Contents lists available at ScienceDirect Behavioural Processes journal homepage: www.elsevier.com/locate/behavproc Comparison of potentially real versus hypothetical food outcomes in delay and probability discounting tasks Stephen H. Robertson, Erin B. Rasmussen ⁎ Department of Psychology, Idaho State University, Pocatello, ID 83209-8812, United States ARTICLE INFO Keywords: Delay discounting Food Food discounting Impulsivity Potentially real outcome ABSTRACT Much of the research on human delay and probability discounting involves the use of hypothetical outcomes, in which participants indicate preferences for outcomes but do not receive them. Research generally shows that hypothetical and potentially real outcomes are discounted at similar rates. One study, however, shows that potentially real cigarettes are discounted more steeply than hypothetical cigarettes in smokers, calling into question the generality of the finding that potentially real and hypothetical money are discounted at similar rates. Using a within-subject design, we tested the extent to which potentially real and hypothetical monetary (Experiment 1) and food-related (Experiment 2) outcomes were discounted at similar rates. We found mixed results for monetary outcomes, in that potentially real outcomes were discounted more steeply than hypothetical outcomes when all participants were included; however, this effect disappeared when only systematic re- sponders were used. In addition, potentially real and hypothetical monetary outcomes were significantly cor- related. For food-related outcomes, we found robust and consistent effects that potentially real and hypothetical food outcomes are discounted similarly and that they correlate strongly. Generally, these findings suggest that using hypothetical outcomes generate similar levels of discounting, in particular for food, which is useful for researchers interested in characterizing food-related impulsivity. 1. Introduction Delay discounting refers to the tendency for an outcome to lose value as the delay to its receipt increases (Ainslie, 1975; Madden and Johnson, 2010; Rachlin, 1995) and is considered a behavioral process that is a facet of impulsivity (Bickel et al., 1999; Bickel and Marsch, 2001; Green et al., 1994). Relative to controls, higher levels of delay discounting (i.e., impulsivity) have been observed in cigarette smokers (e.g., Bickel et al., 1999), cocaine-dependent (e.g., Heil et al., 2006), heroin-dependent (e.g., Kirby et al., 1999), and obese individuals (e.g., Fields et al., 2011; Hendrickson and Rasmussen, 2013; Jarmolowicz et al., 2014; Rasmussen et al., 2010; Weller et al., 2008). Given that excessive delay discounting is associated with a wide-range of health- related conditions, it is considered a trans-disease process (Bickel and Mueller, 2009; Bickel et al., 2012). Delay discounting is assessed by presenting choices between a smaller outcome that is immediately available vs. a larger outcome available after a delay (e.g., “Would you prefer $9 now or $10 in a day?” Most choose the large amount to this question). A pattern of preferences for smaller, sooner outcomes are considered impulsive and a pattern of preferences for larger, later outcomes are considered self- controlled (Bickel and Mueller, 2009; Bickel et al., 2012). Delayed outcomes also inherently include the property of un- certainty. For example, if a person chooses $100 after a year, there is a question of whether the outcome will still be available after this period of time elapses. Therefore, some researchers conceptualize probability discounting as a separate, though related process to delay discounting (Green et al., 1999; Holt et al., 2003; Myerson et al., 2003). Probability discounting, which measures sensitivity to uncertainty, refers to the extent to which an outcome loses its value as the odds against receiving that outcome increase. In measuring probability discounting, an individual is asked to make choices between smaller, certain outcomes vs. larger, less prob- abilistic outcomes (e.g., “Would you prefer $3 for sure or $10 with a 50% chance of receiving that outcome”). Individuals who consistently prefer less probabilistic outcomes are characterized as risk averse and those who prefer them are considered risky (Green et al., 1999; Estle et al., 2007; Lawyer et al., 2011; Rasmussen et al., 2010). Typically in both delay and probability discounting studies using humans, hypothetical outcomes are used for assessing choices; that is, participants do not actually receive the outcome associated with the choice (Madden et al., 1997; Odum et al., 2006; Rasmussen et al., https://doi.org/10.1016/j.beproc.2018.01.014 Received 7 August 2017; Received in revised form 8 November 2017; Accepted 18 January 2018 ⁎ Corresponding author. E-mail address: rasmerin@isu.edu (E.B. Rasmussen). Behavioural Processes 149 (2018) 8–15 Available online 31 January 2018 0376-6357/ © 2018 Published by Elsevier B.V. T