What Works for Whom?
Bayesian Adaptive Design for Prevention Research
Mariel Finucane, PhD
M50 | Mathematica Policy Research
ABSTRACT: A Bayesian approach to randomized program evaluations efficiently identifies what works for whom. The Bayesian design adapts to accumulating evidence: Over the course of an evaluation, more study subjects are allocated to treatment arms that are more promising, given the specific subgroup from which each subject comes. We identify conditions under which there is more than a 90% chance that inference from the Bayesian adaptive design is superior to inference from a standard design, using less than one third the sample size.