Assessing reinforcement learning and vigor with an online decision-making battery

Reinforcement learning and vigor are two fundamental aspects of decision-making modulated by dopamine. Although vigor is highly relevant for mood disorders, it has been overlooked by previous batteries characterizing decision-making abilities. We developed a test battery with two RL tasks (RL/working memory, go/no-go task) and two vigor tasks (physical effort discounting, foraging task) and administered it to an online community sample (N = 162). We used computational modeling and factor analysis to identify two factors reflecting RL and vigor. The RL factor reflected participants' general decision-making ability, while vigor corresponded to a motivational propensity to obtain rewards. Both factors demonstrated moderately strong test-retest reliability (ICC(2,1) = .65 and .67, respectively) over a one-to-two-week interval. Moreover, both factors were meaningfully related to self-report scales. The RL and vigor factors were negatively and positively associated with mania, respectively, although further analyses revealed that vigor was not fully independent of RL. Neither factor was significantly associated with depression symptoms. Collectively, we present preliminary evidence that vigor may represent a separate factor from RL in line with the putative dopamine hypothesis, paving the way for future longitudinal assessments of RL and vigor in mood disorders.