Association for Behavior Analysis International

People discount the value of delayed or uncertain outcomes. That is, the subjective values of delayed and uncertain outcomes decrease depending on how delayed or how unlikely they are. How steeply individuals discount is thought to reflect how impulsive they are. From this perspective, steep discounting of delayed outcomes (which fails to maximize long-term welfare) and shallow discounting of probabilistic outcomes (which fails to adequately take risk into account) reflect similar decision-making processes and the same trait of impulsivity. However, several manipulations differentially affect delay and probability discounting, and correlational studies show that how steeply one discounts delayed rewards is relatively independent of how steeply one discounts probabilistic rewards. Thus, calling them both "impulsivity" may serve only to indicate that they can both cause people real problems. To study such problems, animal models frequently are needed, and indeed, animals' discounting seems similar to human discounting, although there are differences, some more apparent than real. People reportedly discount delayed rewards on a time scale of weeks or months rather than seconds as in animals, but when choices involve real, consumable rewards, people are as impatient as animals. Interestingly, correlational analyses reveal people's discounting of real, consumable rewards on a small time scale and hypothetical money on a large time scale are uncorrelated, suggesting that even within delay discounting, there may be multiple "impulsivities," each of which is important for understanding a different aspect of decision-making.

Delayed reward discounting (DRD) is a behavioral economic index of impulsivity. Numerous studies reported more impulsive DRD in groups exhibiting addictive behavior compared to control groups. This presentation will synthesize findings across the literature using a meta-analysis of these studies. Specifically, the meta-analysis will characterize: (a) overall patterns of findings, qualitatively and quantitatively; (b) systematic variability by sample and study type; and, (c) possible small study (publication) bias. Literature reviews identified 310 candidate articles from which 47 studies reporting 67 comparisons were identified (total N = 56,013). From the total comparisons identified, a highly statistically significant but small magnitude effect was evident with extremely high heterogeneity of effect size. Based on systematic observed differences, large studies assessing DRD with a small number of self-report items were removed and an analysis of 60 comparisons (n = 3,329) using equivalent methods and exhibiting acceptable heterogeneity revealed a highly significant medium magnitude effect. Further analyses revealed significantly larger effect sizes for studies using clinical samples (d = .62) compared to studies using nonclinical samples (d = .44). Indices of small study bias among the various comparisons suggested varying levels of influence by unpublished findings, ranging from minimal to moderate. These results reveal consistent evidence of more impulsive DRD in individuals exhibiting addictive behavior that is of medium effect size in general and significantly larger in individuals who meet criteria for an addictive disorder. Moreover, these findings suggest that multi-item assessments of DRD that offer more precise estimates of discounting are more sensitive to group differences. However, the studies reviewed use cross-sectional designs that do not permit a clear inference of whether these differences are causative (etiological) or consequential (symptomatic) of addictive behavior. To address this, several lines of research will be reviewed, including: (a) longitudinal evidence of DRD predicting addictive behavior; (b) evidence of the temporal stability of DRD; (c) evidence of DRD predicting clinical outcomes; (d) evidence from animal studies bearing on etiology; and (e) evidence of genetic influences on DRD and related phenotypes. On balance, most of the studies in these areas suggest that DRD plays an etiological role but there is also evidence these areas suggest that DRD plays an etiological role but there is also evidence that DRD is further exacerbated by drug abuse and aspects of addiction, such as withdrawal and craving. A potentially interactive and recursive process will be discussed.

A human delay discount function measures overlap between the interests of a person now and that same person's interest later; the steeper a person's delay discount function is, the more impulsive (less restrained) she is. A human social discount function measures overlap between the interests of a person now and those of another person now; the steeper a person's social discount function is, the more selfish (less altruistic) he is. Just as delay discount functions are obtained by asking people to choose between smaller rewards for themselves now and larger rewards for themselves at later times, social discount functions are obtained by asking people to choose between smaller rewards for themselves now and larger rewards for other people now at various social distances. In typical game theory analyses, each player is assumed to value only immediate or delayed rewards with delayed rewards being discounted. Looking at games from the viewpoint of social discounting, benefits from immediate and discounted delayed rewards to the player would be supplemented by discounted rewards to the other player or players. Given social discounting, degree of cooperation by Player-A would be predicted to increase as a function of the amount of reward to the other players or the number of other players receiving benefits (all else held constant), consequent on cooperation. This prediction was confirmed in experimental tests.

Leonard Green, James MacKillop, and Howard Rachlin will participate in a panel discussion and answer questions on the topics presented in Cluster 3.

Session Moderator: Amy Odum.

available onsite