University of Cincinnati Lindner College of Business

Back to Search Results

Mediation Analysis: A New Test when All or Some Variables are Categorical

Author(s): Jiaxiu He, Xin (Shane) Wang, David Curry

Status: Accepted
Year: 2017
Publication Name: International Journal of Research in Marketing
Volume: 4, Issue: 2017


Abstract

Statistical tests for mediation in consumer research typically use a regression coefficients (RegCoeff)-based framework. We present a new test based on likelihood ratio principles to complement the RegCoeff approach. We compare the new test’s performance to conventional methods that use PROCESS, MEDIATION, and extensions thereof to categorical measures. Our tests address situations in which the assumptions of current methods tend not to hold; that is when one or more variables may be categorical or relationships among constructs may be nonlinear. In such environments, results significantly favor the new test, which we call LRT for likelihood ratio test. For example, among 72 tests of power, LRT loses just 9 times to PROCESS. It wins outright in 45 cases and ties in 18 others. Differences are non-trivial; they are statistically significant at the 0.001 level in 21 of the 45 wins. To assist researchers, we detail how and when the new test can best complement existing methods and offer software that performs the new test for any combination of continuous and categorical variables.


Read

Google Scholar Link