期刊名称:Proceedings of the National Academy of Sciences
印刷版ISSN:0027-8424
电子版ISSN:1091-6490
出版年度:2013
卷号:110
期号:20
页码:8308-8313
DOI:10.1073/pnas.1220738110
语种:English
出版社:The National Academy of Sciences of the United States of America
摘要: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.