Assessing susceptibility of a temporal discounting task to faking.

ABSTRACT

OBJECTIVE: Temporal discounting describes the devaluation of delayed rewards. Because temporal discounting is predictive of substance misuse, its clinical assessment could improve prevention (e.g., identifying at-risk youth) and treatment (e.g., predicting relapse). However, if discounting rates can be faked (e.g., to avoid treatment), their clinical utility may be limited. For the first time, we measured the influence of deception in a temporal discounting task. METHOD: We recruited 200 participants (44% female, Mage= 33) through Amazon Mechanical Turk. Participants completed a discounting assessment with instructions to (a) respond honestly, (b) fake good (i.e., simulate better versions of themselves), or (c) fake bad. RESULTS: Generalized linear mixed effects analysis showed that in Experiment 1, nonclinical samples faked good ( Mhyperbolic discounting rate= 0.002) or bad ( M= 0.086) compared to the Honest group ( M= 0.008). In Experiment 2, cigarette smokers faked good ( M= 0.003) compared to the Honest group ( M= 0.025). CONCLUSIONS: Temporal discounting transects the disciplines of psychology, biology, and behavioral economics. Before its promise as an endophenotype can be realized, assessments must be translated for clinical use. Opaquer temporal discounting tasks, or secondary measures of lying, may be required before temporal discounting can be confidently extended to clinical settings.