Duke Institute for Brain Sciences, Duke University, Durham 27710, NC, USA; Center for Cognitive Neuroscience, Duke University, Durham 27710, NC, USA; Department of Psychology and Neuroscience, Duke University, Durham 27708, NC, USA.
Department of Statistical Science, Duke University, Durham 27708, NC, USA.
Cognition. 2021 Jul;212:104703. doi: 10.1016/j.cognition.2021.104703. Epub 2021 May 6.
Valence framing effects occur when participants make different choices or judgments depending on whether the options are described in terms of their positive outcomes (e.g. lives saved) or their negative outcomes (e.g. lives lost). When such framing effects occur in the domain of moral judgments, they have been taken to cast doubt on the reliability of moral judgments and raise questions about the extent to which these moral judgments are self-evident or justified in themselves. One important factor in this debate is the magnitude and variability of the extent to which differences in framing presentation impact moral judgments. Although moral framing effects have been studied by psychologists, the overall strength of these effects pooled across published studies is not yet known. Here we conducted a meta-analysis of 109 published articles (contributing a total of 146 unique experiments with 49,564 participants) involving valence framing effects on moral judgments and found a moderate effect (d = 0.50) among between-subjects designs as well as several moderator variables. While we find evidence for publication bias, statistically accounting for publication bias attenuates, but does not eliminate, this effect (d = 0.22). This suggests that the magnitude of valence framing effects on moral decisions is small, yet significant when accounting for publication bias.
当参与者根据选项是描述积极结果（例如拯救的生命）还是消极结果（例如失去的生命）来做出不同的选择或判断时，就会出现效价框架效应。当这种框架效应出现在道德判断领域时，人们对道德判断的可靠性产生了怀疑，并提出了这些道德判断在多大程度上是不言而喻或有其自身依据的问题。在这场争论中，一个重要因素是框架呈现方式的差异对道德判断的影响的程度和可变性。尽管心理学家已经研究了道德框架效应，但目前还不清楚已发表研究中汇总的这些效应的总体强度。在这里，我们对 109 篇已发表的文章进行了荟萃分析（总共涉及 146 个独特的实验，共有 49564 名参与者），这些文章涉及道德判断中的效价框架效应，结果发现，在被试间设计中存在中等效应（d=0.50），以及几个调节变量。虽然我们发现存在发表偏倚的证据，但对发表偏倚进行统计学处理后，虽然会减弱但不会消除这种效应（d=0.22）。这表明，在考虑到发表偏倚的情况下，效价框架效应对道德决策的影响程度虽然很小，但却具有统计学意义。
