中文摘要：

通过两个研究,损益值大小效应在1元～1亿元的广阔范围内,以及在只采用选择反应模式的条件下得到了进一步探讨,其机制也在齐当别模型的全新视角下得到了讨论;此外,齐当别模型本身以及模型的经典判断任务也得到了探讨。结果发现：基于选择反应模式时,损益值大小效应在获得情景中稳定存在,并只存在一个风险倾向的拐点,而它在损失情景中则不存在;损益值大小效应的机制能够被齐当别模型很好地解释;齐当别模型的经典判断任务对模型预测效能的检验力不足,但其与新设计的维度内判断任务联合后却有很大改善;齐当别模型对损失情景中的均含0结果值的决策任务的解释力还有待进一步探讨。

英文摘要：

Various researches showed that the magnitude of the outcome could affect risk preference in decision making,which was inconsistent with the common phenomenon of risk aversion in gain.Weber（2005） investigated this effect,but three points of their study were needed to be further discussed.First,the magnitude of the value was still in a narrow range;second,a lot of researches didn＇t showed the effect in loss condition by choice response paradigm,but it was not discussed very well;third,the mechanism underlying this effect had not been discussed perfectly up to now.So the current research tested the magnitude of outcome effect in a wider range of value and choice response conditions respectively.What＇s more,the underlying mechanism of the effect was discussed in an equate to differentiate way,and some problems about the equate to differentiate model itself were also discussed.Two experiments were conducted in the current research,chi square;phi and phi square were used for the data analyses.The first experiment included four conditions： risky gain vs.sure gain,risky gain vs.riskier gain,risky loss vs.sure loss and risky loss vs.riskier loss.Each condition included four tasks that had the same risking probability and different magnitude of the value.The magnitude of the value in all tasks varied from 1 Yuan to 100000000 Yuan.135 undergraduate students participated in all of the four conditions of this experiment,but they only finished tasks with a different kind of value in each condition.After they finished each task,the subjects were asked to finish a traditional judgment task which was designed to test for whether the equate to differentiate model could give an interpretation to the magnitude of outcome effect or not.The second study was almost as the same as the first experiment.The only difference between them was that,a new judgment task was designed for the equate to differentiate model in the second study.123 undergraduate students took part in the second experiment.The results of the two experiments