mixed population) and school type (elementary or secondary school).
To determine the inter-rater reliability of the coded variables, we calculated the intraclass correlation coefficient (ICC) for continuous variables and the kappa for categorical variables. We found an average ICC of 1 and an average kappa of .99.
Statistical Analyses
We calculated the correlation coefficient (r) because we were interested in the association between the TSR and students’ social-emotional outcomes. All statistics were converted to the correlation coefficient (r), and subsequently transformed in Fischer’s z-scores to be analyzed. Since extreme effect sizes may have a disproportionate influence on conclusions drawn from statistical analyses, we checked for outliers by searching for effect
sizes with standardized scores larger than 3.29 or smaller than −3.29 (Tabachnik & Fidell, 2013). No outliers were identified.
Most studies reported on multiple TSR dimensions or social-emotional aspects, and therefore more than one effect size could be extracted from these studies. A multilevel random effects model, including three levels, was used for the calculation of combined effect sizes and for the moderator analyses in order to account for statistical dependency of effect sizes (Hox, 2002; Van den Noortgate & Onghena, 2003). While level 1 is random sampling error, level 2 accounts for variance within studies, and level 3 for variance between studies (Assink et al., 2015). This model was used to obtain an overall estimate of the effect size and in case of significant variation between effect sizes from the same study and/or between studies, it was subsequently extended by including moderators to determine whether this variation can be explained by within or between study characteristics.
For the statistical analyses we used the function “rma.mv” of the metafor package (Viechtbauer, 2010) in the R environment (version 3.2.0; R Core Team, 2015). The R syntax was written so that the three sources of variance as described by for instance Van den Noortgate, López-López, Marin-Martinez, and Sánchez-Meca (2013, 2014) were modeled (Assink & Wibbelink, 2016). The t-distribution was used for testing individual regression coefficients of the meta-analytic models and for calculating the corresponding confidence intervals (Knapp & Hartung, 2003). When models were extended with categorical moderators, the omnibus test of the null hypothesis that all group mean effect sizes are equal, followed an F-distribution. To determine whether the variance between effect sizes from the same study (Level 2), and the variance between studies (Level 3) were significant, the deviance of the full model was compared to the deviance of a model excluding one of the variance parameters. The sampling variance of observed effect sizes (Level 1) was estimated by using the formula of Cheung (2014). All model parameters were estimated using the
restricted maximum likelihood estimation method, and before moderator analyses were conducted, each continuous variable was centered around its mean and dichotomous dummy variables were created for all categorical variables with three or more categories. The log-likelihood-ratio-tests were performed one-tailed and all other tests were performed two-tailed. We considered p-values equal to or greater than .05 as statistically significant.
Publication Bias
When conducting a meta-analysis, the risk of publication bias exists. Articles reporting non-significant results are less likely to be published than articles reporting significant results. To examine the possibility of publication bias, we conducted funnel plot analysis as described by Duval and Tweedie (2000a, 2000b) by using the function “trimfill” of the metafor package (Viechtbauer, 2010) in the R environment (Version 3.2.0; R Core Team, 2015). If there is no publication bias, the distribution of effect sizes is shaped as a symmetrical funnel, with the standard error on the y-axis and r (the observed effect size) on the x-axis. Among the available techniques for assessing the possibility of publication bias in a meta-analysis, the trim and fill method provides an estimate of the degree to which publication bias might affect the overall mean effect size (Nakagawa & Santos, 2012). In short, the trim and fill method restores the symmetry of an asymmetric funnel plot by imputing missing effect sizes that are calculated on the basis of existing effect sizes.
Results
Overall Association Between TSRs and Students’ Social-emotional Outcomes
The present study describes 17 studies (k) from 2004 to 2016. The total sample consisted of 6872 students (N), and the size of the samples described in the included studies (at start of the study) ranged from 44 to 1310 participants. The mean age of the participants at start of the studies was 10,15 years (SD = 2.34). Studies were conducted in the USA (k = 10),
Europe (k = 6), and in Asia (k = 1). In total, the coded studies produced 98 separate effect sizes, each reflecting the association between (a dimension of) TSRs and an aspect of students’ social-emotional outcome(s).
An overview of the overall association between the TSR and students’ social-emotional outcomes is presented in Table 2. A moderate and significant association was found between TSR and social-emotional outcomes (r = .302; 95% CI = .211 to .388; p < .001), indicating that a higher quality of TSR was associated with more positive social-emotional outcomes. The funnel plot (see Figure 1) showed that bias may be present at the right side of the funnel plot. The trim and fill analysis yielded a somewhat larger overall effect size after adding 11 effect sizes (r = .363; 95% CI = .289 to .433; p < .001).
Concerning the heterogeneity of the effect sizes, the likelihood ratio-tests showed significant variance within studies (% variance level 2: 26.8) and between studies (% variance level 3: 68.9). We concluded that the heterogeneity in effect sizes might be explained by TSR, social-emotional outcomes, study, and sample characteristics. Therefore, moderator analyses were conducted.
Moderator Analyses
The results of all moderator analyses are presented in Table 3, where moderators are classified into TSR characteristics, social-emotional outcomes characteristics, study characteristics, sample characteristics, student characteristics, teacher characteristics and school characteristics.
TSR and social-emotional outcomes characteristics. The TSR dimensions did not moderate the association between the TSR quality and students’ social-emotional outcomes, although the type of assessment of the TSR was found to have a moderating effect on the association between the TSR and students’ social-emotional outcomes. Stronger associations between the TSR and social-emotional outcomes were found for teacher reports of the TSR,
compared to students’ reports. The reliability of the TSR assessment was identified as a moderator between the TSR and students’ social-emotional outcomes as well. For ease of interpretation the unstandardized regression coefficient in Table 3 was standardized (β = .434,
p <.001), indicating that a higher reliability of TSR measurement yielded larger effect sizes.
Considering the social-emotional outcomes characteristics, the aspect of students’ social-emotional outcomes moderated the association between the TSR and social-emotional outcomes. For the sociability outcome, we found a weaker mean association than for the social skills outcome. We also found a moderating effect of the social-emotional outcomes assessment. Compared to teacher reports of students’ social-emotional outcomes, smaller effects for parent reports on outcomes were found. Reliability of the outcomes did not moderate the relation between TSR and social-emotional outcomes.
Study, sample, teacher and school characteristics. Publication year was found to moderate the association between TSR and students’ social-emotional outcomes. For ease of interpretation the unstandardized regression coefficient in Table 3 was standardized (β = -.529, p <.001), indicating that more recently published studies showed smaller effect sizes than early publications. Impact factor of the journal, study design, and continent of the study did not moderate the association between TSR and social-emotional outcomes in students. Considering sample characteristics, no moderating effects were found for ethnicity, proportion white, proportion ethnic minority and SES.
Students’ mean age moderated the association as well (β = -.537, p < .001), indicating that weaker associations between TSR and social-emotional outcomes were found in older students. No moderating effects were found for students’ gender, teachers’ age, teachers’ gender and teachers’ years of experience.
Considering school characteristics, moderating effects for both school population and type were found. A smaller effect of TSR quality on students’ social-emotional outcomes was
found for a regular school population, compared to a mixed school population. A stronger effect was found for students attending elementary school, compared to students attending secondary school.
Discussion
To examine the association between TSRs and students’ social-emotional outcomes, we conducted a multilevel meta-analysis. A moderate and significant association was found of r = .302, indicating that TSR quality is positively related to social-emotional outcomes of students. The results of the moderator analyses indicated moderating effects for the assessment of TSRs and social-emotional outcomes, reliability of TSRs assessments, aspects of social-emotional outcomes, publication year, mean age, school population and school type. Weaker associations were found for studies using student-reports of TSRs than studies using teacher-reports of TSRs. This moderating effect can be explained by teachers’ professional and therefore less biased perceptions of relationships with their students (Granot, 2016). As a result of their education and experience, teachers may be more self-reflective and objective in their perceptions on TSRs than students. We also found a moderating effect of the reliability of TSR measures: more reliable assessments of TSRs yielded larger effect sizes. This emphasizes the importance of reliable TSR measurement. Studies reporting more reliable TSR measurements often used attachment-based questionnaires for student and/or teacher-reports (Granot, 2016; Koomen & Jellesma, 2015; Sanchez Fowler, Banks, Anhalt, Hinrichs, & Kalis, 2008).
We found smaller effects between TSRs and social-emotional outcomes for studies measuring students’ sociability than social skills. A possible explanation is the distinction between trait and state. Sociability may be seen as a personality trait, which therefore is less open to change. We found larger effect size for teacher-reports than for parent-reports. This may be explained by the context in which respondents experience the child. The association
between TSR and student’s social-emotional development may be stronger in school than at home because teachers experience their students within the school context, where they interact with other students. They observe students’ social-emotional behavior directly in the classroom, and may be affected by or affect students’ behavior at the same time. Parents only experience their child at home, where children may show social-emotional behavior that not only differs from their behavior at school (Stams, Juffer, Rispens & Hoksbergen, 2000), but is also more influenced by parent-child interaction than teacher-child interaction (Stams et al., 2000). Future studies should therefore examine the degree to which possible effects of TSRs on students’ social-emotional behavior may generalize to the home environment.
Considering study characteristics, we found weaker associations for more recent studies. The quality of more recent studies may be higher than the quality of older studies, as the statistical and methodological knowledge has increased largely in social science research over the last decades (Bernard, 2013). Students’ mean age moderated the association as well in that weaker associations were found in older students. This may be explained by the intensity of teacher-student contact. As students grow older, they become less dependent on adults. This is in line with our finding of a moderating effect of school type. We found stronger associations between TSRs and students’ social-emotional outcomes for students attending elementary school rather than secondary school. In elementary school, when students are younger, the intensity as well as the frequency of teacher-student contact is high (Lynch & Cicchetti, 1997). In secondary school the relationship between teacher and student becomes less intense and frequent, as students switch between classrooms and teachers. Many adolescents feel increasingly detached from adults during this period, and as they become less dependent, TSRs tend to be less personal and less positive (Eccles, 1993; Murray, 2002). Considering the school population, we found stronger associations for a mixed than regular school population. In a mixed population, students with problem behavior
or learning disabilities are at higher risk for negative social-emotional outcomes (Blacher, Baker & Eisenhouwer, 2009; Murray & Greenberg, 2001). High quality TSRs may be more important for these students as a protective factor buffering developmental risks. This moderating effect may also be explained by a high extent of variability in social-emotional development in mixed samples. Because of this increased variability, larger effects may be found.
There are some limitations of the current study that need to be mentioned. The first limitation concerns the operationalization of the construct social-emotional development. This construct contains a broad range of related aspects, as the distinction between these aspects may not always be completely clear. Second, there might be case of selection bias. Effect sizes at the right sight of the funnel (Figure 1) are missing, which means there may be more case of selection bias than publication bias. Therefore, some samples may be underrepresented in the studies, like students attending schools for special education and samples of children with learning and/or behavior problems. A third limitation can be the amount of effect sizes our categories contain. Some categories contain less effect sizes than others, for example, observation of social-emotional outcomes, studies in Asia, student freedom, leadership and a low and mixed SES. Therefore, generalization of the results may be hampered, and statistical power was reduced.
Results of the present meta-analysis may have implications for clinical practice and future research on the role of TSRs for students’ social-emotional development. As already mentioned, samples of children with learning and/or behavioral problems may be underrepresented in this study. These children are also at higher risk for more negative social-emotional outcomes than their peers without any difficulties (Blacher, Baker & Eisenhouwer, 2009; Murray & Greenberg, 2001). Future research should therefore focus on this sample to investigate the role of TSRs in children with special needs. Moreover, future research should
examine whether the association between TSRs and school achievement may be mediated by social-emotional development. As mentioned, Roorda et al. (2011) found empirical evidence for the impact of both positive and negative aspects of TSRs on the students’ school engagement and achievement. Students’ social-emotional development may be of influence on attention, working memory and other prefrontal cortical processes contributing to learning in the classroom (Ursache, Blair, & Raver, 2012), and therefore may mediate the association between TSRs and students’ academic outcomes.
The results of this study indicate that there may be a role for TSRs in interventions targeting students’ social-emotional skills. Teachers, as well as other professionals within the school-context, should be aware of the impact they might have on their students’ social-emotional development. By improving a student’s social-social-emotional skills, teachers may need to change the way they interact with the student. When the teacher can manage to create a high quality TSR, marked by warmth and proximity, the student is more likely to engage in more positive social-emotional behavior (Murray, 2002). Thus, improving the quality of TSRs may result in more prosocial behavior within the school-context.
Focusing on TSRs as components of an intervention may also reduce students’ behavior problems within the school context (Baker, 2006; Meehan, Hughes, & Cavell, 2003; Murray & Pianta, 2007). Meta-analyses of Lei, Cui and Chiu (2016) and Van der Werff (2017) on the TSR and students’ psychopathology found a negative association between TSRs and problem behavior, indicating that high quality TSRs are associated with less problem behavior of students. Further research should focus on the role of TSRs in reducing students’ psychopathology, or preventing them from developing psychopathology by means of enhancing students’ social-emotional development.
Ang, R. P., Chong, W. H., Huan, V. S., Quek, C. L., & Yeo, L. S. (2008). Teacher–Student Relationship Inventory: Testing for invariance across upper elementary and junior high samples. Journal of Psychoeducational Assessment, 26, 339–349.
doi:10.1177/0734282908315132
Assink, M., & Wibbelink, C. J. M. (2016). Fitting three-level meta-analytic models in R: A step-by-step tutorial. The Quantitative Methods for Psychology, 12, 154-174. doi: 10.20982/tqmp.12.3.p154
Aviles, A. M., Anderson, T. R., & Davila, E. R. (2006). Child and adolescent social-emotional development within the context of school. Child and Adolescent Mental
Health, 11, 32-39. doi:10.1111/j.1475-3588.2005.00365.x
*Baker, J.A. (2006). Contributions of teacher-child relationships to positive school adjustment during elementary school. Journal of School Psychology, 44, 211-229. doi:10.1016/j.jsp.2006.02.002
Birch, S. H., & Ladd, G. W. (1997). The teacher-child relationship and children’s early school adjustment. Journal of School Psychology, 35, 61-79. doi:10.1016/S0022-4405(96)00029-5
*Blacher, J., Baker, B. L., & Eisenhouwer, A. S. (2009). Student-teacher relationship stability across early school years for children with intellectual disability or typical
development. American Association on Intellectual and Developmental Disabilities,
114, 322-339. doi:10.1352/1944-7558-114.5.322
Boocock, S. S. (1973). The school as a social environment for learning: Social organization micro-social process in education. Sociology of Education, 46, 15-50. doi:
10.2307/2112204
Bernard, H. R. (2013). Social Research Methods. Qualitative and Quantitative Approaches. California: SAGE Publications.
Bowlby, J. (1998). A secure base: clinical applications of attachment theory. London: Routledge.
Burt, K. B., Obradovic´, J., Long, J. D., & Masten, A. S. (2008). The interplay of social competence and psychopathology over 20 years: Testing transactional and cascade models. Child Development, 79, 359–374. doi:10.1111/j.1467-8624.2007.01130.x California Department of Education (2016). Social-Emotional Development Domain.
California Infant/Toddler Learning & Development Foundations. Retrieved from http://www.cde.ca.gov/sp/cd/re/itf09socemodev.asp
Cheung, M. W. L. (2014). Modeling dependent effect sizes with three-level meta-analyses: A structural equation modeling approach. Psychological Methods, 19(2), 211–229. doi:10.1037/a0032968
Connell, J. P., & Wellborn, J. G. (1991). Competence, autonomy, and relatedness: A
motivational analysis of self-system processes. In M. Gunnar & A. Sroufe (Eds.), The
Minnesota symposium on child development: Self-processes and development (Vol.
23, pp. 43-77). Hillsdale, NJ: Lawrence Erlbaum.
Cornelius-White, J. (2007). Learner-centered teacher-student relationships are effective: A meta-analysis. Review of Educational Research, 77, 113-143.
doi:10.3102/003465430298563
*Decker, D. M., Dona, D. P., & Christenson, S. L. (2007). Behaviorally at-risk African American students: The importance of student-teacher relationships for student outcomes. Journal of School Psychology, 45, 83-109. doi:10.1016/j.jsp.2006.09.004 Davis, H. A. (2001). The quality and impact of relationships between elementary school
children and teachers. Journal of Contemporary Educational Psychology, 26, 431– 453.
Davis, H. A. (2003). Conceptualizing the role and influence of student-teacher relationships on children's social and cognitive development. Educational Psychologist, 38, 207-234. doi:10.1207/S15326985EP3804_2
Durlack, J.A., Weissberg, R.P., Dymnicki, A.B., Taylor, R.D., & Schellinger, K.B. (2011). The impact of enhancing students’ social and emotional learning: A meta-analysis of school-based universal interventions. Child Development, 82, 405-432.
doi:10.1111/j.1467-8624.2010.01564.x
Duval, S., & Tweedie, R. (2000a). A nonparametric ‘trim and fill’ method of accounting for publication bias in meta-analysis. Journal of the American Statistical
Association,95(449), 89–99. doi:10.1080/01621459.2000.10473905
Duval, S., & Tweedie, R. (2000b). Trim and fill: A simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis. Biometrics, 56(2), 455- 463. doi:10.1111/j.0006-341X.2000.00455.x.
Eccles, J. S., Midgley, C., Wigfield, A., Buchanan, C. M., Rueman, D., Flanagan, C., et al. (1993). Development during adolescence: The impact of stage-environmental fit on young adolescents' experiences in schools and in families. American Psychologist, 48, 90–101. doi:10.1037/0003-066X.48.2.90
*Fisher, S. D., Reynolds, J. L., & Seehan, C. E. (2016). The protectivite effects of adaptability, study skills, and social skills on externalizing student-teacher relationships. Journal of Emotional and Behavioral Disorders, 24, 101-110. doi: 10.1177/1063426615598767
*Gehlbach, H., Brinckworth, M. E., & Harris, A. D. (2012). Changes in teacher-student relationships. British Journal of Educational Psychology, 62, 690-704.
*Granot, D. (2016) Socioemotional and behavioural adaptation of students with disabilities: the significance of teacher–student attachment-like relationships. Emotional and
Behavioural Difficulties, 21, 416-432. doi:10.1080/13632752.2016.1235324
Hawkins, & Catalano (1992). Communities that care: Action for drug abuse prevention. San Francisco: Jossey-Bass.
*Henricsson, L., & Rydell, A. M. (2004). Elementary school children with behavior problems: teacher-child relations and self-perception. A prospective study.
Merill-Palmer Quarterly, 50, 111-138. doi:10.1353/mpq.2004.0012
Hirschi, T. (1969). Causes of deliquency. Berkeley, CA: University of California Press. Hoeve, M., Stams, G.J.J.M., Van der Put, C.E, Dubas, J.S., Van der Laan, P.H., & Gerris, R.
M.. (2012). A meta-analysis of attachment and juvenile delinquency. Journal of
Abnormal Child Psychology, 40, 771-785.
Hox, J. (2002). Multilevel analysis: Techniques and applications. Mahwah, NJ: LawrenceErlbaum Associates.
Jerome, E. M., Hamre, B. K., & Pianta, R. C. (2009). Teacher-child relationships from kindergarten to sixth grade: Early childhood predictors of teacher-perceived conflict and closeness. Social Development, 18, 915-945.
doi:10.1111/j.1467-9507.2008.00508.x
Knapp, G., & Hartung, J. (2003). Improved tests for a random effects meta-regression with a single covariate. Statistics in Medicine, 22(17), 2693–2710. doi:10.1002/sim.1482 *Koomen, H. M. Y., & Jellesma, F. C. (2015). Can closeness, conflict, and dependency be
used to characterize students’ perceptions of the affective relationship with their teacher? Testing a new child measure in middle childhood. British Journal of
Ladd, G.W., Birch, S.H., & Buhs, E.S. (1999). Children’s social and scholastic lives in kindergarten: Related spheres of influence? Child Development, 70, 1373-1400. doi:10.1111/1467-8624.00101
Leflot, G., Onghena, P., & Colpin, H. (2010). Teacher–child interactions: Relations with children's self-concept in second grade. Infant and Child Development, 19, 385–405. doi:10.1002/icd.672.
Lei, H., Cui, Y., & Chiu, M. M. (2016). Affective teacher-student relationships and students’ externalizing behavior problems: A meta-analysis. Frontiers in Psychology, 7, 1311. doi:10.3389/fpsyg.2016.01311
Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Thousands Oaks, CA: Sage.
*Luckner, A. E., & Pianta, R. C. (2011). Teacher-student interactions in fifth grade classrooms: Relations with children’s peer behavior. Journal of Applied
Developmental Psychology, 32, 257-266. doi:10.1016/j.appdev.2011.02.010
Lynch, M., & Cicchetti, D. (1992). Maltreated children’s reports of relatedness to their teachers. In R. C. Pianta (Ed.), Beyond the parent: The role of other adults in
children’s lives (pp. 81– 107). San Francisco: Jossey-Bass.
*Martin, D. P., & Rimm-Kaufman, S. E. (2015). Do student self-efficacy and teacher-student interaction quality contribute to emotional and social engagement in fifth grade math?
Journal of School Psychology, 53, 359-373. doi:10.1353/mpq.2004.0012
McDonald-Connor, C. M., Son, S., Hindman, A., & Morrison, F. J. (2005). Teacher
qualifications, classroom practices, family characteristics and preschool experience: Complex effects on first graders’ vocabulary and early reading outcomes. Journal of
Meehan, B.T., Hughes, J.N., & Cavell, T.A. (2003). Teacher-student relationship as
compensatory resources for aggressive children. Child Development, 74, 1145-1157. doi:10.1111/1467-8624.00598
Murray, C., & Greenberg, M. T. (2000). Children’s relationship with teachers and bonds with school an investigation of patterns and correlates in middle childhood. Journal of
School Psychology, 38, 423-445. doi:10.1016/S0022-4405(00)00034-0
Murray, C., & Greenberg, M. T. (2001). Relationships with teachers and bonds with school: social emotional adjustment correlates for children with and without disabilities.
Psychology in the Schools, 38, 25-41.
doi:10.1002/1520-6807(200101)38:1<25::AID-PITS4>3.0.CO;2-C
Murray, C. (2002) Supportive Teacher-Student Relationships: Promoting the Social and Emotional Health of Early Adolescents with High Incidence Disabilities. Childhood
Education, 78, 285-290. doi:10.1080/00094056.2002.10522743
Murray, C., & Pianta, R. (2007). The importance of teacher-student relationship for adolescents with high incidence disabilities. Theory Into Practice, 46, 105–112. doi:10.1080/00405840701232943
Nakagawa, S., & Santos, E. S. A. (2012). Methodological issues and advances in biological meta-analysis. Evolutionary Ecology, 26(5), 1253–1274. doi:10.1007/s10682-0129555-5
Nurmi, J. (2012). Students’ characteristics and teacher-child relationships in instruction: A meta-analysis. Educational Research Review, 7, 177-197.
doi:10.1016/j.edurev.2012.03.001
*Obsuth, I., Murray, A. L., Malti, T., Sulger, P., Ribeaud, D., & Eisner, M. (2016). A non-bipartite propensity score analysis of the effects of teacher-student relationships on
adolescent problem and prosocial behavior. Jounal of Youth and Adolescence, 46, 1661-1687. doi:10.1007/s10964-016-0534-y
*Ocak, S. (2010). The effects of child-teacher relationships on interpersonal problem-solving skills of children. Infants and Young Children, 23, 312-322.
doi:10.1097/IYC.0b013e3181f27769
Oberle, E., Schonert-Reichl, K. A., Guhn, M., Zumbo, B. D., & Hertzman, C. (2014). The role of supportive adults in promoting positive development in middle childhood: A population-based study. Canadian Journal of School Psychology, 29, 296-316. doi:10.1177/0829573514540116
Pianta, R. C., Nimetz, S. L., & Bennett, E. (1997). Mother-child relationships, teacher-child relationships and adjustment in preschool and kindergarten. Early Childhood
Research Quarterly, 12, 263–280.
Pianta, R. C. (1999). Enhancing relationships between children and teachers. Washington, DC: American Psychological Association.
Pianta, R. C. (2001). Student-Teacher Relationship Scale. Professional manual. Lutz, FL: Psychological Assessment Resources
Pianta, R. C., Hamre, В., & Stuhlman, M. W. (2003). Relationships between teachers and children. In W. M. Reynolds, G. E. Miller, & I. B. Weiner (Eds.), Handbook of
psychology: Educational psychology. Hoboken, NJ: John Wiley.
Pianta, R. C., La Paro, K. M., Payne, C., Cox, M. J., & Bradley, R. (2002). The relation of kindergarten classroom environment to teacher, family, and school characteristics and child outcomes. The Elementary School Journal, 102, 225–240.
doi:10203-0003$05.00
*Pianta, R. C., & Stuhlman, M. W. (2004). Teacher-child relationships and children’s success in the first years of school. School Psychology Review, 33, 444-458.
*Poulou, M. (2015). Teacher-student relationships, social and emotional skills, and motivational and behavioral difficulties. International Journal of Educational
Psychology, 4, 84-108. doi:10.4471/ijep.2015.04
R Core Team (2015). R: A language and environment for statistical computing. Vienna,Austria: R Foundation for Statistical Computing (Retrieved from http://www.Rproject.org/).
Roorda, D.L., Koomen, H.M.Y., Spilt, J.L., & Oort, F.J. (2011). The influence of affective teacher-student relationships on students' school engagement and achievement: A meta-analytic approach. Review of Educational Research, 81, 493-529.
doi:10.3102/0034654311421793
Ryan, R. M. & Deci, E. L. (2002). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55, 68-78 Sabol, T. J., & Pianta, R. C. (2012). Recent trends in research on teacher-child relationships.
Attachment & Human Development, 14, 213-231.
doi:10.1080/14616734.2012.672262
*Sanchéz, Fowler, L. T., Banks, T. I., Anhalt, K., Hinrichs, H., & Kalis, T. (2008). The association between externalizing behavior problems, teacher-student relationship and academic performance in young urban learners. Behavioral Disorders, 33, 167-183. *Skalická, V., Stenseng, F., & Wichstrøm, L. (2015). Reciprocal relations between
student-teacher conflict, children’s social skills and externalizing behavior: A three-wave longitudinal study from preschool to third grade. International Journal of Behavioral Development, 39, 413-425. doi:10.1177/0165025415584187
Spilt, J. L., Koomen, H. M. Y., & Jak, S. (2012). Are boys better of with male and girls with female teachers? A multilevel investigation of measurement invariance and gender
match in teacher-student relationship quality. Journal of School Psychology, 50, 363-378. doi:10.1016/j.jsp.2011.12.002
Spilt, J.L., Koomen, H.M.Y. & Thijs, J.T. (2011). Teacher wellbeing: The importance of Teacher-Student Relationships. Educational Psychology Review, 23, 457-477. doi:10.1007/s10648-011-9170-y
Stams, G.-J. J., Juffer, F., Rispens, J., & Hoksbergen, A. C. (2000). The development and adjustment of 7-year-old children adopted in infancy. Journal of Child Psychology and Psychiatry, 41, 1025-1037. doi:10.1111/1469-7610.00690
Tabachnik, B. G., & Fidell, L. S. (2013). Using multivariate statistics (6th ed.). Boston: Allynand Bacon.
Ursache, A., Blair, C., & Raver, C. C. (2012). The promotion of self-regulation as a means of enhancing school readiness and early achievement in children at risk for school failure. Child Development Perspectives, 6, 122-128. doi:10.1111/j.1750-8606.2011.00209.x
Van den Noortgate, W., & Onghena, P. (2003). Multilevel meta-analysis: A comparison with traditional meta-analytical procedures. Educational and Psychological Measurement,
63(5), 765–790. doi:10.1177/0013164403251027
Viechtbauer, W. (2010). Conducting meta-analyses in R with the metafor package. Journal of
Statistical Software, 36(3), 1-48. Retrieved from
http://brieger.esalq.usp.br/CRAN/web/packages/metafor/vignettes/metafor.pdf Wentzel, K.R. (1993). Motivation and achievement in early adolescence: The role of multiple
classroom goals. Journal of Early Adolescence, 13, 4-20. doi:10.1177/0272431693013001001
31 Table 1
Characteristics of Included Studies
Author Year N IF Design Continent #ES Relationship outcome measures (#ES)
Social-emotional outcome measures (#ES)
Poulou 2015 962 Cross EU 24 Leadership (3)
Helping (3) Understanding (3) Student freedom (3) Uncertain (3) Dissatisfied (3) Admonishing (3) Strict (3) Social skills (8) Inappropriate assertiveness (8) Peer problems (8)
Luckner & Pianta 2011 894 1,97 Cross USA 2 Emotional support (2) Sociable/cooperative behavior- with peers (1)
Prosocial behavior (1) Koomen & Jellesma 2015 64 2,00 Cross EU 6 Negative -expectations (2)
Closeness (2) Conflict (2)
Peer problems (3) Prosocial behavior (3) Decker, Dona & Christenson 2007 44 3,00 Cross USA 6 Quality TSR (2)
Psychological proximity- seeking (2)
Emotional quality (2)
Social skills (6)
Henricsson & Rydell 2004 95 1,41 Cross EU 8 Total quality (2) Conflict (2) Dependency (2) Closeness (2)
Prosocial behavior (4) Social initiative (4)
Gehlbach, Brinkworth & Harris
2012 119 2,00 Cross USA 4 TSR-positivity (2) TSR-negativity (2)
Self-efficacy (4)
Baker 2006 1310 3,36 Cross USA 3 Total quality (1)
Closeness (1)
32 Conflict (1)
Fisher, Reynolds & Sheehan 2016 209 1,95 Cross USA 1 Total quality (1) Social skills (1) Sanchez Fowler, Banks,
Anhalt, Hinrichs & Kalis
2008 230 1,61 Cross USA 3 Conflict (1) Closeness (1)
Total positive relationship (1)
Prosocial behavior (3)
Skalicka, Stenseng & Wichstrøm
2015 981 1,61 Long EU 6 Conflict (6) Social skills (6)
Martin & Rimm-Kaufman 2015 387 3,36 Cross USA 1 Emotional support (1) Social engagement (1) Blacher, Baker &
Eisenhower
2009 98 2,30 Long USA 7 Conflict (3) Closeness (3) Total quality (1)
Social skills (7)
Eisenhower, Baker & Blacher
2007 140 3,36 Cross USA 2 Emotional support (2) Social engangement (2)
Granot 2016 65 0,51 Cross AS 3 Student attachment security (1)
Student overall appraisal of -teacher as secure figure (1) Total quality (1)
Sociability (3)
Ocak 2010 102 0,59 Cross EU 3 Conflict (1)
Closeness (1) Dependency (1)
Interpersonal problem solving-skills (3)
Obsuth, Murray, Malti, Sulger, Ribeaud & Eisner
2016 682 3,56 Cross EU 15 Total quality (15) Prosocial behavior (15)
Pianta & Stuhlman 2004 490 1,75 Long USA 4 Conflict (2) Closeness (2)
Social competence (4)
Note. N = number of participants; # ES = number of effect sizes (mean); IF = impact factor of journal; design = cross-sectional or longitudinal; Continent = location of study; Cross = cross-sectional design; Long = longitudinal design; AS = Asia
33 Table 2
Overall Association Between the Teacher-student Relationship and Students’ Social-emotional Outcomes
Outcome k #ES Mean r 95% CI p σ2level2 σ2level3 % Var. level 1 % Var. level 2 % Var. level 3 16 94 .302 0.211 – 0.388 <.001*** <.001*** <.001*** 4.3 26.8 68.9 Note. s = number of studies; k = number of effect sizes; CI = confidence interval; Mean r = mean effect size (r); CI = confidence interval; % Var = percentage of variance explained; σ2
level2 = variance between effect sizes within the same study; σ2
34 Table 3
Moderator Effect of Association Between the TSR and Students’ Social-emotional Outcomes Moderator variables k #ES β0
(mean r) β0 (Fischer’s z) t0 β₁ t1 F(df1, df2) TSR characteristics TSR 17 98 F(5, 88) = 1.526 Total quality (RC) 9 25 .269 .276 4.125*** Support 13 30 .316 .327 5.932*** .051 .808 Non-attuned 10 34 .316 .327 5.712*** .051 .782 Dependency 2 3 .311 .322 2.977** .047 .424 Student freedom 1 3 .147 .148 1.682* -.128 -1.362 Leadership 1 3 .337 .351 3.996*** .076 .804 Assessment TSR 17 98 F(2, 91) = 3.233* Student report (RC) 7 47 .264 .270 4.927*** Teacher report 13 48 .350 .365 7.555*** .095 2.096* Observation 2 3 .130 .131 .997 -.138 -.968 Reliability TSR (c) 12 63 .288 3.942*** 0.853 3.942*** F(1, 91) = 15.540*** Social-emotional outcomes characteristics Social-emotional outcomes 17 98 F(5, 88) = 2.348* Social skills (RC) 9 40 .362 .379 6.313*** Peer problems 2 11 .278 .286 3.998*** -.093 -1.741 Prosocial behavior 5 26 .322 .334 4.644*** -.045 -.523 Self-efficacy 1 4 .189 .191 1.053 -.188 -.985 Sociability 4 9 .143 .144 1.734* -.235 -2.375* Inappropriate assertiveness 1 8 .370 .388 5.064*** .009 .154 Assessment outcomes 17 98 F(2, 91) = 5.760** Teacher report (RC) 13 50 .335 .349 6.453*** Student report 6 41 .292 .301 4.876*** -.048 -.990 Parent report 2 4 .052 .052 .593 -.297 -4.005*** Observation 2 3 .243 .248 1.726* -.100 -.726 Reliability outcomes (c) 14 63 .341 6.560*** .977 1.729 F(1, 92) = 2.990 Study characteristics Publication year (c) 17 98 .296 6.810*** -.025 -2.379* F(1, 92) = 5.661* Impact factor (c) 16 74 .319 5.801*** .003 .055 F(1, 68) = 0.003 Study design 17 98 F(1, 92) = 0.547 Cross-sectional (RC) 17 95 .304 .314 6.344*** Longitudinal 2 3 .245 .250 2.595* -.063 -.740 Continent 16 94 F(2, 91) = 2.232 USA (RC) 10 33 .379 .399 6.522*** Europe 6 62 .204 .207 2.980** -.161 -.812 Asia 1 3 .234 .238 1.262 -.192 -2.075* Sample characteristics
Proportion ethnic minority (c) 11 59 .333 5.076*** .002 1.295 F(1, 57) = 1.677
Ethnicity 14 68 F(2, 65) = 0.280 Mixed sample (RC) 11 36 .358 .375 5.551*** Majority sample 2 30 .264 .270 1.941* -.104 -.674 Minority sample 1 6 .277 .284 1.358+ -.091 -.414 Proportion white (c) 11 57 .309 4.600*** -.004 -1.975 F(1, 51) = 3.900 SES 5 12 F(2, 9) = 0.822 Middle (RC) 2 6 .356 .372 1.957 Low 1 3 .605 .701 2.652* .330 1.013 Mixed 2 7 .234 .238 .877 -.134 -.403
35 Student characteristics Proportion male (c) 16 83 .335 7.067*** .005 1.345 F(1, 77) = 1.809 Mean age (c) 12 78 .254 5.955*** -.047 -3.674*** F(1, 76) = 13.498*** Teacher characteristics Proportion male 11 43 .366 5.367*** -.004 -.604 F(1, 37) = 0.365 Gender 12 54 F(1, 52) = 0.073 Mixed (RC) 10 52 .323 .335 4.447*** Female 2 6 .278 .286 1.755 -.049 -.270 Years of experience (c) 4 14 .197 2.582* -.008 -.529 F(1, 8) = 0.280 School characteristics School population 16 94 F(1, 92) = 7.696** Regular (RC) 12 79 .242 .247 5.314*** Mixed 4 15 .471 .511 6.170*** .263 2.774** School type 17 98 F(1, 92) = 4.302* Elementary (RC) 17 88 .308 .319 6.724*** Secondary 1 10 .185 .187 2.446* -.131 -2.074*
Note. In case of a continuous measure the regression coefficient is standardized.
Note. k = number of independent studies; #ES = number of effect sizes; β₀ = intercept/mean effect size (r); t0 =
difference in mean r with zero; β₁ = estimated regression coefficient; t1 = difference in mean r with reference category; F(df1, df2) = omnibus test; (RC) = reference category; (c) = continuous variable.
36 Figure 1
Funnelplot Trim and Fill Analysis
Funnelplot - social-emotional Fischer z S tandar d E rr or 0. 156 0. 121 0. 087 0. 052 0. 017 -0.20 0.00 0.20 0.40 0.60 0.80
