Accession Number : ADA303592
Title : The Optimality of Sequential Personnel Assignments Using a Decision Index.
Descriptive Note : Interim technical paper Jul 93-Jul 94,
Corporate Author : ARMSTRONG LAB BROOKS AFB TX HUMAN RESOURCES DIRECTORATE
Personal Author(s) : Grobman, Jeffrey H. ; Alley, William E. ; Pettit, Raymond S.
PDF Url : ADA303592
Report Date : SEP 1995
Pagination or Media Count : 22
Abstract : The purpose of this study was to investigate the optimality of the decision-index (DI) as it might be used for personnel assignment in a sequential or 'first come-first served' manner. Monte Carlo methods were implemented to generate sets of person-job payoff matrices representing the utilities or payoffs achievable from the prospective assignment of individuals to specific jobs. These matrices were generated with various properties including varying batch sizes, personnel rejection rates, person-to-job ratios, and personnel payoff distributions. Simulated sequential 'assignments' were made from these matrices using 'highest payoff' and decision-index strategies. The 'highest payoff' method assigned individuals to jobs so that the potential 'payoff' for each individual was maximized. This approach contrasts with the decision-index method that transformed individual payoff scores, prior to personnel assignment, in order to increase the overall optimality of assignments. Additionally, linear programming techniques established optimal and minimal assignment solutions. Comparisons of the results from each of these methods demonstrated that sequential assignments with a decision index could attain approximately 92% of the utility of an optimal system for all conditions. Furthermore, use of a 'highest payoff' decision rule produced results nearly equivalent to those from Di-based sequential assignments when payoff matrix row and column means were roughly equal. These results quantity the losses in overall utility that result from using Di-based sequential methods to assign personnel to jobs. (AN)
Descriptors : *MANPOWER UTILIZATION, MATHEMATICAL MODELS, ALGORITHMS, OPTIMIZATION, HUMAN RESOURCES, PERFORMANCE(HUMAN), PROBABILITY DISTRIBUTION FUNCTIONS, MATRICES(MATHEMATICS), LINEAR PROGRAMMING, MONTE CARLO METHOD, SEQUENTIAL ANALYSIS, DECISION THEORY, ALLOCATIONS, PERSONNEL SELECTION, BILLETS(PERSONNEL), BATCH PROCESSING, REJECTION, JOB SHOP SCHEDULING, MANAGEMENT BENCHMARKING.
Subject Categories : Personnel Management and Labor Relations
Distribution Statement : APPROVED FOR PUBLIC RELEASE