Confounding in Causal Inference

Causal inference is an important but controversial topic in the social sciences in that it is difficult to statistically control for all possible confounding variables. To address this concern, this monograph introduces a reference distribution of the confounding that is the product of two dependent correlation coefficients and illustrates how to use the reference distribution to investigate the robustness of a cause inference to the impact of a confounding variable. The methodology discussed in this monograph would also allow for multiple partial causes in the complex social phenomena under study, so as to inform causal inferences in the social sciences from statistical linear models.

Autorentext

Dr. Wei Pan is an Associate Professor of Quantitative Research Methodology at the University of Cincinnati, USA. His current research interests are causal inference, resampling, hierarchical linear models, structural equation modeling, meta- analysis, and their applications in the social, behavioral, and health sciences.