Peer Effects and Gender in the College Classroom

Kristin Jones, Hartwick College
Carlena Cochi Ficano, Hartwick College

The presence of peer effects in higher education has been investigated in an extensive empirical literature. Studies have addressed the impacts of peer academic ability on own academic performance (Sacerdote 2001, Zimmerman 2003, Stinebrickner and Stinebrickner 2005, Lyle 2007, Carrell et. al. 2008); peer characteristics such as family income on own academic performance (Stinebrickner and Stinebrickner 2005); and peer characteristics on decisions such as joining a Greek organization or athletic team (Sacerdote 2001). This paper estimates the effects of overall and gender-differentiated peer academic rating (where peer groups are defined as roommates, floormates, and classmates) on own GPA in the first semester of college. We find evidence of gender-selective peer effects. Male students perform significantly better when their male classmates have higher average academic performance but do not respond to female peer academic performance, while females respond to neither own-gender or general peer performance