Power Analysis for Two-Level Hierarchical Models With Treatment at Cluster Level in Nonrandomized Designs

• In Event: 26.081 - Poster Session 3
  In Poster Session: 26.081-4 - Division D Section 2 Poster Session: Statistical Methods

Abstract

The purpose of this study is to examine the power of two-level hierarchical models with treatment applied at the cluster level in nonrandomized designs. Power analysis under nine conditions were examined before (pre-matching) and after the propensity score analysis; that included two kinds of propensity score matching mechanisms, one-step matching and two-steps matching, in three kinds of population scenarios. The results suggested the minimum numbers of second level units for the appropriate power in different conditions, pointed out possible inflated standard error issues, and further brought out the validity issues that commonly ignored in simulation studies.

Authors

• Chi Chang, Michigan State University

• Richard T. Houang, Michigan State University

• Kimberly S. Maier, Michigan State University