Marion, Donna

This study of 91 dyads investigated concurrent and prospective bidirectional
associations between friendship quality and psychosocial adjustment in young
adolescents, using multiple-group structural equation modeling to detect gender
differences. Friend reports ofboth positive (social support) and negative features of
friendship, self-reports of two adjustment variables (internalizing problems and
externalizing problems), and self-reports of three dimensions of self-esteem (global selfworth,
behavioral conduct esteem, and close friendship esteem) were examined at two
time periods approximately one year apart. Principal findings were that grade 6
friendship negativity was inversely associated with grade 7 self-esteem, and positively
associated with grade 7 internalizing problems and externalizing problems, in the boys'
model but not girls' model. Chi-square difference testing confirmed a significant
difference between these paths in the girls' model and these paths in the boys' model.

This study investigated friend influence on school engagement in a sample of 160 stable same-sex friendship dyads (94 female dyads and 66 male dyads) from five senior high schools and four vocational schools in a small city in central Finland. Longitudinal data were collected during the first and second years of upper secondary school, approximately one year apart, and self-reports were available from both members of each friendship dyad. The framework of the Actor-Partner Interdependence Model (APIM; Kenny, Kashy & Cook, 2006) was used to estimate friend influence on school engagement in a model that did not distinguish same-sex friends, in a direct-effects model that distinguished friends based on relative levels of school burnout, and in a multiple-group model for distinguishable friends that investigated perceptions of maternal affection as a moderator of friend influence. Results suggest that the higher burnout partner in a friendship dyad influenced a decline in the lower burnout partner's school engagement only when the lower burnout partner perceived low maternal affection. When the lower burnout partner perceived high maternal affection, there was no evidence of negative influence by a higher burnout partner. Patterns of influence did not vary as a function of sex or school track. The importance of distinguishing friends on a theoretically and statistically meaningful basis to learn who influences whom, and of investigating indirect effects models when studying friend influence is also discussed.