The Association Between Female Students’ Computer Science Education and STEM Major Selection: Multilevel Structural Equation Modeling

Abstract

Concerning the underrepresentation of female students in computer science (CS) classes at the K-12 level and math-intensive STEM fields, this study investigated the relationship between female students earning less credits in CS courses during high school and their STEM major choices. Data were drawn from a nationally representative sample of U.S. high school graduates from the restricted data of the Educational Longitudinal Study of 2002 which were collected by the National Center for Educational Statistics. The multilevel structural equation modeling revealed that (a) female students’ lower frequency of taking CS courses was associated significantly with their underrepresentation in STEM fields, (b) earning less credits in CS courses was a stronger factor for female students’ underrepresentation in STEM fields compared to the well-documented predictors for STEM major choice—math self-efficacy and math scores on the ACT, and (c) no significant difference in ACT math scores emerged between male and female students, which challenges the dominant view that male students’ greater math aptitude causes the persistent gender gap that leads to women’s underrepresentation in STEM fields. The study suggests that integrating CS in traditional classrooms as well as promoting a rigorous CS curriculum at the secondary level would be an effective strategy to increase female participation in math-intensive STEM fields.