Abstract: In our daily life, both layman and experts are often called to evaluate the probability of uncertain events and the time duration of some tasks or projects, such as who is the winner in the World Cup and how long it will take to finish this new project. Such assessments and judgments play an very substantial role in decision making. Psychologists, economists and researchers in other fields have taken deep exploration in human subjective judgment. One effect that has been studied in great detail across a variety of domains is the unpacking effect, which refers to the phenomenon that the total value assigned to a hypothesis is often greater if the hypothesis is decomposed into individual parts which are then evaluated separately. The summation of these individual parts yields a higher valuation than if the hypothesis had been evaluated as a whole. In the present paper, we comprehensively summarized research on the unpacking effect over 20 years. First we introduced the origin of the unpacking effect. Tversky and Koehler put up with the unpacking effect in the Support Theory for the first time in 1994 and they regarded it as the subadditivity in probability judgment. For example, the sum of subjective probabilities for a person dying from “heart disease, cancer, or other natural causes” tends to be judged greater than the subjective probability for the same person dying simply from ‘‘natural causes”. A lot of research demonstrated that the unpacking effect is not only happened in evaluation of lay people but also founded in experts’ judgments. Although Rottenstreich and Tversky (1997) highlighted that probability judgments are generally subadditive rather than superadditive and examples of superadditivity represent the exception rather than the rule of probability judgment, other researchers revealed that superadditivity is also widespread in probability judgment. Next we summarized the unpacking effect in probability judgment. Researchers showed that the unpacking effect is robust, not culturally influenced, in a wide range of field such as economy, medical treatment and trial. A lot of moderating variables also are explored. Research indicated that when using frequency rather than probability to represent results, the unpacking effect will be alleviated. Other important moderating variables include the nature of partition, temporal distance and the number of partition. Besides, the controversy of subadditivity and superadditivity is still no final conclusion. Considering the similarity between probability and time in human decision-making, it can be speculated that such an unpacking effect could be detected in time judgments. A large amount of research confirmed this conjecture. Researchers found that the unpacking effect can weaken plan fallacy and the difficulty of task is a moderating variable. The unpacking effect in time judgment also exists in many situations such as consumption and travel planning. Some factors like construal level and typicality of events could moderate the unpacking effect in time judgment. Although the original research concerning the unpacking effect dealt exclusively with probability judgments, there is reason to believe that the unpacking effect is not unique to judgments of probability. Just as time judgment, researchers revealed this effect in a lot of other social judgment. The unpacking effect plays an important role in contribution judgment in a team and it can reduce people’ egocentrism. Unpacking life into different domains can also improve human whole life satisfaction. And the unpacking effect is also detected in the field of public policy, negotiation, political and emotion predict. Such results indicate that the unpacking effect reflects a general characteristic of human judgment. At last, based on previous research, we put forward some future directions for research of the unpacking effect in decision making and judgment. Future researchers need to clarify the question that unpacking leads to subadditivity or superadditivity. More studies are needed to move a step forward to understand the impact of some moderating variables on the unpacking effect. In addition, to improve our understanding of the cognitive mechanism underlying the unpacking effect, researchers should pay more attention on the processes of the unpacking effect during decision making and judgment.

Key words: unpacking effect, support theory, subadditivity, superadditivity, probability judgment

摘要： 分解效应是人类主观判断中的一种较稳固的行为偏差，并且判断结果会对随后的决策产生重要影响，因此，对该领域研究成果的全面梳理具有重要的理论意义和实际价值。本文主要介绍了支持理论中的分解效应，总结梳理了概率判断与时间判断中的分解效应研究，综述了其他社会判断中的分解效应研究成果，展望了决策与判断中的分解效应的未来研究方向。通过上述内容的论述，以期为该领域研究提供新的思路，推动国内相关领域研究的发展。

关键词: 分解效应, 支持理论, 次可加性, 超可加性, 概率判断