Identification of academic peer effects in college: Does data aggregation matter?

<span lang="EN-US">This study exploited random roommate assignments in a small Chinese college to estimate the causal effects of roommates’ scores on national College Entrance Test (CET) first-year students’ Grade Point Average (GPA). Analyzing data an entire cohort enrolled students, we found that level aggregation, for both peer-ability measure and one’s own academic-outcome measure, matters identification academic peer effects. Specifically, while average CET score has barely significant impact, highest-scoring roommate’s strong positive impact. Peer are also larger GPA required courses than elective courses. Finally, types decline over time increase time, suggesting students this tend substitute their ability as they become more academically independent.</span>