2025 ASEE Annual Conference & Exposition

Purpose: In this work-in-progress empirical research paper, we describe results from a study that considers how the format and content of letters of recommendation (LOR) might differ according to applicants’ gender and ethno-racial background. We are conducting this study because there is little research that examines the role of subjective application materials such as LOR in relation to engineering admissions outcomes. To date, most research about equity in admissions has focused on the role of standardized tests, yet the use of GRE scores is declining in graduate admissions. This is a gap in our understanding of the factors that influence diversity, equity, and inclusion in engineering education.

Methods: We conducted interviews with engineering faculty in four large programs at a public university to learn how they use information from LOR to make PhD program admissions decisions. The results suggest that faculty perceive LORs that contain detailed information and LORs from research supervisors to be most credible as sources of information about applicants' potential for success. Next, we conducted a preliminary analysis to see if the value faculty place on these types of letters is reflected in admissions outcomes. We used LOR from 280 applications to four PhD programs to examine the relationship between whether an applicant is admitted to the program and two variables that are related to how “credible” the LOR is: 1) the length of the letter (a proxy for containing detailed information) and 2) whether the letter writer is the applicant’s research supervisor. We further examine how these variables differ by the gender and ethno-racial underrepresented status of the applicant.

Findings: Our results show that, in comparison to applicants who are denied entrance to the program, applicants who are admitted have more LORs from research supervisors and have longer LORs. Additionally, we find that recommenders write shorter letters for underrepresented students of color than white and Asian students. In comparison to white and Asian students, underrepresented students of color are also less likely to receive at least one LOR from a research supervisor. There are no differences in letter length or whether recommenders are likely to be a research supervisor by applicant gender.

Implications: Currently, how faculty decide who to admit to graduate engineering programs is an opaque process with substantial consequences for equity in STEM. The findings from this study will provide policy-relevant knowledge about the role that LOR play in structuring access to engineering programs.

keywords: bias , equity, gender, PhD students, mixed-methods research, admissions