The Role of Institutional Context in Shaping Program Outcomes: A Multisite Longitudinal Evaluation Study

• In Event: Assessing STEM Student Success in Higher Education: Insights From Multidisciplinary Research

Abstract

How program evaluators use statistical analyses shapes their research toward or away from a critical agenda that considers power dynamics and the role of normative and oppressive systems. Employing a critical quantitative approach to statistical modeling for an undergraduate science, technology, engineering, mathematics, and medicine (STEMM) intervention program evaluation, we demonstrate why random effects are more appropriate to utilize rather than often-used fixed effects for a regression model with multiple sites. This study is significant for education program evaluators and researchers interested in improving methods to better capture institutional contexts, especially for institutions that are under-resourced yet often might not be getting the credit they deserve for developing the talent of historically excluded groups in STEMM.

Authors

• Krystle Cobian, University of California - Los Angeles

• Jayashri Srinivasan, University of California - Los Angeles