A computational framework for the study of confidence across species

Confidence judgments, the self-assessment of the quality of a subject's knowledge, are considered a central example of metacognition. Prima facie, introspection, and self-report appear the only way to access the subjective sense of confidence or uncertainty. Could confidence be also studied in nonhuman animals so one could probe its neural substrates? Indeed, behavioral paradigms that incentivize animals to evaluate and act upon their own confidence can yield implicit reports of confidence. Here, we suggest that a computational approach can clarify the issues involved in interpreting these tasks and provide a much-needed springboard for advancing the scientific understanding of confidence. We first review relevant theories of probabilistic inference and decision making. We then critically discuss behavioral tasks employed to measure confidence in animals and show how quantitative models can help to constrain the computational strategies underlying confidence-reporting behaviors. In our view, post-decision wagering tasks with continuous measures of confidence appear to offer the best available metrics of confidence. Since behavioral reports alone provide a limited window into mechanism, we argue that progress calls for measuring the neural representations and identifying the computations underlying confidence reports. We present a case study using such a computational approach to study the neural correlates of decision confidence in rats. This work shows that confidence assessments may be considered higher order, but can be generated using elementary neural computations that are available to a wide range of species. Finally, we discuss the relationship of confidence judgments to the broader behavioral uses of confidence and uncertainty.