Project Ranking 2
Final technical rept. Jul 1984-Mar 1986
AUBURN UNIV AL ENGINEERING EXPERIMENT STATION
Pagination or Media Count:
This report describes the methodology developed for a defense department laboratory for project selection and the results of use of the computerized ranking procedure for other complex managerial decisions. In addition to project selection, the system is feasible and instructive in personnel selection, comparison of management strategies, purchasing decisions, and merit-based distribution of awards. The massive input required by many existing systems is not necessary for this absolute deviation ranking system. Up to 20 attributes and 200 projects can be compared. The system allows many missing comparisons. Measures of consistency and reliability are proposed for estimating the validity of the final weights. Preliminary experiments and actual use of the system provide empirical validation and show the results are satisfactory. Further research is underway for the refinement of the system. The impact on reliability of the number of items compared, the comparison scale employed, and the number of missing observations provide fertile fields for future work. Extensions of the basic and additive individualized multicriteria models provide for voting, hierarchy, and constraints options for group decision making. The current version of the system is an organized, structured way of making decisions which enhances the identification.
- Administration and Management
- Computer Programming and Software