Files
Abstract
Pre-college access programs operating in high schools commonly focus on assisting underserved and historically excluded populations of students in preparing for, applying to, and enrolling in postsecondary education. Through two analyses, this study examines the efficacy of two types of access programs in facilitating postsecondary enrollment and persistence outcomes for students and schools served by access programs, over the past two decades. The first analysis uses a propensity score weighting approach and data from the Educational Longitudinal Study of 2002 to examine academic undermatching outcomes when applying to and enrolling in college for students participating in the federal Talent Search, Upward Bound, and Gear Up programs. Results suggest that participation in one of the federal pre-college access programs did not reduce rates of undermatching for students neither when applying to, nor enrolling in college.The second analysis utilizes a difference-in-differences design to investigate school-level impacts of schools treated by near-peer advising programs in the state of Michigan between 2007 and 2020, compared to schools that were untreated but located in the same school districts, and a comparable set of schools identified through Coarsened Exact Matching. Using a panel dataset of Michigan high schools and data provided by the National College Advising Corps, and the Michigan College Access Network, this analysis specifically investigates whether near-peer programs impacted schoolwide graduation, college-going, and college credit attainment rates of high school graduates. Given few school-level analyses of such programs, and the focus these programs place on serving historically excluded populations, the second analysis also disaggregates these outcomes by student racial and ethnic subgroups. The results from the analysis are mixed though they consistently suggest near-peer programs increased rates of college-going for the female student population. Results for other student subgroups vary contextually.