APPLICATION OF DISCRIMINANT ANALYSIS IN PREDICTING GROUP MEMBERSHIP OF EMPLOYEES

ABSTRACT

A researcher will frequently encounter many questions that, at best, need answers while doing a certain study. When the dependent variable is categorized as an interval variable, regression analysis is a technique that many researchers are quite familiar with. The researcher can utilize a discriminant analysis (DA) or a logistic regression (LR) method when the dependent variable is categorized as nominal. The generated model, which produced good projecting precision, was tested using DA with three criteria in this study. About 75.8% of the analysis's case samples have been accurately evaluated and classified by the discriminant function. Since the dependent variable contains three categories, the analysis generated two discriminant functions. The numerical results indicated that function 1 is more significant than function 2, since function 1 can explain 77.1% of the variance among the three groups whereas function 2 can only account for 22.9% of this variance.