Advanced Analysis of Categorical Data
Much of the data in political science (and in the social sciences, in general) are categorical, and overlooking this feature often leads to the application of inappropriate methods of analysis. While categorical data carry the possibility of revealing more complex association structures than those that may be discovered under the common assumption of continuity and normal distribution, the application of inappropriate methods makes fine-tuning of the models impossible. The central theme of the class is the concept of interaction, which covers various forms of effects and associations. Methods for the analysis of interactions have proved useful across the social sciences, where the focus of interest is mostly the joint behaviour of the variables, or effects of some of them on others, as opposed to the individual analyses of the variables. The use of mathematics will be tailored to the level of the audience.