Commentary: Using Potential Outcomes to Understand Causal Mediation Analysis
نویسندگان
چکیده
We begin our discussion by congratulating Maxwell, Cole, and Mitchell (2011; hereafter MCM) for providing a thought-provoking article that tackles the challenging problem of conducting mediation analysis in longitudinal settings. As MCM clearly demonstrated, although mediation analysis plays an essential role in psychological research, conducting such an analysis requires researchers to understand the underlying assumptions and adopt appropriate statistical analysis and research design strategies. In this commentary, we use the potential outcomes framework to highlight the fundamental issues that arise in mediation analysis. Although MCM did
منابع مشابه
Reply to commentary by Imai, Keele, Tingley, and Yamamoto concerning causal mediation analysis.
This comment clarifies how structural causal models unify the graphical and potential outcome approaches to mediation, and why the resulting mediation formulas are identical in both frameworks. It further explains under what conditions ignorability-based assumptions are over-restrictive and why such assumptions require graphical interpretations before they can be judged for plausibility. Finall...
متن کاملCausal Mediation Analysis 1 Running head: IMAI, KEELE, TINGLEY, AND YAMAMOTO Commentary: Practical Implications of Theoretical Results for Causal Mediation Analysis
متن کامل
Understanding how pain education causes changes in pain and disability: protocol for a causal mediation analysis of the PREVENT trial.
INTRODUCTION Pain education is a complex intervention developed to help clinicians manage low back pain. Although complex interventions are usually evaluated by their effects on outcomes, such as pain or disability, most do not directly target these outcomes; instead, they target intermediate factors that are presumed to be associated with the outcomes. The mechanisms underlying treatment effec...
متن کاملReply to Commentary by Imai, Keele, Tingley, and Yamamoto (2014) Concerning Causal Mediation Analysis
This comment clarifies how Structural Causal Models (SCM) unify the graphical and potential outcome approaches to mediation, and why the resulting mediation formulas are identical in both frameworks. It further explains under what conditions ignorability based assumptions are over-restrictive and why such assumptions require graphical interpretations before they can be judged for plausibility. ...
متن کاملA refreshing account of principal stratification.
Pearl (2011) invites researchers to contribute to a discussion on the logic and utility of principal stratification in causal inference, raising some thought-provoking questions. In our commentary, we discuss the role of principal stratification in causal inference, describing why we view the principal stratification framework as useful for addressing causal inference problems where causal esti...
متن کامل