Altering the Shape of Punishment Distributions Affects Decision Making in a Modified Iowa Gambling Task

Abstract

Neuroeconomics research has shown that preference for gambling is altered by the statistical moments (mean, variance, and skew) of reward and punishment distributions. Although it has been shown that altered means can affect feedback-based decision making tasks, little is known if the variance and skew will have an effect on these tasks. To investigate, we systematically controlled the variance (high, medium, and low) and skew (negative, zero, and positive) of the punishment distributions in a modified version of the Iowa Gambling Task. The Iowa Gambling Task has been used extensively in both academic and clinical domains to understand decision making and diagnose decision making impairments. Our results show that decision making can be altered by an interaction of variance and skew. We found a significant decrease over trials in choices from the decks with high variance and asymmetrically skewed punishments and from the decks with low variance and zero skew punishments. These results indicate that punishment distribution shape alone can change human perception of what is optimal (i.e., mean expected outcome) and may help explain what guides our day-to-day decisions.

Post Learning Sleep Improves Cognitive-Emotional Decision-Making: Evidence for a ‘Deck B Sleep Effect’ in the Iowa Gambling Task

“…To ensure participants were well-rested, and to investigate whether self-reported sleepiness levels changed across sessions and between conditions, the Epworth Sleepiness Scale (ESS) was filled out at the beginning of each session [32] . Verbal instructions were given, as previously reported in Seeley et al [33] and the task was administered for 200 trials, after which the participant was told the task was over. The task began with $2000 in virtual money.…”