Children's social emotional competence in Pakistan and Sweden: Factor structure and measurement invariance of the Social Competence Scale (teacher edition).

Abstract

Social emotional competence is fundamental to the positive development of children and youth. Accurately understanding and assessing children's social emotional competencies, using psychometrically sound instruments, are essential to global efforts to support children's social emotional learning, academic achievements, and health. This study examined the psychometric properties of a teacher-reported measure of young children's social emotional competence, the Social Competence Scale - Teacher edition (SCS-T), in two samples of children growing up with varied economic resources/conditions, cultural norms, and educational experiences, namely Pakistan (N = 396) and Sweden (N = 309). Participants were aged 4-6 years old. The study design was cross-sectional. Using structural equation modelling, bi-factor confirmatory factor analysis models implying shared variance, among all items and domain-specific shared variance, among the prosocial items, emotion regulation items, and academic skills items resulted in good fitting models in each respective sample. Invariance testing across samples revealed a subset of items from each factor structure with partial scalar invariance, whereby five items had equal thresholds and could be comparable across the two samples. Thus, results provided partial support for hypotheses 1, 2, and 3, in that the posited three factor model (H1) was not clearly supported and a bi-factor model evidenced the best fit, among tested models, for both samples. Further, partial scalar invariance (H3) was found for five items out of 25 items, concerning social competence and academic skills. In regards, to the posited research question, the results of Z-tests showed significant (p < 0.001) latent mean differences between the samples. Compared to the Swedish sample, the Pakistani sample was 1.80 units lower on social competence (z = -6.41, p < 0.001) and 1.86 units lower on academic skills (z = -7.87, p < 0.001). The implications of these findings in light of efforts to promote positive child development in diverse parts of the world are considered.