ISSN 0439-755X
CN 11-1911/B
主办:中国心理学会
   中国科学院心理研究所
出版:科学出版社

心理学报 ›› 2017, Vol. 49 ›› Issue (2): 262-272.doi: 10.3724/SP.J.1041.2017.00262

• 论文 • 上一篇    下一篇

艾勒悖论再探——三结果决策的齐当别式表征: “好−坏”抑或“最好−最坏”?

李小平   

  1. (安徽师范大学教育科学学院, 芜湖 241000)
  • 收稿日期:2015-09-03 发布日期:2017-02-25 出版日期:2017-02-25
  • 通讯作者: 李小平, E-mail: lxpjq@126.com
  • 基金资助:

    教育部人文社会科学研究青年基金项目(13YJC840021)资助。

Allais Paradox revisit: The representing frame in the decision with three possible outcomes using the equate-to-differentiate model: Better-worse or best-worst comparison

LI Xiaoping   

  1. (Department of Educational Science, Anhui Normal University, Wuhu 241000, China)
  • Received:2015-09-03 Online:2017-02-25 Published:2017-02-25
  • Contact: LI Xiaoping, E-mail: lxpjq@126.com

摘要:

艾勒悖论的第一项决策任务属于三结果决策任务, 这对齐当别模型以及鉴别其不同判断任务间的优势都具有特殊意义。本研究通过3项子研究系统的探讨了不同表征方式(“最好−最坏”, “好−坏”)、不同判断任务(经典判断任务与联合判断任务)、不同数量级的结果值(万元级、百万元级以及亿元级)以及不同决策情境(迫选, 非迫选)上, 齐当别模型对艾勒悖论的解释力。结果显示:无论是迫选情境还是非迫选情境, 经典的艾勒悖论均可被齐当别模型解释, 但应基于“好−坏”而非“最好−最坏”表征维度; 齐当别模型的联合判断任务在三结果决策任务上的预测力也再次被证明优于经典判断任务。

关键词: 艾勒悖论, 三结果决策, 齐当别模型, “好−坏”表征维度, 结果值的数量级大小

Abstract:

The equate-to-differentiate model (EDM) is powerful in explaining the decision effects on the decision task with two possible outcomes. But relatively less is known about its power in explaining the effects on decision with three possible outcomes. Even less is known about the predictive power differences between the “classical matching task” and the “new joint judging task” (designed by Li X.P. etc., 2009) in testing the power of the EDM on the decision task with three possible outcomes. In this paper, therefore, the way EDM describes people’s behavior on the Allais paradox decision tasks involving decision with three possible outcomes was revisited. Specifically, this paper aimed to: (1) compare the “better-worse” and the “best-worst” representing frames in decision involving three possible outcomes (i.e., the Allais paradox decision task); (2) find out whether the predictive power of the EDM on the decision task with three possible outcomes is better explained by the “new joint judging task” than by the “classical matching task” ;(3) find out whether the above conclusion could be replicated when the decision values varied and when the non-imposed decision situation was adopted;(4) revisit the Allais paradox as demonstrated by EDM. Three studies were conducted and a total of 1129 participants were recruited for the research. Three different kinds of decision questionnaires were sent to the participants randomly in Study 1. One of them consisted of the first task of the Allais paradox + “classical matching task” + the “new joint judging task” and both of last two tasks were based on the “better-worse” representation; (2) Another of them consisted of the first task of the Allais paradox + “classical matching task” + the “new joint judging task” and both of last two tasks were based on the “best-worst” representation; (3) The third of them consisted of the second task of the Allais paradox + “classical judging task” + the “new joint judging task”. The magnitude of the decision value on the three questionnaires was ¥1000000. Two different decision values were used in Study 2, which were ¥10000 and ¥100000000 respectively. Otherwise, the design was identical to that in Study 1. Study 3 was similar to Study 2 except that the participants were not imposed to make their decision. They were asked to indicate the differences they saw on the judging task and “new joint judging task” in different ways. They indicated their choice and the differences they saw on a continuous axis similar to that used in Li and Xie (2006). The results revealed that: (1) people used the “better-worse” representation in the decision with three possible outcomes in EDM; (2) the “new joint judging task” was more powerful than the “classical matching task” in testing the EDM on decision with three possible outcomes; (3) the above two conclusions were replicated when the decision values varied and the non-imposed decision situation was adopted; (4) the power of EDM in explaining the Allais paradox could be further strengthened. The findings in this paper show that: (1) the EDM is powerful in describing the effects in decision with three possible outcomes, especially on the Allais paradox tasks; (2) the decision process on tasks with three possible outcomes as described by EDM is not parsimonious.

Key words: the Allais paradox, the decision with three possible outcomes, the equate-to-differentiate model (EDM), the better-worse representing frame, the magnitude of the decision value