On Testing Moderation Effects in Experiments Using Logistic Regression

18 Pages Posted: 13 Feb 2014

See all articles by James D. Hess

James D. Hess

University of Houston

Ye Hu

University of Houston - Bauer College of Business

Ed Blair

University of Houston - C.T. Bauer College of Business

Date Written: February 10, 2014

Abstract

Consumer researchers seeking to explain the probability of a binary outcome in an experiment often attend to the moderation of one treatment variable’s effect by the value of second. The most commonly used method for analyzing such data is logistic regression, but because this method subjects the dependent variable to a nonlinear transformation, the resulting interaction coefficients do not properly reflect moderation effects in the original probabilities. Significant moderation effects may result in non-significant interaction coefficients and vice versa. We illustrate the problem, discuss possible responses, describe how to correctly test moderation effects on probabilities, and demonstrate that addressing this problem makes a practical difference.

Keywords: logistic regression, nonlinear transformation, moderation, experiments

Suggested Citation

Hess, James D. and Hu, Ye and Blair, Ed, On Testing Moderation Effects in Experiments Using Logistic Regression (February 10, 2014). Available at SSRN: https://ssrn.com/abstract=2393725 or http://dx.doi.org/10.2139/ssrn.2393725

James D. Hess (Contact Author)

University of Houston ( email )

4800 Calhoun Road
Houston, TX 77204
United States

Ye Hu

University of Houston - Bauer College of Business ( email )

Houston, TX 77204-6021
United States

Ed Blair

University of Houston - C.T. Bauer College of Business ( email )

Houston, TX 77204-6021
United States

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
137
Abstract Views
597
rank
236,952
PlumX Metrics