Models and Methods for Categorical Data

This course covers statistical methods for analyzing categorical data. Model and theory includes: generalized linear models, including models for binary data, polytomous data (ordered and unordered), counts, contingency tables, matrix and graphical data. Classical and Bayesian inference in these models involves: latent variable representations, conditional likelihood, profile likelihood, and iterative algorithms. More advanced methods include: analysis of repeated measurements, data with cluster structure, nonparametric analysis, adaptive testing in contingency tables, multiple testing and data analysis in high-dimensions. Prerequisite: Statistical Science 521L or 721 and Statistical Science 532 or 732, or consent of instructor.