Academy of Strategic Management Journal (Print ISSN: 1544-1458; Online ISSN: 1939-6104)


Multivariate Discriminant Analysis Managing Staff Appraisal Case Study

Author(s): Muwafaq AlKubaisi, Waleed A. Aziz, Shaju George, Khaled Al-Tarawneh

One of the thought-provoking tasks facing an academic investigator is the data analysis part where the investigator needs to recognize the precise analysis technique and how to interpret the output. The analysis steps can be done using various statistical computer packages with ease. Many researchers are very acquainted and familiar with the regression analysis technique when the dependent variable is classified as an interval variable. However, if the dependent variable is classified as nominal, then the researcher can use a discriminant analysis (DA) or a logistic regression (LR) technique. This research has used DA with three criteria to test the developed model which produced an excellent projecting precision. The discriminant function has properly assessed and classifies about 67% of the cases that are included in the analysis. The analysis produced two discriminant functions, as the dependent variable has three categories. The numerical results showed that function 1 is more critical than function two as 77% of the variance among the three groups can be explained by function one whereas only 23% of this variance can be explained by function 2.
Get the App