The Consequences of Friendships: Evidence on the Effect of Social Relationships in School on Academic Achievement
Jason M. Fletcher, Stephen L. Ross, Zhang, Yuxiu
Abstract
Open-access reader
Jason M. Fletcher, Stephen L. Ross, Zhang, Yuxiu
Abstract
Open-access reader
This paper examines the impact of youth friendship links on student's own academic performance (grade point average) using the Add Health. We estimate a reduced form, high dimensional fixed effects model of within cohort or grade friendship links and use this model to predict each student's number of friends whose mothers completed four years of college. The effects of friendship links are identified using across-cohort, within school variation in demographic composition of the student's cohort or grade. We find that increases in the number of friendship links with students whose mothers are college educated raises grade point average among girls, but not among boys. Additional analyses suggest a positive view of the school environment and a perception of one's self as functioning well in that environment as possible mechanisms. The effects are relatively broad based across students over maternal education, racial and ethnic composition, and across schools that vary in demographic composition.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This paper examines the impact of youth friendship links on student's own academic performance (grade point average) using the Add Health. We estimate a reduced form, high dimensional fixed effects model of within cohort or grade friendship links and use this model to predict each student's number of friends whose mothers completed four years of college. The effects of friendship links are identified using across-cohort, within school variation in demographic composition of the student's cohort or grade. We find that increases in the number of friendship links with students whose mothers are college educated raises grade point average among girls, but not among boys. Additional analyses suggest a positive view of the school environment and a perception of one's self as functioning well in that environment as possible mechanisms. The effects are relatively broad based across students over maternal education, racial and ethnic composition, and across schools that vary in demographic composition.
Key concepts: Friendship, Instrumental variable, Socioeconomic status, Peer effects, Psychology, Developmental psychology, Homophily, Matching (statistics)