Rodent models of adaptive value learning and decision-making

Abstract

Real-world decisions are rarely as straightforward as choosing between clearly “good” vs. “bad” options. More often, options must be evaluated carefully because they differ in relative value. For example, we typically learn about (and make decisions between) options in comparison, where one outcome may be more costly or risky than the other. Several neuropsychiatric conditions are characterized by atypical evaluation of effort and risk costs, including major depression, schizophrenia, autism, obsessive-compulsive disorder, and substance use disorders. Aberrant value learning and decision-making have long been considered a cognitive-behavioral endophenotype of these disorders and can be modeled in rodents. This chapter presents two general methodological domains that the experimenter can manipulate in animal decision-making tasks: Risk and effort. Here, we present detailed methods of rodent tasks frequently employed within these domains: Probabilistic reversal learning (PRL) and effort choice. These tasks recruit regions within rodent frontal cortex, the amygdala, and the striatum, and performance is heavily modulated by dopamine, making these assays highly valid measures in the study of behavioral and substance addictions, in particular.