A Measure of Delay Discounting Within the Academic Domain

Abstract

It is important to better understand the decision-making processes involved in student procrastination, in order to develop interventions that reduce this common problem. Students may procrastinate because studying produces delayed reinforcers; however, no task measuring delay discounting of academic outcomes currently exists. In Experiment 1, we developed and piloted a measure of academic discounting modeled on titrating-amount tasks successfully used in the discounting literature. Participants made hypothetical choices between working for money (the smaller, sooner reinforcer) and working on an assignment that was due at various times (the larger, later reinforcer). Participants showed systematic decreases in the subjective value of the assignment as a function of delay, and the hyperbolic and hyperboloid models described the shape of this decrease in value well. In general, larger delayed rewards are discounted less steeply than smaller delayed rewards (the magnitude effect). In Experiment 2, we observed the magnitude effect in academic discounting: Participants discounted a "not important" assignment more steeply than an "important" assignment. In the hyperboloid model, this change was captured by an increase in the s parameter. Results provide support for the validity of the academic discounting task.

The effect of “should” and “would” instructions on delay discounting of rewards for self and others

“…(2018) did not examine whether the size of the larger, delayed reward moderated the number of unsystematic discounting functions in their meta-analysis, but it is well demonstrated that reward value affects discounting rate (the magnitude effect, e.g., Olsen et al, 2018). Note, however, that if the undiscounted value of $200 is higher for workers located in India, the magnitude effect would predict shallower discounting for that group, whereas we observed steeper discounting.…”

“…Yi et al (2017) found when participants completed the control-type condition before they completed the experimental condition, there was a significant difference in discounting rates between control and experimental conditions. However, when participants completed tasks in the reverse order, no significant differences were observed (see also Olsen et al, 2018, andYi, Pickover, Stuppy-Sullivan, Baker, &Landes, 2016, for similar patterns of effect). Therefore, we may also find that only participants who receive the "typical" decision context (i.e., YouWould) first will demonstrate subsequent differences in discounting rates.…”

“…Delay discounting is affected by condition x order interactions suggesting recent decision type affects subsequent choices (Malkoc & Zauberman, 2006;Malkoc, Zauberman, & Bettman, 2010;Olsen, Macaskill, & Hunt, 2018;Rung, Frye, DeHart, & Odum, 2019;Yi, Stuppy-Sullivan, Pickover, & Landes, 2017). People make a variety of different delay-amount trade-off decisions throughout day-to-day life, and order effects thus raise the intriguing possibility that everyday decision making is also affected by the decision frames an individual has recently encountered.…”

Understanding procrastination: A case of a study skills course

“…Escaping from stressful and aversive situations might relieve stress and are thus rewarding. As an example, students are always faced with a trade-off when choosing between procrastinating or studying (Kirby et al, 2005;Olsen et al, 2018). One alternative is to complete the challenging academic tasks on time which leads to delayed rewards in the form of achieving academic and career goals (see e.g., Sutcliff et al, 2019).…”

The effect of episodic future thinking on delay discounting

“…In contrast to the other atheoretical measures, AUC provides a measure of discounting which takes the value of all of the delays into account. The effects of experimental manipulations on delay discounting rate can differ depending on whether k or AUC is used (Olsen, Macaskill, & Hunt, 2018;Smith & Hantula, 2008). For example, Smith and Hantula (2008) found that k and AUC measures supported different conclusions about the combined effects of using different reinforcer magnitudes and discounting procedures.…”