Examining Racial Variation in Advising Program Outcomes Through LASSO Modeling

• In Event: Critical Analyses of Identity, Belonging, and Racialized Educational Outcomes

Abstract

This study applies LASSO regression within a QuantCrit framework to examine how outcomes vary by race/ethnicity and program participation among college advising participants. Using data from a nonprofit serving majority BIPOC students, nodewise LASSO identified which postsecondary and career outcomes were most strongly associated with subgroup membership. Results show that analyzing dosage alone obscures meaningful variation, with racialized subgroup differences emerging across both aspirations and career preparation. Findings underscore the importance of disaggregated analysis in equity-focused evaluations and highlight the need for programs to tailor support across diverse identities.

Authors

• Gloria A. Segovia, University of Illinois at Chicago

• George Karabatsos, University of Illinois at Chicago