Which analysis is logistic regression in which the outcome variable has more than two categories?

Which analysis is logistic regression in which the outcome variable has more than two categories?

• A. Multiple regression
• B. Multinomial logistic regression
• C. Multivariate
• D. Multilevel linear model (MLM)

Answer: (B)

When the dependent variable is categorical with more than two categories, you use multinomial logistic regression. It extends binary logistic regression by modeling, for each category relative to a baseline, the log odds of being in that category as a function of the predictors. In practice, there are as many logit equations as there are non-baseline categories, which allows you to see how predictors influence the likelihood of each category compared with the reference. If there are only two categories, multinomial logistic regression reduces to standard binary logistic regression. The other analyses described handle different data types or structures: multiple regression is for a continuous outcome, multivariate methods deal with multiple dependent variables, and multilevel linear models handle hierarchical data with random effects.

When the dependent variable is categorical with more than two categories, you use multinomial logistic regression. It extends binary logistic regression by modeling, for each category relative to a baseline, the log odds of being in that category as a function of the predictors. In practice, there are as many logit equations as there are non-baseline categories, which allows you to see how predictors influence the likelihood of each category compared with the reference. If there are only two categories, multinomial logistic regression reduces to standard binary logistic regression. The other analyses described handle different data types or structures: multiple regression is for a continuous outcome, multivariate methods deal with multiple dependent variables, and multilevel linear models handle hierarchical data with random effects.