Relationships and what is taking place socio-emotionally in the classroom may support or hinder students´ learning and development. All students benefit from a positive, supportive classroom climate, especially children with special educational needs. Improving the quality of the learning environment does not mean disregarding academic achievement. Because research on how to improve the classroom climate is limited, this thesis aimed to develop an intervention with the potential to influence the social climate and benefit student outcomes. Furthermore, the three studies in this thesis have connected aims. The first and the second study provided insight into constructs at the student level. The first study examined the psychometric properties of an instrument used to measure students’ prosocial behavior. The second study examined the associations between students´ self-concept, prosocial skills, well-being in school, and academic achievement. Gender differences were also investigated. The third study tested the effects of an intervention involving specific activities (e.g., self-assessment, observation, and coaching). The three studies were empirical investigations of a sample of 143 students in elementary schools in a Swedish metropolitan area. The data sources were students´ self-reports and tests, including teachers´ reports on students’ prosocial skills, teachers’ social climate assessments, and video-recorded classroom climate observations. Study I and II had a cross-sectional design, and study III had an experimental design with cluster randomization at the school level. There were four intervention classes and four control classes. Data were primarily analyzed with structural equation modeling (SEM) techniques in Mplus to examine the hypotheses and research questions.

Study I examined students’ prosocial behavior using confirmatory factor analysis (CFA). A two-factor measurement model was clearly supported, but a single-factor and a three-factor model cannot be excluded as possibilities for future research in the Swedish school context. Study II examined the constructs at baseline with CFA, demonstrating significant associations between self-concept and prosocial behavior, indicators of social-emotional learning (SEL), and well-being. The findings support the association between SEL and academic achievement indicators, confirming previous research. In study III, pre- to post-test changes resulting from a coaching intervention were examined with an autoregressive model. The coaching intervention was considered feasible, but there were no intervention effects from the pre- to post-test on the observed variables: self-concept, prosocial behavior, well-being, academic achievement, or classroom climate.

Overall, this thesis contributes to the research on the whole child approach. Self-concept and prosocial behavior, indicators of SEL and well-being, contribute to understanding academic achievement. Teachers can use these assessment instruments to understand children´s social-emotional and academic development levels and the correlations between them so that appropriate support can be provided.