Program evaluation for interventions aimed at enhancing diversity can fall short when the evaluation unintentionally reifies the exclusion of multiple marginalized student experiences. The present study presents a Quantitative Critical Race Theory (QuantCrit) approach to program evaluation to understand outcomes for Women of Color undergraduates involved in a national biomedical training program called the Building Infrastructure Leading to Diversity (BUILD) initiative. Using longitudinal data, we examined the impact of participation in the BUILD Scholars programs and BUILD-developed novel biomedical curriculum on undergraduate's research self-efficacy. Employing propensity score matching and multiple regression models, we found that Black women who participated in the BUILD scholars program reported higher research self-efficacy, whereas Latine and White undergraduate BUILD scholars had lower research self-efficacy. Additionally, Latine women who participated in novel biomedical curricula reported significantly lower research self-efficacy. We contend that disaggregated and intersectional analyses of subpopulations are necessary for improving understanding of program interventions and identifying areas where systems of exclusion may continue to harm students from minoritized backgrounds. We provide recommendations for future quantitative program evaluation practices and research in science, technology, engineering, mathematics, and medicine (STEMM) equity efforts.