Department of Psychology.
Oxford Internet Institute.
J Exp Psychol Gen. 2023 Feb;152(2):464-482. doi: 10.1037/xge0001284. Epub 2022 Sep 1.
In 10 experiments, we investigated the relations among curiosity and people's confidence in their answers to general information questions after receiving different kinds of feedback: yes/no feedback, true or false informational feedback under uncertainty, or no feedback. The results showed that when people had given a correct answer, yes/no feedback resulted in a near complete loss of curiosity. Upon learning they had made an error via yes/no feedback, curiosity increased, especially for high-confidence errors. When people were given true feedback under uncertainty (they were given the correct answer but were not told that it was correct), curiosity increased for high-confidence errors but was unchanged for correct responses. In contrast, when people were given false feedback under uncertainty, curiosity increased for high-confidence correct responses but was unchanged for errors. These results, taken as a whole, are consistent with the region of proximal learning model which proposes that while curiosity is minimal when people are completely certain that they know the answer, it is maximal when people believe that they almost know. Manipulations that drew participants toward this region of "almost knowing" resulted in increased curiosity. A serendipitous result was the finding (replicated four times in this study) that when no feedback was given, people were more curious about high-confidence errors than they were about equally high-confidence correct answers. It was as if they had some knowledge, tapped selectively by their feelings of curiosity, that there was something special (and possibly amiss) about high-confidence errors. (PsycInfo Database Record (c) 2023 APA, all rights reserved).
在 10 项实验中，我们研究了在接收到不同类型的反馈后，好奇心与人们对一般信息问题答案的信心之间的关系：是/否反馈、不确定时的真/假信息反馈，或无反馈。结果表明，当人们给出正确答案时，是/否反馈几乎完全丧失了好奇心。当他们通过是/否反馈得知自己犯了错误时，好奇心增加了，尤其是对高置信度的错误。当人们在不确定的情况下得到真实的反馈（他们得到了正确的答案，但没有被告知这是正确的）时，高置信度的错误会增加好奇心，但正确的反应则不变。相比之下，当人们在不确定的情况下得到错误的反馈时，高置信度的正确反应会增加好奇心，但错误反应则不变。总的来说，这些结果与近端学习模型一致，该模型提出，当人们完全确定自己知道答案时，好奇心最小，而当人们认为自己几乎知道答案时，好奇心最大。那些让参与者接近“几乎知道”区域的操作会增加好奇心。一个意外的发现是（在这项研究中四次重复），当没有反馈时，人们对高置信度的错误比对同样高置信度的正确答案更感兴趣。就好像他们有一种知识，通过他们的好奇心选择性地挖掘出来，即高置信度的错误有一些特别的（可能有问题的）东西。（PsycInfo 数据库记录（c）2023 APA，保留所有权利）。
