The effect of episodic future thinking on delay discounting

Abstract

Delay discounting refers to the fact that rewards lose their value if they are delayed. Excessive delay discounting is associated with various health-related problems such as over-eating and substance abuse. One phenomenon shown to reduce delay discounting is Episodic Future Thinking (EFT; imagining personal future events). Across multiple experiments and a meta-analysis, the current thesis examined the reliability of the effect of EFT on delay discounting and also sought to clarify the components of EFT that are necessary to reduce delay discounting.

Experiment 1 replicated the EFT effect using a common titrating-amount procedure, and the meta-analysis based on 40 independent studies confirmed that EFT has a reliable, medium-sized effect on delay discounting. The meta-analysis also assessed the robustness of the EFT effect across various methodological features and participant characteristics. A multiple meta-regression revealed that the between-study variability in the size of the EFT effect was accounted for by study design and type of discounting measure. Within-subjects design studies had significantly smaller effect sizes than between-subjects design studies. Studies that used k as a discounting rate measure had significantly smaller effect sizes than studies that used area under the discounting curve or other atheoretical discounting measures. The size of the EFT effect was robust across different participant characteristics and ages, suggesting that EFT may be utilized as an effective intervention for a variety of age groups and impulsive behaviors.

Experiments 2A, 2B and 2C assessed the suitability of an alternative and briefer delay discounting procedure (in which participants simply report their indifference points) for subsequent EFT experiments. The report indifference points procedure produced a high percentage of invalid data, a finding that was replicated across all three experiments. We therefore continued to use the well-established titrating-amount procedure in subsequent studies.

Experiment 3A found that episodic past thinking (imagining personal past events) and semantic future thinking (estimating what a person could buy with the larger, delayed amount) had no effect on discounting, indicating that both episodic thinking and future thinking are necessary components of the EFT effect. Experiment 3A results also indicated that financial relevance alone is not sufficient to reducing discounting. Experiment 3B found that the future events also need to be personally relevant to reduce discounting, and that participants perceived EFT to reduce their discounting by primarily making the larger, delayed reward more valuable. We also showed that demand characteristics, where participants change their behavior to conform to the researcher’s expectations, are an unlikely explanation for the EFT effects found in Experiments 3A and 3B. Further research is warranted to form a better understanding of the mechanism(s) through which EFT reduces delay discounting.