Modeling Polytomous Item Responses Using Simultaneously Estimated Multinomial Logistic Regression Models

Survey items with multiple response categories and multiple-choice test questions are ubiquitous in psychological and educational research. We illustrate the use of log-multiplicative association (LMA) models that are exten-sions of the well-known multinomial logistic regression model for multiple dependent outcome variables to reanalyze a set of items from a large political psychology survey. In particular, LMA models are shown to provide a useful analysis of items when the proper scoring rule for them is unclear. They also clearly reveal the performance of the items when instructions were altered to suppress ‘‘Don’t know’ ’ responses. LMA models can be fit rapidly using com-monly available software