| Prior studies have established that the academic performance of college students varies with both their individual academic ability and with peer ability, or the mean ability of other students who attend the same postsecondary institution. However, there is no consensus on whether academic match--- defined here as the difference between student ability and peer ability--- has an effect on a student's college outcomes. In this dissertation I ask: on average and other things equal, does the statistical effect of student SAT score on college outcomes vary with the mean SAT score of students at the same college? The college outcomes I consider in this dissertation are: bachelor's degree completion, months to bachelor's degree completion, cumulative college grade point average, Science, Technology, Engineering, and Mathematics (STEM) major selection, and decline in educational aspirations. I use a sample of students from the Beginning Postsecondary Students Longitudinal Study 2004-2009 (BPS) and institutional data from the Integrated Postsecondary Education Data System (IPEDS) to determine if and how the effect of student SAT on the different college outcomes considered varies across institutions with different mean SAT scores. The analyses presented in this dissertation provide an insight into why prior research on the effect of academic match on educational outcomes in postsecondary education has yielded conflicting results. The effects of academic match on college outcomes vary so that researchers focusing solely on institutions where the mean student SAT score is high, or solely on institutions where the mean student SAT score is low, will yield paradoxical results. |
