What regression technique is used when the dependent variable is categorical, with binary or multinomial outcomes?

What regression technique is used when the dependent variable is categorical, with binary or multinomial outcomes?

• A. Logistic regression
• B. Linear regression
• C. Poisson regression
• D. Multilevel regression

Answer: (A)

When the outcome is categorical, you need a method that models the probabilities of belonging to each category rather than predicting a numeric value. Logistic regression does this by linking predictors to the log odds of the outcome and producing probabilities between 0 and 1. For a binary outcome, this is the standard binary logistic regression; for more than two categories, the approach generalizes to multinomial logistic regression, which still centers on estimating category probabilities. Linear regression is inappropriate here because it predicts a continuous outcome and can yield impossible values outside [0,1]. Poisson regression is for count data, not category membership. Multilevel regression is a flexible framework for nested data and can handle various outcomes, but the specific method used for categorical outcomes is logistic regression (with its multinomial extension when there are multiple categories).

When the outcome is categorical, you need a method that models the probabilities of belonging to each category rather than predicting a numeric value. Logistic regression does this by linking predictors to the log odds of the outcome and producing probabilities between 0 and 1. For a binary outcome, this is the standard binary logistic regression; for more than two categories, the approach generalizes to multinomial logistic regression, which still centers on estimating category probabilities. Linear regression is inappropriate here because it predicts a continuous outcome and can yield impossible values outside [0,1]. Poisson regression is for count data, not category membership. Multilevel regression is a flexible framework for nested data and can handle various outcomes, but the specific method used for categorical outcomes is logistic regression (with its multinomial extension when there are multiple categories).