Self-efficacy and well-being in adolescents: A comparative study using variable and person-centered analyses

Introduction
While there is already evidence positively linking higher levels of self-efficacy with well-being in adolescence, analyses have been of a bivariate or correlational nature. Previous studies have used both general self-efficacy, and a domain-specific approach. The present study operationalises three domains of self-efficacy (academic, social, and emotional), in a domain specific way, and additionally, examines how clusters, derived from scores on these domains relate to well-being.

Methods
Data were gathered on academic, social, and emotional self-efficacy, mental well-being, psychological and somatic symptomatology, frequency of physical exercise, and parental security. Participants were 3,485 adolescents (M_age = 17.0 [SD = 1.1] in their final year of High school in Northern Ireland.

Results
Results of bivariate analyses pointed to emotional self-efficacy as a key variable in adolescent well-being. In the cluster analyses, five clusters were identified, and in a series of regression analyses, membership of the High self-efficacy cluster (characterized by high scores on all three self-efficacy domains), was found to be optimal.

Conclusions
The application of cluster analysis revealed qualitative differences across a set of five categorical profiles. For that reason, interrelations among Self-Efficacy Questionnaire for Children scores at the group level should no longer be used to make inferences about how these variables function within individuals.