Use of Dummy Variable Regression Techniques in Estimating Joint and Marginal Values of an Outcome

Abstract

This paper proposes a statistical method for estimating the values, including the joint and marginal values of an outcome variable, using dummy variable multiple regression techniques. Estimates are also provided for the direct and indirect effects of a parent independent variable on a dependent variable through its representative dummy variables in the presence of other independent variables in the model. The proposed method is illustrated with some data.

Keywords: Dummy Variables, Total Effect, Direct Effect, Indirect Effect, Parent Variables, Mediation Model, Joint Value, Marginal Value.