Examining the power of performance : an investigation into STEM persistence across field of study and gender
MetadataShow full item record
Are students in the physical sciences/engineering (PS/E) similar, in terms of academic performance and degree persistence, to students majoring in other fields? Previous research tends to focus exclusively on students within STEM; conversely, this dissertation uses a nationally representative sample to examine persistence rates among students whose initial major is in PS/E, life science, business, social science, education, health, or humanities. Students in PS/E majors are sometimes more and sometimes less likely than students in other fields to earn a degree in their initial field of study versus another field. Additionally, students in PS/E are the most academically prepared for college, as measured by high school mathematics and English performance. Despite these high levels of preparation, PS/E students earn lower college grades within their intended field of study than students in other fields. The results show that the gap in college grades, net of background and preparation, explains (in part) why students in business, education, and humanities have higher persistence rates than students in PS/E. Moreover, the association between within-field college grades and persistence is strongest for PS/E students. Taken together, these results demonstrate that students who enter PS/E are indeed unique in terms of academic performance, persistence, and the relationship between the two. Further, among PS/E students, females are more likely than males to earn a degree in fields outside of PS/E compared to within PS/E. Using college transcript data, I investigate the commonly used argument that gender gaps in PS/E persistence can be explained by female underachievement during the college years. Regardless of whether performance is measured using students' PS/E GPA, the proportion of low grades earned, or the difference between PS/E and non-PS/E GPA, there is no evidence that differences in performance explain the gender gap in persistence. This result is not surprising given there is no significant gender gap in PS/E grades. Lastly, I find that the relationship between PS/E GPA and persistence is similar across gender. These results build on the growing body of literature suggesting that gender differences in academic performance are ineffective at explaining gender inequality in PS/E.