Week 6
Multinomial Regression
Before class
Textbook
Xing Liu (2023). Categorical Data Analysis and Multilevel Modeling Using R (1st edition). SAGE Publications Ltd.
Describe the difference between a stratified versus a multinomial approach to modeling an outcome with more than two nominal categories.
Describe how you would interpret the odds ratio of an input variable for a given category in a stratified modeling approach.
Describe what is meant by the ‘reference’ of a multinomial logistic regression model with at least three nominal outcome categories.
Describe how you would interpret the odds ratio of an input variable for a given category in a multinomial modeling approach.
Given a multinomial logistic regression model with outcome categories A, B, C and D and reference category A, describe two ways to determine the coefficients of a multinomial logistic regression model with reference category C.
Describe a process for safely simplifying a multinomial logistic regression model by removing input variables.