Gender Differences in Academic Efficacy across STEM Fields

Cultural processes can reduce self-selection into math and science fields, but it remains unclear how confidence in computer science develops, where women are currently the least represented in STEM (science, technology, engineering, and mathematics). Few studies evaluate both computer skills and self-assessments of skill. In this paper, we evaluate gender differences in efficacy across three STEM fields using a data set of middle schoolers, a particularly consequential period for academic pathways. Even though girls and boys do not significantly differ in terms of math grades and have similar levels of computer skill, the gender gap in computer efficacy is twice as large as the gap for math. We offer support for disaggregation of STEM fields, so the unique meaning making around computing can be addressed.