Peer attachment and class emotional intelligence as predictors of adolescents' psychological well-being: A multilevel approach.

The aim of the study was to analyze whether gender, age, peer attachment, and class-level emotional intelligence could predict adolescents' psychological well-being by applying a multilevel approach. The sample comprised 2182 secondary school students from the Basque Country (northern Spain) (from 118 classrooms, 51.6% girls), aged between 12 and 18 years. A two-level model (with students nested into classes) was used to analyze the influence of three level-one covariates (gender, age, and peer attachment) and one level-two covariate (class-level emotional intelligence) on the positive affect component of psychological well-being. The results showed an overall decrease in well-being as adolescents grow older, and an increase linked to a higher peer attachment. Furthermore, class-level emotional intelligence showed a positive relationship with students' well-being. This group-level covariate also strengthened the effect of peer attachment on the well-being. The advantages of using a multilevel approach for predicting mental health and psychological adjustment are discussed.