| Summary: | 碩士 === 明新科技大學 === 工程管理研究所 === 93 === Facing on the problem of performance evaluation, there are four typical issues for decision maker to consider, they are 1) set up the construction of performance evaluation, 2) criteria weighting setting, 3) criteria grading evaluation, 4) performance scoring calculation. In order to avoid judgment bias and autocratic decision, bringing together the experts to obtain the group consensus is one of the best choices for achieving accurate performance evaluation.
The traditional performance evaluation model is performed under the certainty decision environments. In that case, the data collected is always quantification and constant, both in criteria weighting and criteria grading. While in the situations of real case, in organizational review or in personal review, the application of performance evaluation task is performed under the circumstances: subjective judgments, qualitative measurements, uncertainty, group decision making.
This thesis will focus on literature review in topics of group decision making, interactive group consensus, measuring data analysis for crisp, intervals and fuzzy number, data aggregating method and AHP methodology for setting criteria weighting, also we will propose an algorithm to define an AHP comparison matrix under consistency assurance.
Saaty proposed Analytic Hierarchy Process (AHP) as the method to obtain the corresponding comparative criteria weighting. In practice, people can’t easily and precisely set up paired criteria comparison scale, especially in case that the number of criteria is large. This thesis proposed a two-stage methodology, which applied Brown Gibson’s Rule to set the corresponding criteria priority, then designed an algorithm to process AHP paired comparison matrix under the assumption of full consistence. Each DM only needs (n-1) paired criteria comparisons to get full consistence assured AHP matrix.
|
|---|
