9 Multinomial Logistic Regression
IN THIS CHAPTER
Chapter 7, “Logistic Regression: The Basics,” and Chapter 8, “Logistic Regression: Further Issues,” both discussed problems in binomial logistic regression. That is, the problems all involved outcome variables that take on just two values: for example, True and False, Buys and Doesn’t Buy, Survives and Doesn’t Survive. But you can extend logistic regression to situations that involve outcome variables that take on three or more values, usually but not necessarily strictly nominal values. This chapter discusses those situations.
Working with three or more values ...