著者

竹村 和久 原口 僚平 玉利 祐樹

出版者

日本認知科学会

雑誌

認知科学 (ISSN:13417924)

巻号頁・発行日

vol.22, no.3, pp.368-388, 2015

---

This paper examines two-stage multi-attribute decision strategies in different condi-tions where numbers of alternatives, and attributes are varied. A Monte-Carlo com-puter simulation using the concept of elementary information processes identified de-cision strategies that approximate the accuracy of normative procedures while savingcognitive effort in the two-stage decision making process. The elementary strategiesexamined in the simulation were nine decision strategies: lexicographic, lexicographicsemi-order, elimination by aspect, conjunctive, disjunctive, weighted additive, equallyweighted additive, additive difference, and majority of confirming dimensions strategies.Elementary information process and relative accuracy were computed for all combina-tion of two decision strategies for two-phased decision making process. The result ofthe computer simulation suggested that comparatively effortless and accurate heuristicwas the two-phased strategy that used lexicographic strategy to eliminate until a fewalternatives in the first stage and used weighted additive strategy in the second stage.Lastly, theoretical and practical implications of this study were discussed.