Project Grant 2225209

This National Science Foundation project grant award of $330,423 supports research evaluating letters of recommendation for racial and gender bias in engineering doctoral program admissions. Funded under the federal STEM Education program (CFDA 47.076), the three-year award to the University of Michigan aims to identify differences in how applicants' potential for engineering success is described based on gender and race/ethnicity. Using qualitative methods, content analysis and natural language processing, the project will examine language used in letters of recommendation to identify biases that may perpetuate enrollment disparities. Goals include developing a deeper understanding of disciplinary practices and building skills in text-based data analysis. The research founded on role congruity and stereotype content theories intends to produce an engineering graduate admissions model free of race and gender biases applicable to other disciplines.