Perceived abilities or academic interests? Longitudinal high school science and mathematics effects on postsecondary STEM outcomes by gender and race

Abstract

Abstract Purpose of the study Previous literature has examined the relationship between high school students’ postsecondary STEM major choices and their prior interest and perceived ability in mathematics. Yet, we have limited understanding of whether and how perceived ability and interest in science and mathematics jointly affect students’ STEM major choices. Results Using the most recent nationally representative longitudinal cohort of U.S. secondary school students, we examine the degree to which students’ perceived mathematical and scientific abilities and interests predict their STEM major choices, employing logistic regression and a series of interaction analyses. We find that while both mathematics and science perceived ability positively influence STEM major selection, academic interest in these subjects is a weaker predictor. Moreover, across a series of analyses, we observe a significant gender gap—whereby women are less than half as likely to select STEM majors—as well as nuanced distinctions by self-identified race. The relationships among perceived ability, interest, and STEM major choice are not found to meaningfully vary by race nor consistently by gender. However, perceived ability has a more positive effect for men than women who are pursuing Computing/Engineering majors and a more positive effect for women than men who are pursuing other STEM majors, including less applied Social/Behavioral, Natural, and Other Sciences. Implications These findings suggest potential opportunities to enhance their perceived mathematical and scientific abilities in high school, positioning them to potentially enter STEM fields. School sites with more resources to support the ambitions of STEM students of all backgrounds may be better positioned to reduce postsecondary disparities in STEM fields. Given existing opportunity gaps and resource differentials among schools, corresponding recommendations are suggested.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 42 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.