|其他题名||Heuristic Model or Integrative Model? A Test from the Point of View of Choice Process|
|关键词||整合模型 启发式模型 齐当别模型 决策过程 反应时 眼动|
|摘要||本研究从齐当别模型的视角，通过对决策过程反应时的考察分别对以累积预期理论(cumulative prospect theory)为代表的整合模型和启发式模型家族的重要成员--占优启发式模型(priority heuristic)--进行检验。其中，研究一、研究二、研究三和研究四采用反应时技术，研究五采用眼动技术分别对整合模型和启发式模型进行考察。研究一和研究二的结果表明，决策过程反应时并未随着占优启发式模型所假定的决策步骤的增加而变慢。研究三和研究四的结果表明，决策过程反应时并未随着两个选项的整体值差值的变大而变快。研究五的结果表明，决策者更多的进行基于特征的信启，搜索而不是基于选项的信息搜索。这些研究结果即不利于整合模型也不利于占优启发式模型。文章建议多角度、多指标探讨人们的决策过程，检验、修改、完善，以及建立新的启发式模型或决策过程模型(process model)，以增进对人们如何进行风险决策的理解。|
|其他摘要||A total of five experiments were conducted to compare the integrative model and heuristic model from the view of the equate-to-differentiate model (Li, 1994, 2004a, 2004b) by using a response time approach. Experiment 1 re-examined the priority heuristic by using the decision questions employed by Brandstatter, Gigerenzer and Hertwig (2006) but failed to duplicate their results. The priority heuristic assumed that the increase of steps that are required to be taken would be associated with the increase of time for making a choice. Experiment 2 examined the priority heuristic by manipulating the number of steps assumed by the priority heuristic and the difference between the best-outcome dimension and worst-outcome dimension assumed by the equate-to-differentiate model. It was revealed that the decision time did not increase with the increasing number of steps assumed by the priority heuristic. These results obtained in Experiments 1 and 2 were not friendly to the priority heuristic model. Experiments 3 and 4 were designed to examine the integrative model. Experiment 3 examined the integrative model by comparing the decision time under risk and ambiguity. Interestingly, the average decision time under risk was much longer than that under ambiguity, which was contrary to the implications of the integrative model because integrating probability distributions with outcomes under ambiguity to give expected values will take much longer time than integrating exact probabilities with outcomes under risk. Experiment 4 examined the integrative model by manipulating the difference between CPT values and the difference between the best-outcome dimension and worst-outcome dimension. The results showed that the decision time did not decrease with the increased difference between the CPT values but with the increased difference between the best-outcome dimension and worst-outcome dimension, which were not consistent with integrative model but consistent with the equate-to-differentiate model. Experiment 5 examined people’s information acquisition pattern by using an eye-tracking system. Results conflict with the prediction of the compensatory strategies and indicate that individuals use non-compensatory strategies. However, further investigation revealed that it was not the Priority Heuristic that people used. |
In sum, neither integrative model nor priority heuristic could help account for the data on choice process that we observed. Future work may focus on testing these two sets of models by employing methods which can provide a much richer description of the decision process than the response time approach employed in the present paper.
|汪祚军. 整合模型还是启发式模型？从决策过程角度进行的检测[D]. 北京. 中国科学院研究生院,2010.|