Department of Psychology, University of York, York, YO10 5DD, UK.
Department of Archaeology, University of York, York, YO10 5GB, UK.
Sci Rep. 2024 Sep 30;14(1):22637. doi: 10.1038/s41598-024-72812-2.
Individuals living and working in dangerous settings (e.g., first responders and military personnel) make complex decisions amidst serious threats. However, controlled studies on decision-making under threat are limited given obvious ethical concerns. Here, we embed a complex decision-making task within a threatening, immersive virtual environment. Based on the Iowa Gambling Task (IGT), a paradigm widely used to study complex decision-making, the task requires participants to make a series of choices to escape a collapsing building. In Study 1 we demonstrate that, as with the traditional IGT, participants learn to make advantageous decisions over time and that their behavioural data can be described by reinforcement-learning based computational models. In Study 2 we created threatening and neutral versions of the environment. In the threat condition, participants performed worse, taking longer to improve from baseline and scoring lower through the final trials. Computational modelling further revealed that participants in the threat condition were more responsive to short term rewards and less likely to perseverate on a given choice. These findings suggest that when threat is integral to decision-making, individuals make more erratic choices and focus on short term gains. They furthermore demonstrate the utility of virtual environments for making threat integral to cognitive tasks.
个体在危险环境中生活和工作（例如，急救人员和军人）在严重威胁下做出复杂的决策。然而，由于明显的伦理问题，对威胁下的决策进行的对照研究是有限的。在这里，我们将一项复杂的决策任务嵌入到一个充满威胁的沉浸式虚拟环境中。基于广泛用于研究复杂决策的爱荷华赌博任务（IGT），任务要求参与者做出一系列选择以逃离正在倒塌的建筑物。在研究 1 中，我们证明，与传统的 IGT 一样，参与者随着时间的推移学会做出有利的决策，并且他们的行为数据可以通过基于强化学习的计算模型来描述。在研究 2 中，我们创建了威胁和中性两种环境版本。在威胁条件下，参与者表现更差，需要更长的时间才能从基线中提高，并且在最后的试验中得分更低。计算模型进一步表明，在威胁条件下的参与者对短期奖励更敏感，并且不太可能坚持给定的选择。这些发现表明，当威胁成为决策的一部分时，个体做出更不稳定的选择并关注短期收益。此外，它们还证明了虚拟环境对于将威胁纳入认知任务的实用性。
