The discriminant function analysis is particularly useful as a follow-up to which analysis?

The discriminant function analysis is particularly useful as a follow-up to which analysis?

• A. ANOVA
• B. MANOVA
• C. PCA
• D. Regression

Answer: (B)

Discriminant function analysis is the step that takes the finding that groups differ on a set of dependent variables and turns it into a practical way to understand and use those differences. When MANOVA shows that predefined groups differ across a combination of variables, discriminant analysis finds the linear combinations of those variables—called discriminant functions—that maximize separation between the groups. These functions reveal which variables contribute most to distinguishing the groups and provide a way to classify new cases into the correct group with a measure of accuracy. In short, MANOVA tells you that differences exist across multiple variables together, and discriminant analysis explains how those differences are organized and how to predict group membership from the variables. The other analyses don’t fit as a natural follow-up: ANOVA tests one variable at a time, PCA is about reducing dimensionality without regard to group labels, and regression aims to predict a continuous outcome rather than classify cases into groups.

Discriminant function analysis is the step that takes the finding that groups differ on a set of dependent variables and turns it into a practical way to understand and use those differences. When MANOVA shows that predefined groups differ across a combination of variables, discriminant analysis finds the linear combinations of those variables—called discriminant functions—that maximize separation between the groups. These functions reveal which variables contribute most to distinguishing the groups and provide a way to classify new cases into the correct group with a measure of accuracy. In short, MANOVA tells you that differences exist across multiple variables together, and discriminant analysis explains how those differences are organized and how to predict group membership from the variables. The other analyses don’t fit as a natural follow-up: ANOVA tests one variable at a time, PCA is about reducing dimensionality without regard to group labels, and regression aims to predict a continuous outcome rather than classify cases into groups.