Association for Public Policy Analysis & Management

This paper presents results from the impact evaluation of one of the National Science Foundation’s (NSF) recent initiatives to produce Ph.D.s in STEM: the Bridge to the Doctorate (BD) fellowship. BD was established in 2003 as an activity of the Louis Stokes Alliances for Minority Participation program. Its ultimate goal is to increase the number of underrepresented minority (URM) students successfully completing doctoral degrees in STEM. To achieve this goal, BD provides the initial two years of support for master’s or doctoral studies in STEM. However, as its name indicates, the expectation is that many if not most of the students supported will continue on to Ph.D. programs. The BD program thus serves as a “bridge” to retain minorities on the path to a Ph.D.

The goal of the evaluation was to establish whether BD increased the number of URM students entering the Ph.D. track (the short-term goal) as a signal of progress toward completion of doctoral degrees (the medium-term goal) and eventually joining the STEM workforce (the long-term goal). In this study, we defined entry into the Ph.D. track as having completed two years of graduate-level coursework, equivalent to a master’s degree, and transitioning on to Ph.D.-level work. To measure progress, we constructed measures of key educational milestones on the path to the Ph.D.—including completing master’s degrees, enrolling in a Ph.D. program, entering the Ph.D. track, passing Ph.D. qualifying exams, attaining Ph.D. candidacy and completing a Ph.D.

We used a quasi-experimental design to compare the educational progress of the fellows sample, consisting of students funded between 2003 and 2006 at 43 universities, and a matched sample of students who entered the same programs of studies in the same universities and years but did not receive the fellowship. Based on the data collected, we constructed two comparison groups—one of minority students (the comparison sample, created using propensity score matching) and one of nonminority students (the benchmarking sample). We used these two comparison samples to evaluate the impact of the program and to benchmark outcomes, respectively. The analysis was based on longitudinal, individual-level data submitted yearly between 2008 and 2012 by universities enrolling students.

We find that, overall, compared to other URM students, students supported through BD were more likely to complete master’s degrees and as likely to enter the Ph.D. track and complete doctoral degrees. Among students pursuing master’s degrees, BD has a positive impact on every milestone leading to and including Ph.D. completion. Compared to other URM students in master’s programs, BD fellows are more likely to complete master’s degrees, enroll in doctoral programs, enter the Ph.D. track, complete qualifying exams, and receive doctoral degrees. Among students who receive funding to pursue doctoral studies, however, BD has mixed impacts. Compared to other URM students in doctoral programs, BD fellows are more likely to complete master’s degrees, less likely to enter the Ph.D. track, and equally likely to complete doctoral degrees. In this presentation, we will discuss these findings, advance some explanations, and discuss programmatic and policy implications.