Association for Behavior Analysis International

Delay and probability discounting refer to the devaluation of an outcome as a function of the delay or probability with which the outcome occurs. Discounting rates inform researchers how people make decisions and have been shown to differ between individuals with unhealthy patterns of behavior and control groups. But, everyday choices are not devoid of context and many questions remain about what choice is sensitive to across contexts. This series of talks examines how discounting can be influenced by one’s previous history and current state of access to the commodity under consideration. In addition, these talks highlight novel discounting methodology and data analytic approaches that may allow researchers to better understand what discounting looks like across different groups. Populations discussed will be women with food insecurities, MTurk participants, and the classic undergraduate student. Topics discussed will be amount/delay sensitivity to alcohol; how the size of one’s bank account and intermixing gains and losses influences discounting; and classic population comparisons in discounting rates. Together, the talks that comprise this symposium highlight advances in: methods for measuring discounting, group differences in discounting, and analytic techniques to understand discounting.

Graduate students in behavior analysis; Behavior analysts working in feeding; behavior analysts working with substance abuse populations

Food insecurity, or inconsistent access to foods that meet nutritional needs, has been linked to delay discounting for money, a behavioral measure of impulsivity. The purpose of the present study was to examine differences in delay and probability discounting for food and money outcomes in women with varying levels of food insecurity. Ninety-two women were recruited from a community sample. Participants completed delay and probability discounting tasks for food and money, food security measures, and measures of demographic and health-related variables. Results revealed that food insecure individuals showed significantly higher rates of delay discounting for both food and money compared to food secure individuals. No significant differences were found with probability discounting. These results indicate food insecure individuals are relatively more impulsive for food and money than those of food secure status but not necessarily different in risk-taking. In addition, impulsive behavior may be an underlying mechanism of obesity-related health outcomes in food insecure populations.

Heavy alcohol use on college campuses is a major concern, as it can result in academic problems, assault, alcohol use disorders, and death. One hallmark of excessive drinking is elevated discounting rates of alcohol compared to money. Differences in discounting rates could be due to differences in sensitivity to magnitude of rewards and/or delay. Recent analytic techniques may provide a better understanding of these behavioral mechanisms, and their effects on choice. We used the multilevel modeling technique to evaluate discounting of alcohol and money with undergraduate college students. This approach examines group and individual discounting rates simultaneously using a multilevel logistic regression and nonlinear models. It has been recommended for use, as it is sensitive to magnitudes of reward and relative delay, which are considered separate contributors to choice. We found subjects had differential sensitivity to reward magnitude and delay when discounting of alcohol compared to money. Findings suggest these analytic techniques may lead to the enhanced understanding of these behavioral processes needed to successfully target and decrease alcohol consumption on college campuses.

Previous studies have investigated how drastic changes to income influence delay discounting (e.g. Bickel et al., 2016). Additionally, while researchers have investigated how choices change as a function of the absolute magnitude of an outcome (e.g. Green et al., 1999), no studies have investigated how choices change as a function of magnitude relative to the amount in a participant’s bank. We investigated how endowing participants with a bank of varying amounts influenced choice within delay discounting tasks. 60 undergraduate psychology students completed 4 delay discounting tasks. Each task presented one of two different amounts for the delayed choice alternative (i.e., $3000 and $500,000). In addition, each task presented one of two relative bank amounts (i.e., 1/4 or 4 times the delayed amount). Discounting of the delayed amount was significantly greater when the bank amount was smaller than the delayed amount (e.g. $750 bank; $3,000 delayed amount) only in the small delayed amount ($3,000) condition. When the delayed outcome magnitude was large ($500,000), the bank amount did not have a statistically significant effect on discounting. Our results suggest contrived economic context influences participants’ choices under conditions in which an outcome is delayed.