Teacher-Student Relationships and Students’ Psychopathology
A systematic review and meta-analysis
Name: F.A. van der Werff Student ID: 10002930
Master thesis Forensic Child and Youth Care Sciences Graduate School of Child Development and Education University of Amsterdam
Supervisor: Geert-Jan Stams, Francine Jellesma & Mark Assink Date: 11-07-2017
2
Table of Contents
1. Introduction ... 4
2. Method ... 9
2.1 Inclusion and exclusion criteria... 9
2.2 Sample of studies ... 10
2.3 Publication bias ... 11
2.4 Coding and moderators ... 11
2.5 Statistical analyses... 12
3. Results ... 14
3.1 Descriptives and overall association between TSRs and students’ psychopathology ... 14
3.2 Moderator analyses ... 16
3.2.1 TSR characteristics and psychopathology outcome characteristic ... 16
3.2.2 Study characteristics ... 16 3.2.3 Sample characteristics ... 17 3.2.4 Student characteristics ... 17 3.2.5 Teacher characteristics ... 17 3.2.6 School characteristics ... 17 4. Discussion... 23 References ... 28
3
Abstract
This meta-analysis examined the association between the teacher-student relationship (TSR) and students’ psychopathology (i.e., externalizing and internalizing problem behavior) in elementary and secondary school, and possible moderators of this association. In total, 26 studies, with 196 independent effect sizes were included. The results show a significant negative association between TSRs and students’ psychopathology (r = .311), indicating that a high quality TSR is associated with less psychopathology in students. The association was stronger for externalizing problem behavior (r = .335) than for internalizing problem behavior (r = .227). In addition, non-attuned TSRs (r = .426) showed a larger effect size than supportive TSRs (r = .206). In addition, the assessment of the TSR (teacher-report and observation showed larger effects sizes than student report), reliability of the TSR assessment, assessment of psychopathology outcomes (a larger effect size for teacher- than for parent-report), reliability of psychopathology assessment outcomes, study design (a larger effect for cross-sectional than for longitudinal studies), students’ mean age (smaller effect with the increase of students’ age), and school population (a larger effect for a mixed than for a regular population) moderated the association between the TSR and students’ psychopathology. The discussion focuses on implications for theory and practice.
Keywords: teacher-student relationship, psychopathology, internalizing problem behavior, externalizing problem behavior.
4
1. Introduction
The school is a social environment where children are prepared, cognitively as well as socially, for participation in society (Boocock, 1973; Wentzel, 1993). The focus lies not only on acquiring knowledge and cognitive skills, but also on the development of social competences (Pianta, 1999), self-regulation skills (Baker, 2006; Eisenhower, Baker, & Blacher, 2007), personality development (Heaven, Leeson, & Ciarrochi, 2009), moral socialization (Cooley, 2008), and for some even, acquiring democratic skills (Mager, & Nowak, 2012). Teachers as well as students are participants in social interactions, both at the group level and dyadic (Poulou, 2015; Yeon, Schwartz, Cappella, & Seidman, 2014). Relationships that students have with their teachers have been shown to be associated with both their social-emotional development and learning achievements (Aviles, Anderson & Davila, 2006; Pianta, Nimetz & Bennett, 1997; Roelofs, 2017; Roorda et al., 2011). The present meta-analysis examines the relation between teacher-student relationship (TSR) and students’ psychopathology in elementary and secondary school.
Where Roorda et al. (2011) and Roelofs et al. (2017), in their meta-analyses, found empirical evidence for a relation between the teacher-student relationship (TSR) and students’ learning achievements and socio-emotional development, respectively, the current meta-analysis focuses on the association between the quality of the teacher-student relationship and students’ psychopathology (e.g., internalizing and externalizing behavioral problems). Although Nurmi conducted a meta-analysis in 2012 on the relation between students’ characteristics, among which were their behavioral problems, and the teacher-student relationship, and Lei, Cui and Chiu (2016) more recently conducted a meta-analysis on teacher-student relationships and students’ externalizing behavior problems, a meta-analysis on the quality of teacher-student relationships as a predictor of both internalizing (e.g., depression, anxiety) and externalizing (e.g., aggression, delinquency) problems of
5 students has not yet been conducted. This quantitative review study will fill that gap by conducting a multi-level meta-analysis, which focuses on factors that may moderate the association between TSRs and students’ psychopathology, accounting for both within and between study differences in effect sizes.
There is a vast body of empirical research examining the association between TSRs and students’ psychopathology and school-based outcomes (Davis, 2001; Pianta, 1997). Roorda et al. (2011) integrated the theories and knowledge on the association between TSRs and school-outcomes, such as students’ school engagement and achievement, by conducting a meta-analysis. They found evidence of the impact of both positive and negative aspects of the TSR on the students’ school engagement and academic achievement. Although there has been much empirical research and theories suggesting a relation between the TSR and students’ psychopathology, this knowledge has not yet been sufficiently integrated (Decker, Dona, & Christenson; Howes et al., 2008; Lei et al., 2016).
Teacher-student relationships
Over the past two decades, there has been an increasing interest in the association between the quality of relationships between students and teachers and 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, which are (extended) attachment theory and social-motivational self-determination theory (Ryan & Deci, 2002).
From the attachment perspective, TSRs may be considered as 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 creates emotional security in the child if the attachment figure (caregiver) provides for a
6 secure base and haven (Bowlby, 1988). Emotional security is in turn considered to be a necessary precondition for exploration of the environment. According to attachment theory, teachers’ nurturing and responsiveness to students’ needs can 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 the students’ attempts to cope with the demands of the social 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).
According to social control theory, which is also an attachment-based theory (Hoeve et al., 2012), the feeling of belonging and connectedness within the school context is important for students (Hirschi, 1969). This feeling of connectedness allows students to develop their prosocial skills and behaviors, and increases their involvement with social non-deviant groups. They are less likely to engage in non-deviant behaviors because these behaviors can threaten their relationships with significant others within the school context.
Because TSRs can be seen from the perspective of child-caregiver attachment, the affective quality of the TSRs are often assessed by using three dimensions 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 refers to a more negative relationship (Roorda et al., 2011).
7 According to self-determination theory (Ryan & Deci, 2002) 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 TSRs derived from attachment theory, while relatedness is connected to the concept of emotional security (Connell & Wellborn, 1991).
A relationship that builds on respect for autonomy in social interactions will create a feeling of connectedness, prosocial behavior and responsibility, which may result in less dysfunctional or maladaptive behavior in students (Rudasill, Reio, Stipanovic, & Taylor, 2010). In contrast, negative TSRs nourish psychopathology, due to a lack of acceptance, conflict, and support in autonomy (Pianta & Stuhlman, 2004; Stewart & Suldo, 2011). The TSR can facilitate a protective influence for students’ psychopathology, by creating an environment for students, in which they experience a feeling of connectedness with their environment.
In summary, a TSR that is characterized by closeness, respect for autonomy and competence, and involvement of the teacher may prevent the development of students’ psychopathology or might buffer against risks for psychopathology (Murray & Murray, 2004), because of the emotional security that is provided by teachers (Connell & Wellborn, 1991; Roorda et al., 2011). On the other hand, a TSR characterized by conflict or dependency, low involvement, and a lack of respect for the students’ basic needs, may increase the development of students’ psychopathology (Connell & Wellborn, 1991, Roorda
8 et al., 2011). Both these allegations can be explained from the theories mentioned earlier, that is, attachment theory and social-determination theory.
Impact of TSRs: The Role of student, teacher and study characteristics
The strength of the association between TSRs and students’ psychopathology may be influenced by other factors, such as characteristics related to TSRs, different forms of psychopathology, and sample and study characteristics. Considering TSR characteristics, Lei et al. (2016) found evidence that a TSR characterized by conflict was strongly linked to externalizing problem behavior. The type of assessment may also moderate the association between TSR and psychopathology, since teacher-report has been shown to yield a stronger association with externalizing behavior than student-report (Lei et al., 2016). Both externalizing and internalizing problem behavior have been shown to be associated with quality of the TSR in several studies, with externalizing behavior generally showing the strongest association (Drugli, 2013; Henricsson & Rydell, 2004; Murray & Murray, 2004; Silver, et al., 2005).
Student characteristics may also influence the TSR. Nurmi (2012) concluded that students who already exhibit a higher level of problem behavior tend to experience less closeness in the relationship with their teacher. Various studies found that boys tend to have relationships characterized by conflict and less support, whereas girls were found to have closer relationships with their teachers (Baker, 2006; Drugli, 2013; Leflot, van Lier, Verschueren, Onghena, & Colpin, 2011).
Different sample characteristics and study characteristics may influence the association as well. Rudasill et al. (2011) found that children in families with a low income rated higher on conflict in the TSR.
Considering school characteristics, school type and school population may possibly moderate the association between the TSR and psychopathology outcomes of students.
9 Research on differences in TSRs of students with and without intellectual or other disabilities showed that students with disabilities tend to have lower quality TSRs (Blacher, Baker, & Eisenhower, 2009; Murray & Greenberg, 2001).
Present study
The present study is a multilevel meta-analysis on the association between TSRs and students’ psychopathology. We expect to find a negative association between TSRs and students’ psychopathology. Second, this study examines moderators of the association between TSRs and students’ psychopathology. The following moderators were included: TSR dimension (e.g., support, non-attuned relationships, dependency), assessment of the TSR outcomes (student report, teacher report, or observation), aspects of psychopathology (internalizing or externalizing problem behavior), assessment of psychopathology outcomes (e.g., parent- or teacher-report), reliability of psychopathology assessments, publication year, impact factor of the journal, study design (cross-sectional or longitudinal), country (USA, Europe, Canada or Other), ethnicity (majority, mixed or minority sample), SES (low or middle to high), students’ mean age, and gender (girls, boys, mixed), teachers’ gender (male, female, mixed), school population (e.g., regular or special education), and school type (elementary or secondary school).
2. Method
2.1 Inclusion and exclusion criteria
Multiple inclusion criteria were formulated to select the studies for this meta-analysis. First, psychopathology had to be operationalized in terms of internalizing (i.e., anxiety, depression, and withdrawal) or externalizing problem behavior (i.e., aggression, disruptive behavior, conduct problems, anger, oppositional defiant behavior, delinquency, attention en hyperactivity problems). Other types of psychopathology outcomes were excluded. Second,
10 the study had to report about the association between the TSR and students’ psychopathology in a way that made it possible to calculate effect sizes. Only studies that reported correlations between (dimensions of) TSRs and (aspects of) students’ psychopathology were included. Third, we only included studies with students from elementary school and secondary school. We excluded studies with a sample that consisted of children in Preschool or Kindergarten. Fourth, we only included studies that reported in English.
2.2 Sample of studies
All studies addressing the relation between the TSR and students’ psychopathology, which were published before February 2017 were included in the current meta-analysis. We used PsycINFO, and 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* or child* and teacher* or teacher support*. For the students’ psychopathology, the following terms were used: psychopathology*, internalizing*, internalizing*, anxiety*, anxious*, depress*, withdraw*, externalizing*, externalizing*, aggress*, delinquen*, conduct*, behavior*, behavior*, disrupt*, anger, oppositional, attention*, hyper*. With this search string, we used both American and British English terms. The studies had to describe original data and should include factors that are associated with TSRs and students’ psychopathology.
Our literature search strategy yielded a total of 277 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, 14 studies were found that met the 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, which yielded a total of five studies that met the inclusion criteria. The studies that Lei et al. (2016) used in their meta-analysis on the association between TSR and externalizing
11 behavior were also searched; this yielded a total of three extra studies. The reference sections of review studies were searched for qualifying studies and yielded a total of three studies that met the inclusion criteria. Finally, requesting studies from authors yielded in one study that met the inclusion criteria.
2.3 Publication bias
When conducting a meta-analysis, the risk of publication bias consists. 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 a 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.
2.4 Coding and moderators
In developing a coding form, guidelines proposed by Lipsey and Wilson (2001) were followed. The dimensions of the TSR were coded as they were reported in the studies: conflict, closeness, dependency, emotional support, negative expectations, total quality and other. After reviewing all the reported dimensions and data, we came to the following categorization of TSR: support, non-attuned, dependency, student freedom, leadership and total quality. Furthermore, we coded how the TSR was assessed and the reliability of these measurements. We made a distinction between three types of assessment: student-report,
12 teacher-report and observation. For coding the student’s psychopathology, we used the broadband conceptualization of externalizing (i.e., conduct problems, hyperactivity, aggression, asocial behavior, problem behavior, delinquency and oppositional defiant behavior) and internalizing (i.e. depression and anxiety) problems (Drugli, 2013; Lei et al., 2016). The type of assessment (teacher, student-, parent-, and peer-report, and composite) and the reliability of the measurements of psychopathology outcomes were coded as well.
Finally, various study and sample characteristics with a potential effect on the association between the TSR and students’ psychopathology were coded. Study characteristics were publication year, the impact factor of the journal, whether the study had a cross-sectional or a longitudinal design, and the country where the study was conducted. Sample characteristics were the proportion of subjects with an ethnic minority background, the ethnicity of the sample (minority, majority or mixed), the proportion of Caucasian white subjects and the social economic status (SES; low, and middle). Besides these characteristics, student, teacher and school characteristics were coded. 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 mixed sample was coded. Therefore, we were not able to perform analyses on this possible 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 following school characteristics were coded: school population (regular, special education, conduct problems or mixed population) and school type (elementary or secondary school).
2.5 Statistical analyses
The correlation coefficient (r) was calculated because we were interested in the association between the TSR and students’ psychopathology. All statistics were converted to the correlation coefficient (r), and subsequently transformed in Fischer z-scores to be
13 analyzed. Since extreme effect sizes may have a disproportionate influence on conclusions drawn from statistical analyses, outliers were checked 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 psychopathology 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 (Wibbelink & Assink, 2015). In case of significant variation between effect sizes from the same study and/or between studies, moderator analyses were conducted to determine whether this variation could 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 (Wibbelink & Assink, 2015). 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
14 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 for 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 greater than .05 as statistically significant.
3. Results
3.1 Descriptives and overall association between TSRs and students’ psychopathology
The present study describes 26 studies (k) from 2001 to 2016. The total sample consisted of 12556 (N) students in elementary and secondary school, and the size of the samples described in the included studies (at start of the study) ranged from 44 to 1554 participants. The mean age of the participants at start of the studies was 9.94 years (SD = 2.36). Studies were conducted in the USA (k = 15), Canada (k = 1), Europe (k = 8), Australia (k = 1), and in Israel (k = 1). In total, the coded studies produced 196 separate effect sizes, each reflecting the association between the TSR and an aspect of psychopathology (externalizing or internalizing problem behavior).
An overview of the overall association between the TSR and students’ psychopathology is presented in Tabel 2. A moderate and significant association was found between the TSR and students’ psychopathology (r = .311; 95% CI = .249 to .391; p = < .001), indicating that a higher quality of TSR was associated with less psychopathology in students. The funnel plot (Figure 1) showed that studies were missing on the right side of the funnel plot, suggesting selection bias, but not publication bias. The trim and fill analysis yielded a somewhat larger effect size after adding 45 effect sizes (r = .411; 95% CI = .387 to .491; p = <.001).
15 Figure 1. Funnel plot
Concerning the heterogeneity of the effect sizes, the likelihood ratio-test on the second level of variance showed that there was significant variance within studies. On the third level of variance, there was a significant variance as well, indicating that there was variance in effect sizes between studies. Since it is shown that the level two and three variances were significant, we concluded that the heterogeneity among the effect sizes might be explained by TSR, psychopathology, study, and sample characteristics. Therefore, moderator analyses were conducted.
Funnelplot - psychopathology Fischer z S tandar d E rr or 0. 156 0. 121 0. 087 0. 052 0. 017 0.00 0.50 1.00
16 3.2 Moderator analyses
The results of all moderator analyses are presented in Table 3, where moderators are classified into TSR, psychopathology outcomes, study, sample, student, teacher and school characteristics.
3.2.1 TSR characteristics and psychopathology outcome characteristic
The TSR dimension significantly moderated the association between the TSR and students’ psychopathology. No association was found for the dimension student freedom, while a strong association was found for non-attuned TSRs. Type of TSR assessment moderated the association between the TSR and students’ psychopathology. Stronger associations were found for teacher-report and observation of the TSR than for student- report of the TSR. Greater reliability of the TSR yielded larger effect sizes (β = .073, p < .01). The aspects of students’ psychopathology outcomes did have a moderating effect on the association between the TSR and psychopathology. Externalizing problem behavior showed a moderate effect and a smaller effect was found for internalizing problem behavior. We also found that outcomes assessment moderated the TSR-psychopathology association, as larger effects were found for student- and teacher-report. No significant effects were found for parent-report of psychopathology. More reliable assessment of psychopathology outcomes yielded larger effect sizes (β = .157, p < .001).
3.2.2 Study characteristics
Study design was found to have a moderating effect on the association between the TSR and psychopathology. Studies with a cross-sectional design yielded a moderate effect, while longitudinal studies showed a small effect. None of the other study characteristics (i.e., publication year, impact factor of the journal, and country) significantly moderated the association between TSR and students’ psychopathology.
17 3.2.3 Sample characteristics
None of the sample characteristics (i.e., proportion ethnic minority, ethnicity, proportion white, and SES) significantly moderated the association between the TSR and students’ psychopathology.
3.2.4 Student characteristics
Students’ mean age was found to moderate the association of the TSR and psychopathology, indicating that weaker associations were found when children were older (β = -.262, p < .001). No significant moderator effect was found for proportion of male students.
3.2.5 Teacher characteristics
None of the teacher characteristics (i.e., proportion of male teachers, gender and years of experience) significantly moderated the association between the TSR and students’ psychopathology.
3.2.6 School characteristics
The school population was found to moderate the association between TSR and students’ psychopathology. A strong association was found for a mixed school population, whereas a moderate association was found for a regular school population, and no significant association was found for students with conduct problems and special education classes. No significant effects were found for school type either.
18
Table 1
Characteristics of included studies
Author Year N IF Design Continent #ES Relationship outcome measures (#ES) Social-emotional outcome
measures (#ES)
Baker 2006 1310 3,36 Cross USA 6 Total quality (2)
Closeness (2) Conflict (2)
Externalizing (3) Internalizing (3) Blacher, Baker, & Eisenhower 2009 98 2,30 Cross/Long USA 7 Conflict (3)
Closeness (3) Total quality (1)
Problem behavior (7)
Crum, Waschbusch, & Willoughby 2016 1554 1,95 Cross CAN Closeness (1) Conflict (1)
ODD (2) Decker, Dona, & Christenson 2007 44 3,00 Cross USA 6 Total quality (1)
Psychological proximity- seeking (1) Emotional quality (1)
Problem behavior (3)
Drugli 2013 825 0,41 Cross EU 4 Closeness (2)
Conflict (2)
Externalizing (2) Internalizing (2)
Drugli, Klökner, & Larsson 2011 825 Cross EU 4 Closeness (2)
Conflict (2)
Internalizing (2) Externalizing (2) Eisenhower, Baker, &
Blacher
2007 140 3,36 Cross USA 2 Total quality (2) Problem behavior (2)
Eisenhouwer, Blacher, & Bush 2015 166 1,96 Long USA 12 Conflict (6) Closeness (6)
Externalizing (12)
Granot 2016 65 0,51 Cross ISR 3 Student attachment security (2)
Student overall appraisal of teacher as secure figure (2)
Total quality (2)
Externalizing (3) Internalizing (3)
Henricsson, & Rydell 2004 95 1,41 Cross EU 8 Total quality (2) Conflict (2) Dependency (2) Closeness (2)
Externalizing (4) Internalizing (4)
Koomen, & Jellesma 2015 64 2,00 Cross EU 9 Negative expectations (3) Closeness (3)
Conflict (3)
Internalizing (3) Conduct problems (3) Hyperactivity (3) Leflot, van Lier, Verschueren,
Onghena & Colpin
2011 570 3,31 Cross/Long EU 16 Teacher Support (16) Externalizing (16)
19
Relational aggression (1) Asocial behavior (1)
Madill, Gest & Rodkin 2014 628 1,75 Cross USA 1 Closeness (1) Aggressive-disruptive behavior
Murray, & Greenberg 2001 193 1,04 Long USA 16 Affiliation with teacher (8) Dissatisfaction with teacher (8)
Delinquency (4) Depression (4) Anxiety (4)
Conduct problems (4)
Murray, & Murray 2004 99 1,04 Cross USA 6 Conflict (2)
Closeness (2) Dependency (2)
Externalizing (3) Internalizing (3)
Murray, & Zvoch 2011 64 1,95 Cross USA 6 Conflict (3)
Closeness (3)
Conduct problems (2) Depression (2) Externalizing (2) Obsuth, Murray, Malti, Sulger,
Ribeaud, & Eisner
2016 682 3,56 Cross EU 24 Total quality (24) Aggression (15)
Oppositional defiant behavior (9)
Pianta, & Stuhlman 2004 490 1,75 Cross USA 8 Closeness (4)
Conflict (4)
Internalizing (4) Externalizing (4)
Poulou 2015 962 Cross EU 24 Leadership (3)
Helping (3) Understanding (3) Student freedom (3) Uncertain (3) Dissatisfied (3) Admonishing (3) Strict (3) Internalizing (8) Conduct problems (8) Hyperactivity (8)
Runion, Vitaro, Cross, Shaw, Hall, & Boivin
2014 374 1,41 Cross AUS 2 Conflict (1) Closeness (1)
Aggression (2)
Sanchez Fowler, Banks, Anhalt, Hinrichs, & Kalis
2008 230 Cross USA 3 Conflict (1)
Closeness (1)
Total positive relationship (1)
Externalizing (3)
Skalická, Stenseng, & Wichstrøm 2015 981 1,61 Long EU 6 Conflict (6) Externalizing (6)
Stewart, & Suldo 2011 390 1,04 Cross USA 2 Teacher Support (2) Internalizing (1)
20
Stipek, & Miles 2008 403 3,79 Cross USA 4 Conflict (4) Aggression (4)
Troop-Gordon, & Kopp 2011 410 1,80 Long USA 12 Closeness (4)
Conflict (4) Dependency (4)
Relational aggression (6) Physical aggression (6) 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; ISR = Israel; CAN = Canada, AUS = Australia
21
Table 2.
Overall association between the teacher-student relationship and student’s psychopathology
Outcome k #ES Mean r 95% CI p σ2level2 σ2level3 %
Var. level 1 % Var. level 2 % Var. level 3 26 196 .311 .249-.372 <.001*** <.001*** <.001*** 3.4 60.6 36.0
Note. k = number of studies; #ES = number of effect sizes; 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
22
Table 3.
Moderator effects of association between TSR and students’ psychopathology
Moderator variables k #ES β₀
(mean r/ β) t0 β₁ t1 F(df1, df2) TSR characteristics TSR 26 196 F(5, 190) = 10.650*** Total quality (RC) 8 35 .325 5.228*** Support 22 78 .206 5.634*** -.127 -1.909+ Non-attuned 18 69 .426 11.800*** .119 1.741+ Dependency 3 8 .017 3.914*** -.017 -.172 Student freedom 1 3 .170 1.524 -.164 -1.300 Leadership 1 3 .335 3.088** .013 .099 Assessment TSR 26 196 F(2, 193) = 2.186+ Student report (RC) 8 70 .253 4.973*** Teacher report 21 123 .339 9.691*** .094 1.748+ Observation 1 3 .321 6.835*** -.129 -2.035* Reliability TSR (c) 20 119 .335 8.506*** .511 1.870* F(1, 115) = 4.277** Psychopathology characteristics Psychopathology outcomes 26 196 F(1, 194) = 8.498** Internalizing 12 43 .227 4.862*** Externalizing 26 153 .335 9.571*** .119 2.915** Assessment outcomes 26 196 F(4, 191) = 1.973+ Teacher report (RC) 19 107 .343 8.440*** Student report 6 51 .337 5.251*** -.007 -.102 Parent report 4 10 .133 1.548 -.224 -2.635** Peer report 1 12 .278 5.745*** -.074 1.845+ Composite 1 16 .279 5.678*** .992 1.742+ Reliability outcomes (c) 13 60 .374 7.777*** 2.998 3.266** F(1, 58) = 10.667 *** Study characteristics Publication year (c) 26 196 .317 8.804*** -.007 -.863 F(1, 194) = .605 Impact factor (c) 23 165 .311 8.253*** .008 .195 F(1, 163) = .038 Study design 26 196 F(1, 194) = 8.577*** Cross-sectional 24 147 .339 9.430*** Longitudinal 6 49 .178 2.967** -.177 -2.929** Continent 26 196 F(4, 191) = .572 USA (RC) 15 91 .341 7.282*** Europe 8 95 .298 4.988*** -.048 -.601 Canada 1 2 .071 .338 -.283 -1.301 Australia 1 2 .193 .916 -.159 -.729 Other 1 6 .293 1.636 -.052 -.275 Sample characteristics
Proportion ethnic minority (c) 19 141 .351 8.149*** .001 .624 F(1, 139) = .389
Ethnicity 21 152 F(1, 149) = .960 Mixed sample (RC) 15 93 .347 7.601*** Majority sample 4 50 .339 4.188*** -.009 -.090 Minority sample 2 9 .160 1.170 -.201 -1.379 Proportion white (c) 18 136 .357 8.379*** -.002 -1.232 F(1, 136) = 1.517 SES 5 21 F(1, 19) = .670 Middle 1 2 .589 2.552* Low 4 19 .416 4.223*** -.233 -.819 Student characteristics Proportion male (c) 24 174 .348 9.906*** .004 1.505 F(1, 168) = 2.266+
23 Mean age (c) 21 174 .291 9.480*** -.036 -3.195*** F(1, 170) = 10.206*** Teacher characteristics Proportion male (c) 10 65 .351 6.293*** .003 .776 F(1, 63) = .602 Gender 11 89 F(1, 87) = .009 Mixed 10 83 .311 5.196*** Female 1 6 .293 1.511 -.020 -.095 Years of experience (c) 4 27 .264 3.165** -.003 -.117 F(1, 25) = .014 School characteristics School population 26 196 F(3, 192) = 3.811** Regular (RC) 21 162 .287 8.538*** Conduct problems 1 2 .071 .375 -.224 -1.162 Mixed 4 24 .489 6.533*** .240 2.701** Special education 1 8 .148 1.435 -.146 -1.448 School type 26 196 F(1, 194) = .324 Elementary 25 178 .314 9.302*** Secondary 2 18 .274 3.534*** -.044 -.569
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.
+
p < .10, * p <.05, ** p <.01,*** p < .001
4. Discussion
In order to examine the association between TSRs and students’ psychopathology, we conducted a multilevel meta-analysis. In total 26 studies were included, which consisted of 196 effect sizes. A moderate and significant association was found, with r = .311. This result confirms our hypothesis that high quality TSRs are related to less psychopathology in students. Dimensions of the TSR, reliability and type of TSR and psychopathology assessment, aspects of psychopathology, study design, mean age of the student, and school population were moderators of the relation between TSRs and psychopathology.
The more conflictual non-attuned teacher-student relationships were stronger related to psychopathology than other TSR dimensions, which is consistent with earlier research by Drugli (2013). There is empirical evidence showing that in particular conflictual TSRs increase externalizing behavior in students (Drugli, Klökner, & Larsson, 2011; Stipek & Miles, 2008), which might, more than internalizing behavior, negatively affect TSRs because negative self-reinforcing cycles of coercive interactions may easily develop (Drugli et al., 2011), while teachers may show relatively strong reactions to negative emotions of their
24 students (Hagenauer, Hascher, & Volet, 2015). This is in line with our study finding that TSRs were stronger related to externalizing than to internalizing behavior, although the larger effect size for externalizing behavior may also be explained by the more easily observable and public nature of externalizing behavior compared to internalizing behavior (Drugli et al., 2013). Notably, given that attachment has been shown to be much stronger associated with internalizing problems than with externalizing problems (Colonessi et al., 2011; Fearon, Bakermans-Kranenburg, van IJzendoorn, Lapsley, & Roisman, 2010; Hoeve et al., 2012), the larger effect size for externalizing problems of our meta-analysis may suggest that the TSR-psychopathology association is more a product of coercive teacher-student interactions than attachment related interactions between teachers and students.
Stronger associations were found for teacher-report of the TSR than for student report. Granot (2016) suggests that teachers have more professional (i.e., self-reflective and objective) and therefore less biased perceptions of relationships with their students, which may explain a stronger relation between the TSR and psychopathology. Thus, teacher-report of the TSR tends to be more valid and reliable than other assessment types. Teacher- and student-report of psychopathology showed the strongest association with TSRs, whereas parent report did not show a significant association. This may be explained by the context in which respondents experience the child. The association between TSR and students’ psychopathology 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’ psychopathology directly in the classroom, and may be affected by or affect students’ behavior at the same time. The students’ behavior is often situation specific, which could indicate that problem behavior observed in the school context may not correspond with problem behavior observed by parents in the home environment (Stams, Juffer, Rispens, &
25 Hoksbergen, 2000; Sointu, Savolainen, Lappalainen, & Epstein, 2012), and is also more influenced by teacher-child interaction than by parent-child interaction.
Finally, more reliable assessments of both psychopathology and TSRs yielded larger effect sizes. Cross-sectional studies yielded larger effects than did longitudinal studies, which shows that possible effects of the TSR on students’ psychopathology may level out over time. In addition, effect sizes decreased with student’s age. A possible explanation is that although relationships with adults remain important in adolescence, the influences of peer relationships gain a more central role, which could influence the nature and impact of the TSR (Blakemore & Mills, 2014). In addition, with age connectedness with peers becomes more important than connectedness with school, which may set limits to the effects that the TSR may have on psychopathology in adolescence (Loukas, Cance, & Batanova, 2013).
The school population showed a moderating effect on the association between TSRs and psychopathology. A mixed population showed a strong association, whereas a regular population showed a moderate association. An explanation for this difference may be that a mixed population shows more variance in both psychopathology and quality of TSRs among students. High quality TSRs are an important promotive factor for typically developing children (Baker, 2006; Skalickà, Stenseng, & Wichstrøm, 2015), but they may also function as a protective factor for children who are at risk for developing psychopathology (Murray & Murray, 2004).
There are several limitations of this study that need to be mentioned. First, there may be selection bias. Figure 1 shows that effect sizes are missing on the right side of the funnel plot, which indicates selection bias rather than publication bias. Certain populations may be underrepresented in studies, such as students with special needs, with higher levels of psychopathology. Second, there are more research data on externalizing behavior problems and the TSR than on internalizing problem behavior and the TSR (Drugli, 2013). Therefore
26 the generalization of the results on the association between the TSR and internalizing problem behavior is relatively limited.
Implications for clinical practice and further research concerning the TSR and students’ psychopathology can be derived from the results of this multilevel meta-analysis. First, it is clear that the TSR is a strong predictor of students’ psychopathology (Drugli, 2013; Lei et al., 2016). Interventions should focus on decreasing non-attuned TSRs, because this dimension of the TSR shows the strongest association with psychopathology. It is therefore of great importance to pay attention to restructuring these negative dimensions of the TSR in order to reduce psychopathology.
Second, the research on children with special needs is limited. This is remarkable, given that students with special needs are a vulnerable population in need of explicit and direct support from adults (Murray & Pianta, 2007). In contrast, research has shown that students with disabilities reported higher dissatisfaction of the TSR (Murray, & Greenberg, 2001). In addition, students with intellectual disabilities tend to report less positive relationships, characterized by more conflict and dependency, and less closeness (Eisenhower et al., 2007). This clarifies the need for supportive TSRs in special education in order to reduce problem behaviors.
Further research should focus more on reducing student’s psychopathology by using the TSR as a protective factor. High quality TSRs are associated with lower levels of psychopathology, whereas low quality TSRs, characterized by non-attuned interactions, are associated with higher levels of psychopathology. In addition, future research should examine the association between the TSR and internalizing problem behavior to fill the gap that is currently present. At last, research should focus on students’ needs of the TSR in special education. The TSR has shown its impact on psychopathology through this meta-analysis,
27 which also means that there is a lot to gain in special education TSRs. More important, the lack of research in this area is striking.
28
References
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
1
*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
*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
Blakemore, S.J., & Mills, K. L. (2014). Is adolescence a sensitive period for sociocultural processing? The Annual Review of Psychology, 65, 1–21. doi: 10.1146/annurev-psych-010213-115202
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
Bowlby, J. (1988). A secure base: Clinical applications of attachment theory. London: Routledge.
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
Colonessi, C., Draijer, E.M., Stams, G.J.J.M., Bruggen, C.O. van der, Bögers, S.M., & Noom, M.J. (2011). The relation between insecure attachment and child anxiety: a
29 meta-analytic review. Journal of Clinical Child & Adolescent Psychology, 40, 630-645. doi: 10.1080/15374416.2011.581623
Cooley, A. (2008). Legislating character: moral education in North Carolina’s public schools. Educational Studies, 43, 188-205. doi: 10.1080/00131940802117563
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
*Crum, M.S., Waschbusch, D.A., & Willoughby, M.T. (2016). Callous-unemotional traits, behavior disorders, and student-teacher relationship in elementary school students. Journal of Emotional and Behavioral Disorders, 24, 16-29. doi:
10.1077-1063426615569533
Davis, H. A. (2001). The quality and impact of relationships between elementary school children and teachers. Journal of Contemporary Educational Psychology, 26, 431– 453. doi: 10.1006/ceps.2000.1068
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
*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
30 *Drugli, M.B. (2013). How are closeness and conflict in student-teacher relationships
associated with demographic factors, school functioning and mental health in
Norwegian schoolchildren aged 6-13. Scandinavian Journal of Educational Research, 57, 217-225. doi: 10.1080/00313831.2012.656276
*Drugli, M.B., Klökner, C., & Larsson, B. (2011). Do demographic factors, school
functioning, and quality of student–teacher relationships as rated by teachers predict internalising and externalising problems among Norwegian schoolchildren?
Evaluation & Research in Education, 24, 243-254. doi: 10.1080/09500790.2011.626033
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.
*Eisenhower, A.S., Baker, B.L., & Blacher, J. (2007). Early student-teacher relationships of children with and without intellectual disability: contributions of behavioral, social, and self-regulatory competence. Journal of School Psychology, 45, 363-383. doi: 10.1016/j.jsp.2006.10.002
*Eisenhower, A.S., Blacher, J., & Bush, H.H. (2015). Longitudinal associations between externalizing problems and student–teacher relationship quality for young children with ASD. Research in Autism Spectrum Disorders, 9, 163-173. doi:
10.1016/j.rasd.2014.09.007
Fearon, R. P., Bakermans-Kranenburg, M. J., Van IJzendoorn, M. H., Lapsley, A., &
31 the development of children’s externalizing behavior: A meta-analytic study. Child Development, 81, 435–456. Doi: 10.1111/j.1467-8624.2009.01405.x.
*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 Hagenauer, G., Hascher, T., & Volet, S.E. (2015). Teacher emotions in the classroom:
associations with students’ engagement, classroom discipline and the interpersonal teacher-student relationship. European Journal of Psychology Education, 30, 385-403. doi: 10.1007/s10212-015-0250-0
Heaven, P.C.L., Leeson, P., & Ciarrochi, J. (2009). Personality development at school: assessing a reciprocal influence model of teachers’ evaluation and students’ personality. Journal of Research in Personality, 43, 815-821. doi:
10.1016/j.jrp.2009.06.009
*Henriccson, L., & Rydell, A.-M. (2004). Elementary school children with behavior problems: Teacher-child-interactions and self-perception. A prospective study. Merrill-Palmer Quaterly, 50, 111-138.
Hirschi, T. (1969). Causes of delinquency. Berkeley, CA: University of California Press.
Hoeve, M., Stams, G.J.J.M., Put, C.E. van der, Dubas, J.S., Laan, P.H. van der, & Gerris, J.R.M. (2012). Meta-analysis of attachment to parents and delinquency. Journal of Abnormal Child Psychology, 40, 771-785. doi: 10.1007/s10802-011-9608-1 Howes, C., Burchinal, M., Pianta, R., Bryant, D., Early, D., Clifford, R., & Barbarin, O.
(2008). Ready to learn? Children’s pre-academic achievement in pre-kindergarten programs. Early Childhood Research Quarterly, 23, 27-50. doi:
32 Hox, J. (2002). Multilevel analysis: Techniques and applications. Mahwah, NJ: Lawrence
Erlbaum Associates.
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 Educational Psychology, 85, 479-497. doi:10.1111/bjep.12094
*Leflot, G., Lier, P.A.C. van, Verschueren, K., Onghena, P., & Colpin, H. (2011). Transactional Associations Among Teacher Support, Peer Social Preference, and Child Externalizing Behavior: A Four-Wave Longitudinal Study. Journal of Clinical Child & Adolescent Psychology, 40:, 87-99, doi: 10.1080/15374416.2011.533409 Lei, H., Cui, Y., Chiu, M.M. (2016). Affective teacher-student relationship 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. Loukas, A., Cance, J. D., and Batanova, M. (2013). Trajectories of school
connectedness across the middle school years: examining the Roles of adolescents’ internalizing and externalizing problems. Youth and Society. 48, 557–576. doi: 10.1177/0044118X13504419
*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
*Madill, R.A., Gest, S.D., & Rodkin, P.C. (2014). Students’ perceptions of relatedness in the classroom the roles of emotionally supportive teacher-child interactions, children’s
33 aggressive disruptive behavior and peer social preference. School Psychology Review, 43, 86-105. doi:
Mager, U., & Nowak, P. (2012). Effects of student participation in decision making at school. A systematic review and synthesis of empirical research. Educational Research Review, 7, 38-61. doi: 10.1016/j.edrev.2011.11.001
*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. (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:
10.1002/1502-6807(200101)38:1<25::AID-PITS4>3.0.CO;2-C
*Murray, C., & Murray, K. (2004). Child level correlates of teacher-student relationships: An examination of demographic characteristics, academic orientations, and behavioral orientations. Psychology in Schools, 41, 751–762. doi: 10.1002/pits.20015 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
*Murray, C., & Zvoch, K. (2011). Teacher-student relationships among behaviorally at-risk African American youth from low-income backgrounds: student perceptions, teacher perceptions, and socioemotional adjustment correlates. Journal of Emotional and Behavioral Disorders, 19, 41-54. doi: 10.1177/1063426609353607
Nakagawa, S., & Santos, E. S. A. (2012). Methodological issues and advances in biological meta-analysis. Evolutionary Ecology, 26(5), 1253–1274. doi: 10.1007/s10682-012 9555-5
34 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. Journal of Youth and Adolescence, 46, 1661-1687. doi: 10.1007/s10964-016-0534-y
Pianta, R.C. (1997). Introduction: Developmental perspectives on school outcomes for risk and non-risk populations. Journal of School Psychology, 35, 1-2. doi: 10.1016/S0022-4405(96)00035-0
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. (2003). Relationships between teachers and children. In W. M. Reynolds, G. E. Miller, & I. B. Weiner (Eds.), Handbook of psychology: Educational psychology (Vol. 7. no. 1 99-234). Hoboken, NJ: John Wiley.
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. doi: 10.1016/S0885-2006(97)90003-X
*Pianta, R.C., & Stulhman, 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 emotional and behavioural difficulties. International Journal of Educational Psychology, 4,
84-35 108. 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/).
Roelofs, M. (2017). Teacher-student relationship and students’ social-emotional outcomes: a systematic review and meta-analysis (master thesis).
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
Rudasill, K.M., Reio Jr., T.G., Stipanovic, N., Taylor, J.E. (2010). A longitudinal study of student–teacher relationship quality, difficult temperament, and risky behavior from childhood to early adolescence. Journal of School Psychology, 48, 389-412. doi: 10.106/j.jsp.2010.05.001
*Runions, K. C., Vitaro, F., Cross, D., Shaw, T., Hall, M., and Boivin, M. (2014). Teacher– child relationship, parenting, and growth in likelihood and severity of physical aggression in the early school years. Merrill Palmer Quarterly, 60, 274–301. doi: 10.13110/merrpalmquar1982.60.3.0274
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. doi: 10,1037/0003-066X.55.1.68
*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. Silver, R. B., Measelle, J.R., Armstrong, J.M., & Essex, M.J. (2005). Trajectories of
36 classroom externalizing behavior: Contributions of child interactions, family
characteristics, and the teacher-child relationship during the school transition. Journal of School Psychology, 43, 39-60. doi: 10.1016/j.jsp.2004.11.003
*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
Sointu, E.T., Savolainen, H., Lappalainen, K., & Epstein, M.H. (2012). Parent, teacher and student cross informant agreement of behavioral and emotional strengths: students with and without special education support. Journal of Chile and Family Studies, 21, 682-690. doi: 10.1007/s10826-011-9250-x
Spilt, J.L., Koomen, H.M.Y., Thijs, J.T. (2011). Teacher wellbeing: The importance of teacher-student relationships. Educational Psychology, 23, 457-477. doi: 10.1007/s10648-011-9170-y
Spruit, A., Schalkwijk, F., Vugt, E. van, Stams, G.J.J.M. (2016). The relation between self-conscious emotions and delinquency: a meta-analysis. Aggression and Violent Behavior, 28, 12-20. doi: 10.1016/j.avb.2016.03.009
Stams, G.J.J.M., Juffer, F., Rispens, J., & Hoksbergen, R.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
*Stewart, T., & Suldo, S. (2011). Relationships between social support sources and early adolescents’ mental health: the moderating effect of student achievement level. Psychology in the Schools, 48, 1016-1033. doi: 10.1002/pits.20607
*Stipek, D., and Miles, S. (2008). Effects of aggression on achievement: does conflict with the teacher make it worse? Child Dev. 79, 1721–1735. doi:
10.1111/j.1467-37 8624.2008.01221.x
Tabachnik, B. G., & Fidell, L. S. (2013). Using multivariate statistics (6th ed.). Boston: Allyn and Bacon.
*Troop-Gordon, W., and Kopp, J. (2011). Teacher–child relationship quality and children’s peer victimization and aggressive behavior in late childhood. Social Development, 20, 536–561. doi: 10.1111/j.1467-9507.2011.00604.x
Van den Noortgate, W., López-López, J. A., Marin-Martinez, F., & Sánchez-Meca, J. (2013). Three-level meta-analysis of dependent effect sizes. Behavior Research Methods, 45(2), 576–594. doi: 10.3758/s13428-012-0261-6
Van den Noortgate, W., López-López, J. A., Marin-Martinez, F., & Sánchez-Meca, J. (2014). Meta-analysis of multiple outcomes: A multilevel approach. Behavior Research Methods. doi: 10.3758/s13428-014-0527-2 (Advance online publication).
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
Wibbelink, C. J. M., & Assink, M. (2015). Handleiding voor het uitvoeren van een drie-level meta-analyse in R [Manual for conducting a three-level meta-analysis in R].
Amsterdam: Universiteit van Amsterdam.
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
38 Role of social contexts in middle grades school climate. American Journal of Community Psychology, 54, 28-45. doi: 10.1007/s10464-014-96
