STEM identity and STEM career intention: a meta-analysis

STEM identity has gained increasing attention for its impact on shaping students’ intention to pursue careers in STEM fields. However, its role remains unclear due to inconsistent findings across various studies. This meta-analysis aimed to explore the strength and potential moderators of their relationship by synthesizing evidence from different studies. We estimated the average weighted correlation coefficient (r) and explored six potential moderators, including several study characteristics and demographic characteristics, to explain the heterogeneity among studies. Utilizing a random-effects model, we analyzed 75 effect sizes from 35 quantitative studies and revealed a moderate and statistically significant average weighted correlation of r = .47 (SE = .04, p < .001) between STEM identity and STEM career intention. Through moderator analysis, we found the magnitude of this correlation varies significantly with identity type (e.g., science identity, physics identity, and math identity), gender, and different interactions between identity type and demographic characteristics. These findings suggest that while STEM identity is a meaningful predictor of students’ career intention, its influence is more moderate than often assumed. Importantly, the relationship is not uniform; it varies across subgroups and contexts. This nuanced understanding highlights the need for more targeted and inclusive identity-based interventions in STEM education and offers valuable insights for educators, researchers, and policymakers in fostering sustained student engagement in STEM pathways.