Data from: Are capuchin monkeys (Sapajus spp.) sensitive to lost opportunities? The role of opportunity costs in intertemporal choice

Principles of economics predict that the costs associated with obtaining rewards can influence choice. When individuals face choices between a smaller, immediate option and a larger, later option, they often experience opportunity costs associated with waiting for delayed rewards because they must forego the opportunity to make other choices. We evaluated how reducing opportunity costs affects delay tolerance in capuchin monkeys. After choosing the larger option, in the High cost condition subjects had to wait for the delay to expire, whereas in the Low cost different condition they could perform a new choice during the delay. To assess the effect of intake rate on choices, the Low cost same condition had the same intake rate ratio as the High cost condition. We found that capuchins attended both to intake rates and to opportunity costs. They chose the larger option more often in the Low cost different and Low cost same conditions than in the High cost condition, and more often in the Low cost different condition than in the Low cost same condition. Understanding how non-human primates represent and use costs in making decisions not only helps to develop theoretical frameworks to explain their choices but also addresses similarities with and differences from human decision making. These outcomes provide insights into the origins of human economic behaviour.Each row represents a single trial. Responses are coded in the choice column where 0 represents the smaller, sooner option and 1 represents the larger, later option. Please see README.txt. Funding provided by: Istituto di Scienze e Tecnologie della Cognizione, Consiglio Nazionale delle Ricerche, Rome*Crossref Funder Registry ID: Award Number:These data were collected in an experiment with 10 adult capuchin monkeys, belonging to four social groups, housed at the Primate Center of the Consiglio Nazionale delle Ricerche, Rome between July and November 2018. The data were generated by a custom-built computer program that recorded each trial as a row. We processed the raw data by doing the following: Removing extraneous columns Renaming columns Fixing two incorrectly input responses Fixing two incorrectly input trial numbers Removing forced choice trials Removing extra trials from subjects who were tested beyond stability Recoding factor label