Opposite-gender friendships and learning performance of students: Evidence from China

Abstract

Based on data from China Education Panel Survey, this paper uses the fixed-effect model and PSM-DID method to examine the causal effect of opposite-gender friendships on the learning performance of Chinese students. We find that opposite-gender friendships significantly reduce girls' test scores, but have no impact on boys. The change in learning habits caused by the quality of social networks is the possible mechanism that drives the effect. The paper also suggests that parental regulation increases the probability that children associate with friends of the opposite gender, which leads to lower test scores. In contrast, class rearrangement can be effective in improving girls' learning performance by reducing their opposite-gender friendships. The evidence in this paper sheds new light on the debate over whether and how to intervene in students' social interactions.