Math Self-Efficacy and STEM Intentions: A Person-Centered Approach

Research pertaining to STEM interest and persistence continues to be a top priority in the educational research arena. The current study employed a person-centered approach to examine the impact of math self-efficacy and various distal predictors, such as individuals' demographic information, beliefs about math, and social group identification, on STEM interest and intentions. Specifically, we conducted a latent profile analysis (LPA), thereby inferring three homogeneous subgroups of individuals or latent classes from their response patterns on the 18-item sources of math self-efficacy measure. Our analyses showed that individuals' ethnicity, implicit theories of math ability, and other group orientation were predictive of class membership (Mastery, Moderate, and Unconfident). We also found that there were significant differences in interest in STEM subjects, interest in STEM activities, individuals' majors, and retention grade point average across the three latent classes. Our findings support the importance of math self-efficacy in choice of major as well as overall academic performance regardless of whether a student is in a STEM field or a non-STEM field.