Rats show adaptive choice in a metacognitive task with high uncertainty.

Metacognition refers to the use of one's cognitive processes to coordinate behavior. Many higher cognitive functions such as feeling-of-knowing judgment and theory of mind are thought to be metacognitive processes. Although some primate species also show this ability in the form of behavioral control, a rodent model of metacognition is required for advanced studies of this phenomenon at behavioral, molecular, and neural levels. Here we show that rats could reliably be trained in a metacognitive task. The rats were trained to remember the location of a nose-poke hole and later indicate the location via a behavioral task. Rats had options of either demonstrating their memory or switching to an easier task (escape). Four rats were used in a two-choice metacognitive task, and 3 were used in a six-choice task. In the six-choice task, rats increased the likelihood of receiving a reward by utilizing the option to escape, in exchange for a decrease in the amount of reward received per correct trial. Furthermore, rats escaped more in sample-omitted trials than in standard trials. Results are consistent with the hypothesis that rats have metacognition, and could be utilized as a benchmark for further metacognition studies in rats. However, rats in the two-choice task did not use the escape response adaptively. These results were consistent with those seen in capuchin monkeys. Similarity between rodents and primates in task switching should expand the possibility of comparative studies of metacognition. (PsycINFO Database Record