Refer to Exhibit 1. Use the same variables you used for Discriminant Analysis and fit a Nominal Logistic Model (Analyze>Fit Model, Cast the variable "Personal Loan" to Y field, all continuous variables to Model Effects Box, and the variable "Validation 2" to Validation field). . The overall misclassification rate for Logistic (on the validation set) is 17.23% . Discriminant correctly classifies 127 v the 192 acceptors (in the validation set) as acceptors, while Logistic only classifies 171 v of the 192 as acceptors (in the validation set). . So, the Discriminant v method does a better job of correctly classifying acceptors (when 0.5 is the cutoff probability)