Design and Analysis in the Interventional Mediation Framework

• In Event: Friday Roundtable Session 9:45 am
  In Roundtable Session: Innovation in Statistical Methods (Table 31)

Abstract

A critical limitation to conventional and most contemporary mediation frameworks involving multiple mediators is that the validity of their inferences requires the correct a priori specification of the causal structure among mediators. Recent work has developed an alternative framework based on ‘interventional effects’ that relaxes this assumption. The purpose of this study was to extend the interventional mediation framework to multilevel settings. We extend this framework to multilevel settings by delineating the assumptions for identification and developing principles of estimation, sampling variability and statistical power under several different multilevel structures, designs and estimators. The results provide tools to design and analyze multilevel mediation studies using the much more flexible interventional framework.

Authors

• Fangxing Bai, Montana State University

• Benjamin Kelcey, University of Cincinnati