Training diversity promotes absolute-value-guided choice.
Training diversity promotes absolute-value-guided choice.
复制标题
DOI：
10.1371/journal.pcbi.1010664
复制
发表时间：
2022-11
影响因子：
4.3
通讯作者：
中科院分区：
文献类型：
--
作者：
许多决策研究表明，人类在选择选项中学习期望值或相对偏好，但对什么环境条件促进一种策略而不是另一种策略知之甚少。在这里，我们测试新的假设，人类适应的程度，他们形成的绝对值的多样性的学习环境。由于绝对值更好地推广到新的选项集，我们预测，一个人了解的选项越多，他们形成绝对值的可能性就越大。为了测试这一点，我们设计了一个为期多天的学习实验，包括20个学习阶段，其中受试者选择成对的图像，每个图像与不同的奖励概率相关。我们通过要求受试者在他们在不同的会议中学习的图像之间进行选择来评估受试者形成绝对值和相对偏好的程度。我们发现，在一个会话中同时学习更多的图像增强了绝对值，抑制了相对偏好，学习。相反，在多个会话中将每个图像与大量其他图像累积并不影响学习的形式。这些结果表明，人类编码偏好的方式适应于即时学习环境提供的经验的多样性。学习一对选项之间的相对偏好对指导它们之间的选择是有效的，但可能会导致错误的选择之间没有配对，即使我们知道每个选项都很好。随着选项数量的增加，将相对偏好推广到新的决策环境的问题也会增加，因为选项越多，我们就越有可能在新的选项集合中遇到选择。为了解决这个问题，人们可以学习与每个选项相关的期望回报，也就是它的“绝对值”，通过它可以比较任何一对选项。因此，我们假设一个人了解的选项越多，他们就越有可能形成绝对值，而不是相对偏好。我们构建了一个新颖的多日奖励学习实验来专门验证这一假设。我们发现，同时学习更多的图像确实增强了绝对值学习，抑制了相对偏好学习。研究结果阐明了什么样的学习条件促进了可概括的偏好的形成，这些偏好可以帮助在不同的环境中做出最佳决策，这种能力在真实的世界中至关重要，因为在这个世界中，经验是有限的，并且在多个不断变化的环境中是支离破碎的。
Many decision-making studies have demonstrated that humans learn either expected values or relative preferences among choice options, yet little is known about what environmental conditions promote one strategy over the other. Here, we test the novel hypothesis that humans adapt the degree to which they form absolute values to the diversity of the learning environment. Since absolute values generalize better to new sets of options, we predicted that the more options a person learns about the more likely they would be to form absolute values. To test this, we designed a multi-day learning experiment comprising twenty learning sessions in which subjects chose among pairs of images each associated with a different probability of reward. We assessed the degree to which subjects formed absolute values and relative preferences by asking them to choose between images they learned about in separate sessions. We found that concurrently learning about more images within a session enhanced absolute-value, and suppressed relative-preference, learning. Conversely, cumulatively pitting each image against a larger number of other images across multiple sessions did not impact the form of learning. These results show that the way humans encode preferences is adapted to the diversity of experiences offered by the immediate learning context. Learning relative preferences between a pair of options is effective in guiding choice between them, but might lead to error in choosing between options that have not been paired against each other even if we know each option well. This problem of generalizing relative preferences to novel decision contexts increases as the number of options gets larger, since the more options there are the more likely we are to encounter choices between new sets of options. To solve this problem, people may learn the expected reward associated with each individual option—that is, its ‘absolute value’, by means of which any pair of options can be compared. Thus, we hypothesized that the more options a person learns about, the more likely they would be to form absolute values as opposed to relative preferences. We constructed a novel multi-day reward learning experiment to specifically test this hypothesis. We found that concurrently learning about more images indeed enhances absolute-value learning and suppresses relative-preference learning. The findings clarify what learning conditions promote the formation of generalizable preferences that can help reach optimal decisions across different contexts, an ability that is vital in the real world where experience is limited and fragmented across multiple continuously shifting contexts.
登录
查看更多内容
影响因子：
5.7
作者：
Bartra, Oscar;McGuire, Joseph T.;Kable, Joseph W.
通讯作者：
Kable, Joseph W.
影响因子：
16.6
作者：
Biderman N;Shohamy D
通讯作者：
Shohamy D
影响因子：
13.6
作者：
Bavard, Sophie;Rustichini, Aldo;Palminteri, Stefano
通讯作者：
Palminteri, Stefano
影响因子：
4.3
作者：
通讯作者：
--
影响因子：
64.8
作者：
Daw, Nathaniel D.;O'Doherty, John P.;Dayan, Peter;Seymour, Ben;Dolan, Raymond J.
通讯作者：
Dolan, Raymond J.