Elicitation of Priors for Intervention Effects in Educational Trial Data
摘要
Effect sizes for educational interventions are commonly small, and hence decisions to re-grant efficacy trials (small trials with homogeneous populations under idealized conditions) as effectiveness trials (larger trials with heterogeneous populations) are often based on limited evidence from the efficacy trial itself. However, supplementary evidence may be available on how (past) effectiveness trials with similar outcomes tend to perform. This work proposes a Bayesian approach of making use of such evidence for re-granting decisions.