APS
2024 APS Annual Convention · 2024
Propensity Score Analysis with Missing Data: Imputation Algorithm Matters
- Evelyn Shifflett
University of California, Merced - Fan Jia
University of California, Merced - Haiyan Liu
University of California, Merced
Abstract
Propensity score analysis (PSA) combined with multiple imputation (MI) is a viable strategy for addressing imbalanced groups and confounding in quasi-experimental designs with missing data. Literature is divided on how best to combine PSA with MI. Simulation study results indicate both MI-Within and MI-Across produce unbiased treatment effect estimates.
Other
