Analyzing and Visualizing Interactions in SAS
February 7 @ 9:00 am - 12:00 pm
In regression, we are often interested in an interaction, which is the modification of the effect of an independent variable by another. To understand the magnitude, direction, and significance of the interaction we need to decompose it into simple effects. Simple effects can be analyzed in three ways, by 1) testing each effect against zero, 2) testing differences among effects, and 3) visualizing each effect with graphs. This seminar will demonstrate how to do all three analyses in SAS PROC GLM / PLM for the following regression models: quadratic, continuous-by-continuous, categorical-by-continuous, categorical-by-categorical, and categorical-categorical-continuous interaction. The notes for the workshop are here. The sign-up link is here.