Targeted Validation, Broader Benefit: Advancing Public Health with Two-Phase Designs

This session highlights recent innovations in two-phase studies, especially extreme-tail and related targeted sampling strategies, to emphasize how design choices yield broader public health benefits. Our speakers will showcase diverse applications: reducing bias in community-level food access metrics, extending causal inference to new populations despite incomplete covariate overlap, balancing efficiency across multiple models in pharmaceutical and observational studies, and improving cost-effective evaluation of cancer screening tests. Collectively, these talks demonstrate how modern applications and extensions of extreme tail sampling and two-phase designs advance beyond traditional paradigms to meet the needs of complex health research. In line with the JSM theme, Communities in Action: Advancing Society, the session underscores how innovations in extreme-tail and targeted two-phase sampling can generate more accurate, equitable, and actionable insights for the communities we aim to serve.