In the last chapter we analyzed data on criminal court dispositions that could be explained by the main effects of the two factor variables. In this chapter we introduce more complexity into the analysis by adding another variable and another level to the type of offense variable. Because of these changes, a main-effects model will no longer fit the data and a model including interactions among the factors will be required. Besides introducing the modeling of interaction in this chapter, we also consider a new type of model—the modular model—in our attempt to characterize the variation in the data.
KeywordsFactor Variable Prison Sentence Main Effect Model Modular Model Physician Type
Unable to display preview. Download preview PDF.