Recommendations on the Improvements of the Scoring Method for the Most Advantageous Tender.

• Main Authors: Chuan-Cheng Cheng, 鄭全成
• Other Authors: Hao-Hsi Tseng
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/18915003409529665056

碩士 === 國立宜蘭大學 === 土木工程學系碩士班 === 94 === Using the Most Advantageous Tender (MAT) for government procurement projects has become popular in recent years. However, the problem of fairness and justice of MAT have been criticized because the results might be significantly altered by a single committee member with prejudice. At present, the most popular evaluation methods for MAT are the Overall Evaluated Score Method, Price Per Score Point Method, Ranking Method, and other approaches approved by the authorities. Unfortunately, none of the methods are capable of resolving the problems of MAT. Therefore, this study aims to improve the current evaluation method by locating the defects of the MAT by adopting the Arrow's Impossibility Theorem and the voting rule of the Public Choice. A new score model namely Optimum Overall Evaluated Score Method which corrects the flaws of MAT is also proposed. In the method, the raw scores of all committee members are unified into a unification base line such that the extreme scores are corrected logically. Through this way, the results can reflect every committee member’s preference fairly. Furthermore, in order to guarantee no remained extreme score exists, the final scores are then corrected by the Geometric Average Method. As a result, the committee members can give the scores based on the Overall Evaluated Score Method, and the individual impacts on final result is avoided. The result of practical and simulated case shows that the proposed model is capable to obtain the equivalent ranking as fine as the other scoring systems in normal situations. In addition, if certain extreme scores emerge, the improved method can also effectively resolve the situation. By testing all the scoring methods with Arrow's impossibility theorem, it appears that the proposed scoring model is more precise and logical than the others.