Social scientists and policy analysts frequently wish to study categoric variables such as the choices that people make in economic, political and social settings. This applied workshop teaches researchers how to conduct statistical analyses with categoric dependent variables using the popular Stata and R programs.
Starting with the linear regression model, the workshop focuses on binomial logit models, ordinal logit models, the multinomial logit model and the mixed logit model. Examples using national survey data on voting in the 2016 UK Brexit referendum, the 2015 Canadian federal election and the 2016 US presidential election are presented. Assuming only familiarity with introductory statistics, the workshop emphasizes how to do multivariate analyses of categoric dependent variables with real-world data, to interpret results and present them in informative graphs.