= 160) were recruited through Amazon Mechanical Turk (MTurk), a crowdsourcing Internet marketplace where small tasks are posted as human intelligence tasks (HITs) for human workers to complete. Only MTurk workers who had a 95% acceptance rate on previous HITs and who currently resided in the United States could access the survey. Thirty-one percent of the participants were minority, 48% were female, and the average age was 35.25 (SD = 9.42). All participants had graduated high school, with 79% having completed at least one year of college education. Participants were excluded from completing the survey if they endorsed tobacco or illicit substance use (e.g., opiates, stimulants) with the exception of marijuana, or if they were not at least 18 years old. There are two reasons why we decided to include marijuana users while not including other drugs of abuse. First, while drug use is typically associated with greater delay discounting, that is not the case with marijuana. For example, tobacco [17
], cocaine [19
], and opioid [20
] users have higher discounting than non-users. However, delay discounting does not differ between users and non-users of marijuana [21
], suggesting that including marijuana users would not influence our results. Second, our study relied on Amazon’s Mechanical Turk to collect online data from individuals across the United States. Currently, 22 states have decriminalized or legalized marijuana [22
], and we could have lost a significant number of potential participants if we excluded marijuana use, in addition to limiting the generalizability of our results. For these reasons, we did not exclude marijuana users from participation.
Individuals were told they were participating in a study that sought to examine factors that influence decision-making. Participants were paid $1.00 for completing the survey and awarded a $5.00 bonus if they followed instructions and carefully completed 100% of the survey questions. All subjects gave their informed consent for inclusion before they participated in the study. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the Social and Behavioral Sciences Institutional Review Board of the State University of New York at Buffalo (STUDY00002214).
2.2. Experimental Design and Procedures
Before participation, an overview of the study was provided. Eligible participants were randomized to one of four conditions in a 2 × 2 factorial design, with “EFT/ERT” and “matched/unmatched cue presentation” as the between-group variables. ERT controls for engaging in episodic memory and has the participant generate recent cues rather than future cues. Eligible participants completed one 45-min survey that included consent, cue creation depending on group assignment, a delay-discounting task, a demographics questionnaire, and future orientation measures. No deception was used in this research protocol.
In terms of episodic cue generation, depending on group assignment, participants generated either EFT or ERT cues. Within the EFT condition, participants were either assigned to an EFT-matched group, where the created cues later matched the time delays of the DD task, or to one of four EFT-unmatched groups, where the created cues were presented out of order to the time delays of the DD task (see Table 1
for cue presentation order). Regardless of EFT group assignment, there were no differences during the cue creation portion of the experiment. EFT groups were asked to list and describe future positive events that were going to take place, or could reasonably take place in 1 day, 1 week, 1 month, 6 months, and 2 years. Because positive events are more vivid than negative events, only positive events were used [23
]. Participants in the EFT groups listed one future event that they were looking forward to per time period. The future events were matched to the five time delays of the DD task. Participants rated the salience (importance), valence (enjoyment), arousal (exciting), and vividness of each event on scales from “1—not at all” to “5—very much”. For events that were rated low in vividness (less than 3), the survey asked the participant if a different, more vivid event was available within the same time frame. Vividness of cues is important because previous researchers have found that the vividness of EFT cues predicts reduced DD [23
The ERT groups were asked to list positive and enjoyable events that took place within the last 60 h (e.g., 12 h ago, 24 h ago, 36 h ago). Participants in the ERT-matched group created cues that were later matched to the time delays of the discounting task, while participants in the ERT-unmatched groups created cues that were later presented out of order to the way they would normally be paired with time delays of the DD task (i.e., sequentially starting with 12 h ago; see Table 1
for cue presentation order). All five groups first listed all past recent events for each specified time period (e.g., “12 h ago I enjoyed a cup of coffee at home”, “24 h ago I met my friend for lunch”). Salience, valence, arousal, and vividness (“1—not at all” to “5—very much”) ratings were collected for each event. Events that were rated low in vividness (less than 3) were queried to see if another, more vivid event for the same time period could be listed instead.
After listing events, both EFT and ERT groups were asked to provide further episodic details about each event. Specifically, participants were asked to imagine and describe who they were with, what they were doing, where they were, and how they were feeling at each event. These elaborations were incorporated into the original one-sentence event description to create five individualized cues that were roughly 3–6 sentences long. Before each of the five time delays of the DD task, the corresponding cue, depending on group assignment, was presented on the screen for participants to read out loud. After completing each of the trials of DD, participants were asked to rate each event on how frequently they thought of their event (“1—never” to “5—always”) and how vivid their event was (“1—not vivid at all” to “5— highly vivid”). An overall imagery score was calculated by averaging vividness and frequency scores into one variable that assessed the degree to which the participants thought of their episodic events during DD.
In terms of the delay discounting task, participants made choices between immediate and delayed hypothetical monetary rewards across five time delays. In each of the five delays, the immediate amounts were available now while delayed rewards were available 1 day, 1 week, 1 month, 6 months, and 2 years into the future. The value of the delayed reward remained fixed at $
100 while the value of the immediate reward was adjusted in specific dollar increments based on response, until an indifference point was reached. The indifference point represents the equivalence point between choosing the immediate reward and the future reward. Participants were prompted to vividly imagine their EFT or ERT cue during the DD task by reading their event out loud before each of the five time delays. The cue was presented in large font on the computer screen prior to each time delay for participants to read out loud and was also embedded within the DD task so participants could keep the event in mind as they made their monetary choices. A different cue was presented for each time delay: The EFT-matched group imagined future events that corresponded to the time delay of the DD task, while the EFT-unmatched groups imagined a different future event during each time delay (Table 1
). The ERT-groups imagined a different recent event during each time delay (Table 1
). Discounting of rewards in the DD task was determined by calculating area under the curve (AUC) values. AUC is a method of measuring DD by calculating the area under the empirical discounting function [24
]. Individual AUC scores can range from 0 (highest possible discounting) to 1 (no discounting). An attention check was embedded in the middle of the delay discounting task to ensure participants were reading instructions and making choices carefully. The question asked whether they would prefer $
0.00 now or $
100 in 6 months. If participants chose $
0.00 now, it was considered a failure to pay attention to the task, and they would be removed from data analyses. No participants failed the attention check.
In terms of demographics, participant demographic information was collected (e.g. sex, age, race/ethnicity, family income, and educational level).
In terms of time orientation measures, the Future Orientation Scale (FOS) assesses the extent to which individuals consider the potential future outcomes of their current behavior and the extent to which they prefer long-term, as opposed to short-term goals [25
]. The FOS was scored for three five-item subscales: time perspective (e.g., “Some people spend very little time thinking about how things might be in the future BUT Other people spend a lot of time thinking about how things might be in the future”); anticipation of future consequences (e.g., “Some people usually think about the consequences before they do something BUT Other people just act—they don’t waste time thinking about the consequences”); and planning ahead (e.g., “Some people like to plan things out one step at a time BUT Other people like to jump right into things without planning them out beforehand”). The respondent was asked to choose between the two descriptors and then rate whether it was “sort of true for me” or “really true for me”. Higher scores indicated higher future orientation. To assess for the time period most relevant to an individual’s finances, participants were asked the following question, “In planning your, or your family’s, saving and spending, which of the time periods is more important to you and your partner, if you have one?” The answer choices provided were categorical and ranged from not planning, planning for the next 6 months or less, the next year, the next 5 years, and the next 10 years or more. Financial planning was converted to a continuous variable for analysis using midpoints of the categories [26
]. Subjective probability of living to age 75 was measured by asking, “What do you think are the chances you will live to be 75 or more? (where 0 means there is no chance you will live to 75 or more, and 100 means you will definitely live to 75 or more)” [26
]. Higher values were indicative of greater future orientation. These measures of time orientation allowed for a comparison between groups and whether these measures moderated the effect of EFT on reducing DD.