Abstract
Probability discounting is loss in reinforcer value as a function of uncertainty. In typical tasks measuring probability discounting, participants repeatedly choose between a smaller, certain amount and a larger amount at one of several probabilities, and do not experience the outcome they select. Most participants show a gain–loss asymmetry, discounting gains more steeply than losses. We examined whether gain–loss asymmetry also occurred when participants received feedback about the number of points they had gained or lost following choices, and whether discounting rates changed as participants accrued experience of outcomes across the session. We assessed discounting of hypothetical monetary gains and losses twice using a mixed description-experience task. In these tasks, participants chose repeatedly between options that were described, where each choice was then followed by feedback on the number of points gained or lost from the chosen option, but not the unchosen option (i.e. a partial feedback procedure). Neither the first nor second task showed a consistent mean difference in discounting of gains and losses. Discounting in the first task was affected by an interaction of condition order and type of outcome (gains or losses). Furthermore, discounting of losses was more stable across task administrations than discounting of gains. Participants’ tendency to choose the probabilistic option was significantly correlated across conditions. Together, this highlights the crucial role of recent experience of outcomes on probability discounting of gains and losses.