Testing for interaction in analysis of variance


The regression model, even one employing dummy variables, assumes that the effects of the independent variables on the dependent variable are additive. However, in the presence of interaction, these effects will not be strictly additive. For example, in the dummy variable regression of income on gender and minority group status, interaction would be present if the effect of gender on income was not the same for both minorities and nonminorities. As in the case of multiple regression with continuous variables, we can test for the presence of interaction using interaction variables that are multiplicative functions of the independent variables. These interaction variables also enable us to estimate the effects on the dependent variable of any interaction between the independent variables. In analysis of variance models, of course, both the independent variables and the interaction variables are dummy variables.


Regression Model American Woman Multiple Regression Model Interaction Variable Multiplicative Function 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Plenum Press, New York 1997

Personalised recommendations