Anthony Walsh
In the social world we often come across circumstances where the effect of one variable on another variable varies depends on different levels of a third variable. In statistical terms this is known as Interaction. The concept of interaction can often be confusing to students, who tend to think in terms of straight linear relationships. The present study offers a simple and effective way to teach this complex concept in the context of sociopolitical ideology. We examine the role of ideology in determining the sentencing recommendations of probation officers using ANOVA and OLS regression and contend that interaction terms can be presented to students in an easily understandable manner. Our approach helps integrate substantive theory with statistics to the benefit of both.