Abstract
This research examined the magnitude of personality differences across different colleges and universities to understand (a) how much students at different colleges vary from one another and (b) whether there are site-level variables that can explain observed differences. Nearly 8,600 students at 30 colleges and universities completed a Big Five personality trait measure. Site-level information was obtained from the Integrated Postsecondary Education System database (U.S. Department of Education). Multilevel models revealed that each of the Big Five traits showed significant between-site variability, even after accounting for individual-level demographic differences. Some site-level variables (e.g., enrollment size, requiring letters of recommendation) explained between-site differences in traits, but many tests were not statistically significant. Student samples at different universities differed in terms of average levels of Big Five personality domains. This raises the possibility that personality differences may explain differences in research results obtained when studying students at different colleges and universities. Furthermore, results suggest that research that compares findings for only a few sites (e.g., much cross-cultural research) runs the risk of overgeneralizing differences between specific samples to broader group differences. These results underscore the value of multisite collaborative research efforts to enhance psychological research.