|
BPMSG |
 |
|
Wu, W.-H., Chiang, C.-T., & Lin, C.-T. (2008). Comparing the aggregation methods in the analytic hierarchy process when uniform distribution. WSEAS Transactions on Business and Economics, 5(3). Added by: Klaus D. Goepel (07 Jun 2019 13:12:53 Asia/Singapore) |
| Resource type: Journal Article BibTeX citation key: Wu2008 Email resource to friend View all bibliographic details  |
Categories: AHP/ANP Keywords: aggregation of individual judgments (AIJ), aggregation of individual priorities (AIP), Analytic Hierarchy Process (AHP), arithmetic mean, geometric mean Creators: Chiang, Lin, Wu Collection: WSEAS Transactions on Business and Economics |
Views: 10/506 Views index: % Popularity index: 11.25% |
| Abstract |
|
The Analytic Hierarchy Process (AHP) is a popular methodology for group decision making. Individual judgments can be aggregated in several ways, with the most effective approach being the aggregation of individual judgments (AIJ) and Individual priorities (AIP). When the judgments are aggregated, regardless of whether AIJ or AIP are used, arithmetic and geometric means are selected by decision makers. Some articles have discussed these two methods and made relevant suggestions. But when they discuses the issue of these, the distribution from judges’ opinion were not considering. This study performed simulation to generate the weights of the judgers, assuming that the opinion are distributed as Uniform distribution and then generated AHP weights using arithmetic and geometric means. Following performing statistical testing for the relative mean square errors between the parameter and estimator based on simulation, the results demonstrated no significant difference. Finally, based on the results of this study, we conclude the following: (1) if number of judges is not large both then both methods are applicable; (2) if number of judges is large then the geometric mean cannot be obtained, and the arithmetic mean is applicable; and (3) when the opinions of judges coincide, the arithmetic mean is applicable.
|