Assessing the Importance of Sample Choice and Selectivity for Sex Segregation in College Majors: A Replication of Ochsenfeld (2016)

Abstract Ochsenfeld (2016) has found that a substantial part of sex segregation in higher education results from differences in vocational interests (i.e., preferences), while constraints (e.g., relative math grades) play only a minor role. We challenge the validity of these findings because earlier work employed a cross-sectional student sample and might therefore suffer from endogenous selection (i.e., post hoc rationalizations due to simultaneous reporting of majors and preferences) and postoutcome collider bias (i.e., conditioning on the outcome). Our replication study uses panel data (National Educational Panel Study, NEPS-SC4) that allow adjustment for the two sources of bias through the application of a pretransition preference measure and inverse probability weighting. Our analyses demonstrate the validity of prior research. Furthermore, our analysis indicates that the explanatory power of the overall model and the role of constraints for sex segregation in majors vary across the propensity of sample inclusion, thereby demonstrating the importance of sample composition for testing sociological theories.