AN EVALUATION OF THE DISCRIMINANT ASSIGNMENT MODEL.
Abstract
The military service continually faces the problem of assigning personnel to training school in a manner that will result in an optimum use of its human resources. A possible solution to this problem might lie in the use of a discriminant assignment model that is evaluated in this paper. The model makes a probabilistic prediction that an airman's assignment to a given school is desirable by using Bayes' theorem and historical data of individuals' past school performances. Data from a policy capturing experiment on a group of 75 airmen is used as a measure of past school performances. From the analysis of this data, airman-school predictions are made on five airmen to eight schools. The five airmen are then assigned to five schools to demonstrate that the probabilistic predictions may be used to make optimum allocations of airmen to schools. The results of a rank correlation test indicate that the model predictions have a significant rank correlation with counselor assignment ratings. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1966
- Accession Number
- AD0806484
Entities
People
- Peter James Phillips
Organizations
- Air Force Institute of Technology