Director of Institutional Research Pratt Institute Fairfield, Connecticut, United States
Session Abstract: Student success programs often show promising outcomes, but how much is due to the program versus who chooses to participate? This session demonstrates how inverse probability of treatment weighting (IPTW), a propensity score–based method, addresses selection bias in program evaluation by reweighting samples to create balanced comparison groups while retaining full datasets. Using examples from a Math Readiness Program, participants will see how IPTW provides more credible estimates of program impact. The session emphasizes practical implementation, including diagnostic checks for validity and Tableau dashboards that make complex adjustments transparent to stakeholders. Participants will gain frameworks for applying these methods to their own evaluations, moving beyond descriptive comparisons to causal inference that withstands scrutiny and informs strategic decisions.
