Premium
A study on aggregation of group decisions
Author(s) -
Huang YeuShiang,
Liao JingTai,
Lin ZuLiang
Publication year - 2008
Publication title -
systems research and behavioral science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.371
H-Index - 45
eISSN - 1099-1743
pISSN - 1092-7026
DOI - 10.1002/sres.941
Subject(s) - analytic hierarchy process , decision maker , group decision making , operations research , aggregate (composite) , group (periodic table) , decision analysis , decision problem , computer science , order (exchange) , weighted sum model , decision model , management science , simple (philosophy) , artificial intelligence , optimal decision , mathematics , psychology , statistics , machine learning , decision tree , social psychology , engineering , economics , algorithm , epistemology , materials science , chemistry , composite material , organic chemistry , finance , philosophy
The analytic hierarchy process (AHP), proposed by Saaty in 1971, has already become a common decision technique in the field of decision science. In attempting to deal with group decision problems, a vast amount of literature has adopted geometric means or arithmetic means to aggregate the multiple opinions of group decision makers with AHP. However, geometric means and arithmetic means are usually subject to the influence of extreme values, and make use of only simple average types of manipulation to combine the judgments of decision makers which may not practically capture the group preferences among alternatives. In this paper, two aspects of consideration are probed into the construction of a group AHP decision model, which are ‘preferential differences’ to denote the differences of preferential weights among alternatives for each decision maker, and ‘preferential ranks’ to denote the ranks of the alternatives for each decision maker. A comparative analysis is performed to investigate the effectiveness of the proposed model, and a ‘satisfaction index’ is suggested to indicate the satisfactory level of the final obtained group decision. Finally, a feedback interview is carried out to collect the post‐model opinions of decision makers in order to understand their acceptance about the proposed model. Copyright © 2008 John Wiley & Sons, Ltd.