How does social context modulate risky decision-making in long-tailed macaques (Macaca fascicularis)? A research report on problematic issues with a repeated sampling from experience paradigm

Abstract

We compared the risk preferences of long-tailed macaques (Macaca fascicularis) in a social and a nonsocial condition, where we assessed their preference for a safe or a risky option after repeated information sampling. To this end, we devised a food dispenser that could run automatically or involve a human distributor. Only two of the initial set of ten monkeys reached the final test conditions; these two monkeys did not prefer a particular option in any of the conditions. In light of the unexpected difficulties that the monkeys had in learning the reward contingencies, we discuss a potential problem with paradigms that involve extracting statistical information from repeated sampling events. Given that such a paradigm has been used repeatedly and is still being used with different animals, we feel the field might benefit from a critical discussion of how to increase the validity of the data generated by this paradigm. We suggest that it is crucial to obtain a good estimate of the amount of sampling information a particular study population needs for optimal decision-making before any test conditions are run. We would have benefited greatly from the availability of more detailed reporting of similar difficulties in other populations, for example, information regarding timeframe and necessary amount of information on sampling events or previous failures with specific presentation formats. We hope this report can serve as a source of information for others working on this or related topics.