Graduate Program in Neuroscience, University of Minnesota, Minneapolis, MN 55455, USA.
Proc Natl Acad Sci U S A. 2013 May 14;110(20):8308-13. doi: 10.1073/pnas.1220738110. Epub 2013 Apr 29.
Laboratory studies of decision making often take the form of two-alternative, forced-choice paradigms. In natural settings, however, many decision problems arise as stay/go choices. We designed a foraging task to test intertemporal decision making in rats via stay/go decisions. Subjects did not follow the rate-maximizing strategy of choosing only food items associated with short delays. Instead, rats were often willing to wait for surprisingly long periods, and consequently earned a lower rate of food intake than they might have by ignoring long-delay options. We tested whether foraging theory or delay discounting models predicted the behavior we observed but found that these models could not account for the strategies subjects selected. Subjects' behavior was well accounted for by a model that incorporated a cost for rejecting potential food items. Interestingly, subjects' cost sensitivity was proportional to environmental richness. These findings are at odds with traditional normative accounts of decision making but are consistent with retrospective considerations having a deleterious influence on decisions (as in the "sunk-cost" effect). More broadly, these findings highlight the utility of complementing existing assays of decision making with tasks that mimic more natural decision topologies.
实验室研究中的决策制定通常采用二选一、强制选择的范式。然而,在自然环境中,许多决策问题表现为停留/离开的选择。我们设计了一个觅食任务,通过停留/离开的决策来测试大鼠的跨期决策。实验对象并没有遵循只选择与短延迟相关的食物项目的最大化收益策略。相反,大鼠经常愿意等待相当长的时间,因此,与忽略长延迟选项相比,它们的食物摄入量会降低。我们测试了觅食理论或延迟折扣模型是否可以预测我们观察到的行为,但发现这些模型无法解释实验对象选择的策略。实验对象的行为可以很好地用一个模型来解释,该模型考虑了拒绝潜在食物项目的成本。有趣的是,实验对象的成本敏感性与环境丰富度成正比。这些发现与传统的决策制定规范解释相矛盾,但与回顾性考虑对决策产生不利影响(如“沉没成本”效应)相一致。更广泛地说,这些发现强调了用更自然的决策拓扑结构任务来补充现有的决策制定测试的效用。