Categorical Data Analysis: Poisson GLM

This lecture covers the application of Poisson Generalized Linear Models (GLM) to analyze count data cross-classified by different categories. It explores the relationship between Poisson, multinomial, and product multinomial likelihoods, focusing on fitting models to contingency tables. The instructor discusses the connection with categorical data, Bayesian inference, and log-linear models. Examples include analyzing smoking data and jacamar response to butterflies. The lecture emphasizes the equivalence of inferences between Poisson and multinomial models and the estimation of parameters under different scenarios.