Department of Psychology, Wesleyan University, United States.
Department of Psychology, Wesleyan University, United States.
Cogn Psychol. 2020 May;118:101273. doi: 10.1016/j.cogpsych.2020.101273. Epub 2020 Feb 3.
Performance on an intuitive symbolic number skills task-namely the number line estimation task-has previously been found to predict value function curvature in decision making under risk, using a cumulative prospect theory (CPT) model. However there has been no evidence of a similar relationship with the probability weighting function. This is surprising given that both number line estimation and probability weighting can be construed as involving proportion judgment, that is, involving estimating a number on a bounded scale based on its proportional relationship to the whole. In the present work, we re-evaluated the relationship between number line estimation and probability weighting through the lens of proportion judgment. Using a CPT model with a two-parameter probability weighting function, we found a double dissociation: number line estimation bias predicted probability weighting curvature while performance on a different number skills task, number comparison, predicted probability weighting elevation. Interestingly, while degree of bias was correlated across tasks, the direction of bias was not. The findings provide support for proportion judgment as a plausible account of the shape of the probability weighting function, and suggest directions for future work.
在风险决策下，使用累积前景理论（CPT）模型，人们发现，在直观的符号数字技能任务（即数轴估计任务）上的表现可以预测价值函数的曲率。然而，并没有证据表明与概率加权函数有类似的关系。这令人惊讶，因为数轴估计和概率加权都可以被理解为涉及比例判断，即根据数量与整体的比例关系，在有界的尺度上估计数量。在本工作中，我们通过比例判断的视角重新评估了数轴估计和概率加权之间的关系。使用具有双参数概率加权函数的 CPT 模型，我们发现了双重分离：数轴估计偏差预测了概率加权曲率，而在不同的数字技能任务（数字比较）上的表现则预测了概率加权提升。有趣的是，虽然偏差程度在任务之间是相关的，但偏差的方向却不是。这些发现为比例判断作为概率加权函数形状的一种合理解释提供了支持，并为未来的工作提供了方向。
