The Roles of Peers, Teachers, and Families in Shaping Adolescents’ Social and Emotional Skills: Evidence from the OECD SSES in China

• In Event: Assessing Learning Worldwide: Cross-Cultural Evidence, Discourse, and Policy Implications

Proposal

Introduction
Adolescents’ social and emotional skills are shaped within their daily interactions across multiple ecological systems, including peers, teachers, and families (Bronfenbrenner, 1994; Gergen, 2009). Peer relationships are central to emotional regulation and social adaptation during adolescence. Supportive peer interactions in early adolescence significantly contribute to psychological development and social-emotional regulation (Mitic et al., 2021). Moreover, peer and teacher-student relationships jointly influence children’s social-emotional adjustment, highlighting the complementary nature of these social contexts (Wang et al., 2016; Dunsmore et al., 2008).
Teacher-student relationships also play a crucial role in students’ social and emotional development. Positive teacher-student interactions not only foster academic achievement (Lei et al., 2023; Göktaş & Kaya, 2023; Roorda et al., 2009) but are also strongly associated with enhanced social and emotional competencies (Dietrich et al., 2021; Durlak et al., 2011; Skinner et al., 2008). Longitudinal findings further demonstrate that high-quality teacher-student relationships predict stronger future self-regulation and social competence (Hajovsky et al., 2023).
Family, as the innermost microsystem in Bronfenbrenner’s ecological framework, exerts a profound influence on adolescents’ emotional and social development. Parent-child relationships in particular are critical, with research showing moderate associations between parent-child closeness, emotion regulation, and social competence in adolescents (Engels et al., 2001). Positive parent-child interactions, such as emotional support and intimacy, significantly enhance children’s social and emotional skills (Haven et al., 2014; Boyum & Parke, 1995).
Drawing upon ecological systems theory, this study emphasizes that social and emotional skills develop within the nested systems of peer, teacher-student, and family relationships (Bronfenbrenner, 1994; Soto et al., 2022). While prior studies have examined environmental influences, few have integrated these three relational contexts into a single analytical framework. Moreover, most research has relied on regression-based analyses, overlooking advanced methods suited to nested data structures. To address these gaps, this study uses multilevel modeling with data from the OECD SSES in Jinan, China, to analyze the contributions of peer, teacher-student, and family relationships to adolescents’ social and emotional skills.

Data and Methods
Sample
Data were drawn from the OECD-SSES, a large-scale international survey designed to assess the conditions and practices shaping social and emotional development among students aged 10–15. The present study uses the Jinan sample of 15-year-old students (N = 3,699; 48% female, mean age = 15.70, SD = 0.28), obtained through stratified two-stage cluster sampling.

Measures
(1)Social and Emotional Skills. SES were assessed across five broad domains—task performance, emotional regulation, collaboration, openness, and engaging with others—operationalized through 16 subskills (e.g., responsibility, empathy, optimism). A total of 144 items were rated on a 5-point scale. Scores were calibrated via Item Response Theory and Weighted Likelihood Estimation, transformed to a standardized scale (M = 500, SD = 100).
(2)Peer relationships. Peer connectedness was measured with six items (α = 0.723) on a 4-point scale (e.g., “My classmates respect me”).
(3)Teacher–student relationships.Perceptions of teacher respect and closeness were measured with six items (α = 0.733), including one reverse-coded item.
(4)Family relationships. Family quality was operationalized through a single item on satisfaction with parent/guardian relations, rated on a 10-point scale. Parallel items captured satisfaction with peers and teachers, used in supplementary analyses.

Analysis
Hierarchical linear modeling (HLM) was employed to account for the nested structure of the data (students within classes).

Results
Gender differences were evident across several domains: girls outperformed boys in achievement motivation, empathy, and tolerance, consistent with meta-analytic evidence on gender disparities in non-cognitive skills (Feingold, 1994; DiPrete & Buchmann, 2013). Family socioeconomic status positively predicted most subskills, particularly achievement motivation and responsibility, but showed a small negative association with trust, echoing prior findings on stratification and emotional inequalities (Liu, 2018).
Peer relationships emerged as the strongest predictor, significantly and positively influencing all 16 subskills, with the largest effects on trust (b = 0.428), cooperation (b = 0.375), and optimism (b = 0.374). This finding aligns with evidence that peer acceptance fosters prosocial and emotional outcomes (Zárate, 2023; Pinheiro Mota & Matos, 2013).
Teacher–student relationships predicted 15 of 16 subskills (excluding sociability), with the strongest associations for achievement motivation (b = 0.320), perseverance (b = 0.303), and curiosity (b = 0.301), consistent with global evidence linking teacher support to socio-emotional competencies (Roorda et al., 2011; Jennings & Greenberg, 2009).
Family relationship satisfaction significantly predicted eight subskills, including responsibility, perseverance, self-control, vitality, and optimism (b = 0.143). However, its overall predictive power was weaker than that of peers and teachers. Interestingly, family satisfaction negatively predicted empathy (b = –0.058) and sociability (b = –0.042), possibly reflecting adolescents’ growing desire for independence and the greater role of peers in empathy development (Hoffman, 2008; Boele et al., 2019).
Finally, significant between-classroom variation was observed in several subskills, particularly self-control, underscoring the influence of contextual factors beyond individual characteristics.

Discussion
This study highlights the differentiated roles of peers, teachers, and families in shaping adolescents’ social and emotional skills. Teacher–student relationships emerged as the strongest predictor, particularly for task performance and openness, reinforcing prior evidence that supportive teachers foster motivation, perseverance, and openness to diversity (Hallinan, 2008; Ulug et al., 2011). Peer relationships most strongly influenced emotional regulation and collaboration, suggesting that peer interactions provide unique opportunities for practicing empathy, optimism, and cooperation (Rubin et al., 2008; Tetzner & Becker, 2019). Family relationships, though less predictive at age 15, remained crucial for emotional control and perseverance, consistent with theories of emotional socialization in the home (Morris et al., 2007; Ford & Gross, 2019).
The findings underscore that social and emotional skills cannot be understood in isolation from relational contexts. Schools should therefore foster both supportive teacher–student relationships and peer collaboration opportunities, while families should be encouraged to maintain emotionally supportive parenting even as adolescents seek greater autonomy.

Authors

• JIA LI, East China Normal University

• Zhenfei Peng, Faculty of Education

• Xinmin Zhao