The Impact of Experiencing Outcomes on Delay and Probability Discounting

Temporal and probability discounting refer to the decrease in subjective value of rewards that are either delayed or uncertain. Comparatively high degrees of discounting have been linked to other indices of impulsivity in both animals and humans, but the study of human discounting has primarily relied on participant reports of preferences between hypothetical outcomes. A number studies to date have illustrated not only how sensitive these preferences are to manipulations within such hypothetical procedures, but also that tasks where each consequence is actually experienced are able to capture changes in behaviour that hypothetical tasks may not. The current thesis examined temporal and probability discounting using a novel experiential discounting task in the form of a computer game-based discounting task in which the delays or probabilities of the preferred outcomes were experienced following each choice. The game had participants collect points by skiing over jumps, and discounting choices between either immediate and delayed or certain and uncertain point outcomes provided additional opportunities for point gain. Discounting was examined across four samples primarily composed of university students. Experiment 1, a test of the newly created task in two parts, assessed the most appropriate descriptive model of both temporal and probability discounting, as well as examined the relation between the two, hypothetical discounting and self-report measures of impulsivity and risk taking. Experiment 2 examined the effect of magnitude manipulations on both temporal and probability discounting. Experiment 3 examined the effect on degree of temporal discounting of manipulations to post-reward delays and access to alternative reinforcement during this period. Lastly, Experiment 4 attempted to manipulate the association between delay and risk and examined the impact of this on discounting of delayed outcomes. Participants across all studies showed systematic decreases in subjective value of both delayed and uncertain point outcomes. However, the shape of this decrease in value differed across the two types of outcomes, as did the effect of the magnitude manipulations. Furthermore, neither post-reinforcement duration, access to alternative reinforcement, nor experience with risk influenced degree of temporal discounting. The results are discussed in terms of single process accounts of discounting, the implications for individual trait interpretations of discounting and the relevance of experiential and hypothetical discounting to the construct of impulsivity.