Abstract
Does the amount of self-control consumers must exert to choose a goal-consistent action influence their perceptions of goal progress? For example, if you choose to go to the gym when one of your favorite TV shows is on (vs. when nothing interesting is on TV), do you perceive that you have made a differential amount of progress toward your goal, despite completing the exact same workout? In eight studies (N = 7,515), the authors demonstrate that consumers perceive that they have made more progress on their goals when more (vs. less) self-control is required to choose to complete an identical goal-consistent task. This is because when consumers exert more (vs. less) self-control to choose a goal-consistent task over the goal-inconsistent alternatives, they infer higher commitment to the goal. The higher inferred commitment, in turn, leads consumers to perceive that future goal pursuit will be easier. The authors demonstrate this effect across a variety of tasks and means of exerting self-control, as well as with both hypothetical scenarios and real-behavior studies.
The difficulty of choosing to complete a goal-consistent action (e.g., going to the gym) over a goal-inconsistent action (e.g., sitting on the couch and watching TV) often depends on the situation. For example, imagine you have a weekly goal to go the gym three times a week. Today is a workout day and you are planning to go the gym. As soon as you are about to leave, your favorite TV show starts playing. Despite the temptation to watch, you convince yourself to go to the gym and complete your regular workout routine. Now imagine another situation: you again are planning to go the gym; however, in this case, there is nothing interesting to watch on TV. Again, you go to the gym and complete your regular workout routine. In both cases, you would complete the exact same workout routine and thus make the same amount of objective progress toward your workout goal. However, do consumers perceive this to be the case? In this research, we examine how completing a goal-consistent task when choosing to do so requires more (vs. less) self-control influences consumers’ perceptions of goal progress.
Consumers often struggle to reach their goals. For example, they find it difficult to become more fit, learn a new language, or complete tasks on their to-do list. As a result, they frequently seek out companies to help them reach their goals, such as CrossFit for fitness goals, Duolingo for language goals, and Wunderlist for task organization goals. In the health and fitness industry, for instance, more than one in five Americans have at least one health club membership, and health and fitness apps have over 87 million users (IHRSA 2019). Moreover, expanding to goals outside health and fitness (e.g., savings, language learning), millions of consumers are engaging with companies to help them reach their goals. By helping consumers reach their goals, these companies benefit from greater consumer adherence to their offerings as well as greater customer satisfaction.
One area that companies can leverage in motivating consumers is their perceptions of goal progress. Prior research has shown that perceived progress in attaining a goal is a key factor in determining consumers’ subsequent motivation (Fishbach and Dhar 2005; Kivetz, Urminsky, and Zheng 2006; Wallace and Etkin 2017). Yet despite the importance of progress perceptions for motivation, little work has examined the factors influencing progress perceptions to begin with (Campbell and Warren 2015; Cheema and Bagchi 2011; Huang, Zhang, and Broniarczyk 2012; Sharif and Woolley 2020). In this work, we explore a novel driver of goal progress perceptions: the amount of self-control a consumer must exert to make a goal-consistent choice.
Research defines self-control as “the self's capacity to alter its own states and responses” (Baumeister 2002, p. 670). Often when pursuing long-term goals, what consumers want to do (e.g., stay at home and sit on the couch) is at odds with what they think they should do (e.g., go to the gym) (Bazerman, Tenbrunsel, and Wade-Benzoni 1998). Thus, successfully exerting self-control means choosing the goal-consistent task (i.e., go to the gym) and rejecting the goal-inconsistent alternatives (i.e., sit at home and watch TV).
While all goal-relevant tasks require self-control to complete them, the extent of self-control often depends on the situation. In particular, consumers will need to exert more self-control when it is more difficult to choose the goal-consistent option (over the goal-inconsistent one). For example, consumers will need to exert more self-control to go the gym when their favorite TV show is on, as the goal-inconsistent action is more appealing. In addition, they will need to exert more self-control when they are feeling sluggish or are not in the mood to exercise, as the goal-consistent action is less appealing. In these situations, although consumers know they should exercise to achieve their long-term goal, they want to do it less, making it more difficult to choose to complete it. However, regardless of how difficult it is to choose to complete the goal-consistent option, completing it will lead to an equivalent amount of objective progress toward the goal of becoming more fit. Despite their objective progress being the same, we find that consumers perceive that they have made more progress on their goals when they exert more (vs. less) self-control to choose to complete a goal-consistent task. We argue that this is because people tend to infer that the greater effort involved in choosing a goal-consistent action signals greater commitment to the goal, which will make future goal pursuit feel easier.
This research makes several theoretical contributions to research on goal progress, self-control, and motivation. We contribute to, and connect, research on goal progress perceptions and research on self-control. While a large stream of research has examined how perceived goal progress influences motivation (e.g., Fishbach and Dhar 2005; Kivetz, Urminsky, and Zheng 2006; Wallace and Etkin 2017), a much more limited stream has examined the factors that influence goal progress perceptions to begin with (Campbell and Warren 2015; Cheema and Bagchi 2011; Huang, Zhang, and Broniarczyk 2012; Sharif and Woolley 2020). We contribute to the literature by demonstrating a novel factor that influences consumers’ perceptions of goal progress: the extent to which they must exert self-control to choose the goal-consistent action. In doing so, we simultaneously contribute to the research on self-control (Bartels and Urminsky 2011; Bazerman, Tenbrunsel, and Wade-Benzoni 1998; Hoch and Loewenstein 1991; Milkman, Rogers, and Bazerman 2008; Mukhopadhyay and Johar 2005; O’Donoghue and Rabin 1999; Urminsky 2017) by revealing an unexplored consequence of overcoming greater temptation.
Further, while previous research has examined progress perceptions and commitment as two distinct drivers of motivation (Carver 2004; Fishbach and Dhar 2005; Zhang, Fishbach, and Dhar 2007), the relationship between commitment and perceived progress has received less attention. We add to the literature by showing that inferred commitment from exerting greater self-control to engage in a goal-consistent action can elevate progress perceptions by making future goal pursuit feel easier.
Theoretical Development
Goal Progress Perceptions and Self-Control
Although several studies have shown that consumers frequently monitor their progress toward meeting their goals (Carver and Scheier 1998), and their perceptions of goal progress influence their motivation (Harkin et al. 2016; Kivetz, Urminsky, and Zheng 2006; Koo and Fishbach 2012), a very limited stream of research has examined the factors that influence consumers’ perceptions of goal progress to begin with. Prior research has found that consumers’ perceptions of progress depend on whether they are tracking goal-consistent (vs. goal-inconsistent) actions (Campbell and Warren 2015), how far they are from their goal (Huang, Zhang, and Broniarczyk 2012), the ease of visualizing the goal (Cheema and Bagchi 2011), goal distance and step-size ambiguity (Bagchi and Li 2011), as well as the manner in which they categorize their goal-relevant actions (Sharif and Woolley 2020). This stream of research has noted that a variety of factors influence consumers’ judgments about their goal progress.
We build on this existing research by examining how exerting more (vs. less) self-control influences perceptions of goal progress. In particular, we examine how the difficulty of choosing a goal-consistent action during a self-control conflict influences goal progress perceptions. When consumers set a goal (e.g., losing weight), their choices can be either consistent (also known as “should” behaviors) or inconsistent (also known as “want” behaviors) with the goal. For example, a consumer could choose to go to the gym (a goal-consistent action) or stay at home and watch TV (a goal-inconsistent action).
Consumers must make a trade-off when choosing between these actions. Goal-inconsistent actions have an immediate benefit and a delayed cost; goal-consistent actions have an immediate cost and a delayed benefit: whereas watching TV, rather than going to the gym, is enjoyable now (i.e., immediate benefit), it will cost you later in terms of your health/fitness goal; conversely, while going to gym and working out will benefit your health goal in the long run, it is less enjoyable now (i.e., immediate cost). Due to present-biased preferences (Bazerman, Tenbrunsel, and Wade-Benzoni 1998; Loewenstein and Thaler 1989; Milkman, Rogers, and Bazerman 2008; O’Donoghue and Rabin 1999; Urminsky and Zauberman 2015), consumers often place a greater weight on the immediate benefits and delayed costs over delayed benefits and immediate costs. To choose the goal-consistent action (over the goal-inconsistent action), consumers must exert self-control, defined as overriding an urge or immediate response and engaging in a different action (Baumeister, Vohs, and Tice 2007; Bertrams 2020).
Although making the choice to pursue a goal-consistent (vs. a goal-inconsistent) action already involves trade-offs, it can be exacerbated by small hurdles. For example, a consumer might be very tired and not in the mood to work out, their favorite TV show might be on when they are considering going to the gym, they might have to experience a longer commute to a gym on a particular day, or they might have to complete a goal-relevant task they have been putting off. As a result, although many consumers want to become more fit and have set out to do so, they may want to pursue the goal-relevant task less on a given day. This increases the trade-off between the goal-inconsistent (i.e., stay at home and not go to the gym) and goal-consistent (i.e., go to the gym to become more fit) actions and makes it more difficult and effortful to choose the goal-consistent action.
Indeed, even outside the self-control domain, prior research has found that consumers find it difficult and effortful to make choices (Luce, Payne, and Bettman 1999), because weighing the pros and the cons of each option can be cognitively difficult (Bettman, Luce, and Payne 1998; Payne, Bettman, and Johnson 1993). Choosing is also emotionally difficult, because by choosing one option, consumers must give up the other option (Brenner, Rottenstreich, and Sood 1999; Luce, Payne, and Bettman 1999). Further, as the difference in attribute values between choice alternatives increases, consumers experience more conflict in making a choice (Heath, Larrick, and Wu 1999; Wang et al. 2010), because by choosing one option, consumers must give up relatively larger advantages of the other option. As a result, they can find it depleting to choose between options with larger trade-offs (Wang et al. 2010). Building on this research, we suggest that choosing a goal-consistent action (over a goal-inconsistent action) when it is more (vs. less) difficult is mentally effortful for consumers.
Commitment and Progress Perceptions
We propose that choosing to complete a goal-consistent action when it is more effortful to do so (i.e., exerting more self-control) will increase consumers’ perceptions of goal progress by signaling greater goal commitment (Prelec and Bodner 2003). “Commitment” is defined as a person's attachment or determination to reach a goal (Klein, Molloy, and Brinsfield 2012; Locke, Latham, and Erez 1988). Prior research suggests several factors that can influence commitment toward a goal, such as self-efficacy, peer group influence, and external rewards/incentives (Bandura and Locke 2003; Locke, Latham, and Erez 1988).
We build on this research by suggesting a novel antecedent to consumer goal commitment: choosing to complete a goal-consistent action when it is more (vs. less) difficult to do so. Prior research supports this prediction, revealing that consumers often make inferences that they possess desirable (or undesirable) characteristics from the choices that they make (Khan and Dhar 2006). Consumers can infer greater goal commitment after choosing to complete a goal-consistent action (Fishbach and Dhar 2005), which can increase subsequent motivation. Further, consumers’ perceptions of how committed they are to their goal often depends on how much effort they perceive that they have exerted toward the goal-relevant task (Zhang and Huang 2010). Indeed, studies have theorized consumers to be more committed toward their goal, and thus more motivated, when they earned (vs. were endowed) high progress toward their goal (Zhang and Huang 2010). We build on this research by suggesting that exerting more effort to choose a goal-consistent action in the face of a more difficult self-control conflict will similarly lead to a boost in inferred commitment. Prior research has found that consumers make inferences about themselves not just from the options that they choose but also from the options that they reject. For example, Dhar and Wertenbroch (2012) show that consumers felt more satisfied when they chose a virtuous option when the choice set comprised both vices and virtues relative to a choice set that consisted of only virtues.
Thus, by choosing the goal-consistent option when it is more (vs. less) difficult to do so, consumers will make a greater positive inference about themselves. Notably, consumers will perceive they are more committed to reaching the larger goal that the goal-relevant task is working toward (e.g., more committed to becoming fit after going to the gym). Cognitive dissonance may contribute to this positive self-signal; that is, consumers feel that because they exerted additional effort to resolve the conflict, they should change their attitude about their commitment as a result (Festinger 1962). Cognitive dissonance can also affect commitment further down the road because it urges consumers to act consistently (Cialdini 2007). This can also enhance the self-signal we discussed, as inferring commitment from an initial goal-consistent task can signal to consumers that reaching the goal matters to them and encourages them to keep engaging the goal (Baca-Motes et al. 2013).
Ease of Future Goal Pursuit
We posit that higher inferred commitment due to higher self-control exertion will signal to consumers that the possibility of self-control lapses will be lower in the future (Hsiaw 2013). This is in line with a considerable stream of research that shows the enhancing effect of commitment on future self-control (Ariely and Wertenbroch 2002; Bisin and Hyndman 2021; Fudenberg and Levine 2006). With heightened perceptions of commitment, consumers may believe that they are now the type of person that can overcome these kinds of hurdles, which represents a shift in their self-perceptions (Freedman and Fraser 1966). For instance, previous research has shown that when consumers engage in effortful work, they feel reassured that desired outcomes are within their reach (Cutright and Samper 2014). Further, when consumers overcome a difficult self-control hurdle, they may perceive that they have the momentum needed to reach their goal (Dhar, Huber, and Khan 2007; Gollwitzer 1990; Markman and Guenther 2007), as they sense that the psychological hurdles that prevented them from reaching their goal in the past have been removed.
Thus, when consumers infer that succumbing to future self-control hurdles is less likely, they feel that it will be easier for them to complete the remainder of their goal. In assessing goal progress, consumers must consider not only how much progress they have made but also how much work remains. Perceiving the remainder of the goal as easier or more attainable should cause consumers to assume greater goal progress from their undertaken goal-consistent action. Indeed, prior research has found that consumers often rely on the perceived attainability of the goal when determining their goal progress (Campbell and Warren 2015). That is, if consumers perceive that their action confirms the expectation that they will reach the goal, they believe that they have made more progress toward that goal.
Across eight experiments using hypothetical scenarios and real-behavior studies, we show that when consumers exert more (vs. less) self-control to choose to complete a task, they perceive they have made more progress on their goals (holding the task itself constant). Further, we reveal that this effect is driven by increased perceptions of commitment toward the goal, which makes future goal pursuit feel easier.
Study 1a: Exerting Self-Control in the Presence of Attractive Goal-Inconsistent Alternatives
In Study 1a, we examine our main effect: whether exerting more (vs. less) self-control to choose a goal-consistent option impacts progress perceptions. In particular, we manipulate the attractiveness of the goal-inconsistent option, leading to a greater conflict between “wanting” to do the goal-consistent task and how much one “should” do the task. For all scenarios that manipulate self-control by increasing the “want” versus “should” conflict, (Studies 1a, 3, and 4), we conducted two pretests (see Web Appendices A and B). These pretests confirmed that (1) our self-control scenarios successfully created a want–should conflict for participants and (2) perceptions of goal importance and task difficulty did not change as a result of our manipulations.
Method
We preregistered this study (https://aspredicted.org/blind.php?x=ZVC_Z1H) for 700 human intelligence tasks (HITs) on Amazon Mechanical Turk (MTurk). Seven hundred seven MTurk participants (49.1% female; Mage = 40.41 years, age range: 19–83 years) were randomly assigned to one of two self-control (SC) conditions: more SC and less SC.
All participants were asked to imagine having enrolled in a gym to become more fit. They were told that their goal is to go to the gym four days a week, and the normal exercise session consists of five exercises, taking about 60 minutes total. Participants in the more SC condition were then told, “Now imagine that today is a workout day. However, you realize that your favorite TV show will be on during your normal gym time today. This show is only on today, one day of the week. You have been really wanting to watch this episode of your television show. However, although you want to stay and watch the show, you decide to go to the gym. You complete all 5 of your workouts, taking 60 minutes total.” Participants in the less SC condition simply read, “Now imagine that today is a workout day. You complete all 5 of your workouts, taking 60 minutes total.” All participants then responded to the progress and commitment questions, whose order was counterbalanced. To measure progress, we asked, “How much progress do you think you have made from this workout session on your goal of completing all your workouts in the week?” on a scale from 0 (“No progress at all”) to 100 (“A lot of progress”). To measure participants’ perceived commitment to the goal we asked, “How committed do you think you are to your fitness goal?” (1 = “Not committed at all,” and 7 = “Very committed”). Half of the participants responded to the progress question first, and the other half responded to the commitment question first. 1
Results
We removed participants who failed the attention check question, and our final data set consisted of 703 MTurk participants (49.2% female, Mage = 40.41 years, range = 19–83 years).
Perceived progress
We did not find a significant SC (more vs. less) × order (progress first, commitment first) interaction effect on perceived progress (F(703) = .067, p = .80). We therefore collapsed across the two order conditions. As predicted, participants in the more SC condition perceived that they made more progress (Mmore SC = 66.70, SD = 28.59) on their workout goal for the week than those in the less SC condition (Mless SC = 51.75, SD = 29.90; B = 14.95, SE = 2.21, t(701) = −6.774, p < .001, d = .51; 95% confidence interval [CI] of the difference = [−19.28, −10.61]).
Perceived commitment
Similarly, participants in the more SC condition rated their commitment to the goal higher than those in the less SC condition (Mmore SC = 6.34, SD = .90; Mless SC = 5.97, SD = 1.07; B = .37, SE = .07, t(701) = −4.918, p < .001, d = .37, 95% CI of the difference = [−.51, −.22]). We then ran a mediation analysis using the PROCESS macro (Hayes 2017, Model 4), with condition as the independent variable, perceived progress as the dependent variable, and commitment as the mediator. As we hypothesized, exerting self-control had a positive effect on perceived commitment to the fitness goal (B = .37, p < .001). Commitment had a positive effect on progress perceptions (B = 8.98, p < .001). The 95% CI of the indirect effect of condition on progress through perceived commitment did not include zero (b = 3.30, SE = .78, 95% CI = [1.88, 4.92]), providing support to commitment as the mediator.
Discussion
Study 1a provided preliminary support for our central hypothesis that consumers perceive greater goal progress after exerting more (vs. less) self-control to complete an identical task. The results of this study also demonstrated initial evidence for the proposed mechanism: that exerting self-control leads to higher progress perceptions by signaling greater commitment.
Study 1b: Self-Control When the Goal-Consistent Task Is Less Desirable
In an additional preregistered study reported in Web Appendix C (Study 1b: https://aspredicted.org/blind.php?x=jv92kd), we replicated our main effect with another manipulation. Consumers must exert more self-control to choose to take on a goal-relevant task when they are not in the right mood, such as when they feel sluggish. This more difficult conflict forces consumers to exert more self-control to choose to complete the task. We find consistent evidence that consumers perceive greater goal progress in this scenario—that is, when they exert more self-control to choose the goal-consistent option when they are feeling sluggish (Mmore SC = 68.03, SD = 26.34 vs. Mless SC = 60.55, SD = 30.10; B = 7.48, SE = 2.83, t(398) = 2.64, p = .008, d = .26, 95% CI of the difference = [1.92, 13.04]) and, further, that greater commitment mediates the effect (b = 2.76, 95% CI = [.33, 5.23]). In Study 1c, we aim to replicate our effect in a real-behavior setting.
Study 1c: Reflecting on Self-Control Difficulty: Real-Behavior Study
The objective of Study 1c is to examine if priming participants to perceive that choosing the goal-consistent task is more (vs. less) difficult (i.e., requires more [vs. less] self-control to choose to complete it) impacts goal progress perceptions in a real-behavior study. In particular, we primed participants to think of previous instances when they felt sluggish or not in the right mood and did not make it to the gym (vs. to think of an unrelated event). Participants then had the opportunity to complete a short workout. We expected that priming the participants with a difficult self-control conflict in the past (vs. not) will lead them to perceive that choosing to complete the goal-relevant task now is also more difficult, requiring more self-control. If they then successfully complete the task, we hypothesized that they would perceive that they have made more progress to their goal than those who were not primed.
Method
We preregistered this study (https://aspredicted.org/blind.php?x=7ff5uu) for 800 HITs on MTurk. Eight hundred six MTurk participants (59% female; Mage = 35.97 years, age range: 18–81 years) were randomly assigned to one of the two conditions: more SC versus less SC. In the more SC condition, we asked participants to recall and describe two instances in which they had planned to exercise but they did not feel like doing so. Specifically, they read, “Recall a time when you had planned to workout but you did not feel like doing it. For example, think about times where you felt sluggish or lacked motivation. Below please write about one such incident by describing what workout you were planning to do and why you decided to not do the workout.” By having participants recall times in the past when they struggled to exercise due to a self-control conflict (i.e., feeling too sluggish to do the planned workout), we intended to prime them with the sense that choosing to complete this goal-relevant task is more difficult. Participants in the less SC condition were asked to recall and describe instances in which they went shopping for clothes. We chose to have participants think about a domain outside of fitness in the less SC condition so as not to influence their baseline perceptions of how difficult it is to choose to complete a fitness task. After the priming task, all participants were told that they could receive a monetary bonus if they completed a series of four workouts that would take 30 seconds each. They were asked to indicate whether they wanted to complete or skip the workout task.
Participants who opted in were asked to complete the workouts to the best of their ability. After completing the workouts, they rated their progress toward the goal of becoming more fit on a scale from 0 to 100. Then, they were asked about their perceived effort to start the workout and perceived effort to complete the workout, both on a seven-point scale (1 = “No effort at all,” and 7 = “A lot of effort”). We also asked participants to indicate if they indeed completed the workouts (we assured them that their answer to this question would not affect their payment). We preregistered to keep only those participants who reported to have completed all the workouts.
Results
Overall, 92% of the participants chose to complete the workout task (N = 735), which did not differ between conditions (%more SC = 91.7%, %less SC = 90.7%; χ2(1, N = 806) = .21, p = .64). Of the 735 participants who opted in, 86.5% reported to have completed the workout task. As such, and as preregistered, we included a total of 636 participants (58.3% female; Mage = 36.20 years, age range: 18–81 years) in the following analyses. There was no significant difference between the number of participants in the two conditions who completed the workouts (%more SC = 88.3, %less SC = 84.9; χ2(1, N = 735) = 1.79, p = .41).
Priming check
As we expected, and confirming our priming manipulation, participants in the more SC condition rated the effort required to start (i.e., choose to complete) the workout task as greater than the participants in the less SC condition (Mmore SC = 3.89, SD = 2.09 vs. Mless SC = 3.40, SD = 2.00; t(634) = −3.01, p = .003, d = .24; 95% CI of the difference = [−.81, −.17]). However, as intended, there was no difference between the two conditions in the amount of effort required to complete the workout itself (Mmore SC = 4.68, SD = 1.82 vs. Mless SC = 4.56, SD = 1.79; t(634) = −.830, p = .41; 95% CI of the difference = [−.40, .16]).
Perceived progress
As predicted, we found that participants perceived they had made significantly more progress in the more SC condition than in the less SC condition (Mmore SC = 35.99, SD = 30.12 vs. Mless SC = 28.79, SD = 27.20; B = 7.20, SE = 2.27, t(634) = −3.17, p = .002; d = .25; 95% CI of the difference = [−11.68, −2.71]). The results showed the same pattern when we included all participants, even those who indicated to have not completed all the workouts (Mmore SC = 35.00, SD = 30.03 vs. Mless SC = 29.31, SD = 27.44; t(733) = −2.68, p = .007; 95% CI of the difference = [−9.85, −1.53]).
Discussion
This study replicated the results of Studies 1a and 1b in a real-behavior setting, revealing consistent evidence that exertion of greater self-control (or, in this case, reminders of such self-control exertion) to choose a goal-consistent task positively affects consumers’ perceptions of goal progress. This study revealed that merely thinking about a previous self-control conflict leads consumers to perceive they have exerted higher self-control if they complete this same task today, and thus have made more progress. This suggests that companies can utilize this effect and boost customers’ progress perceptions simply by having them reflect on past self-control conflicts. In the next study, we rely on this finding to create two sets of push notifications that can be sent to users who are pursuing a goal. The main objective of Study 2 is to further examine the self-control–commitment–progress causal link. To that end, we manipulate perceived goal commitment (i.e., our proposed mediation) consistent with Pieters’s (2017) recommendation of conducting experimental mediation studies.
Study 2: Commitment Leads to Progress
Thus far, we have demonstrated that exerting higher self-control to choose to complete a goal-consistent task increases consumers’ perceived progress toward their goal by increasing their perceived commitment to the goal. If enhanced commitment indeed drives our effect, then our effect should be attenuated if consumers already infer high commitment for an external reason. For instance, consumers can utilize extrinsic motivation tools to precommit to the goal-relevant tasks that they aim to complete (Rogers, Milkman, and Volpp 2014). In this study, we examine whether a signal of high commitment due to an external factor moderates the effect of self-control on perceived progress. Study 2 also builds on Study 1c by examining another managerially relevant context: the messages that companies send consumers to motivate desirable behavior. Indeed, most health and wellness applications try to encourage and motivate their users by sending them both general and customized push notifications. For instance, Aloe Bud, a wellness app, sends gentle self-care reminders that encourage users to celebrate the effort they make and the resulting victories; the LevelMoney app used to send encouraging personalized messages such as “Boomshakalaka! You’ve spent less than 50% of your Spendable so far this week! Good job!” Consistent with these examples, we manipulate self-control exerted to choose a goal-relevant option by reminding half of participants of their self-control effort while keeping the goal-relevant task identical between conditions.
Method
We preregistered this study (https://aspredicted.org/blind.php?x=J47_2BL) for 2,000 HITs on Prolific. Two thousand five participants (56% female; Mage = 38.00 years, age range = 18–85 years) were randomly assigned to one of the four conditions in a 2 (SC: more vs. less) × 2 (perceived commitment: low vs. high) between-subjects design. All participants were told, “Imagine you have downloaded an exercise app on your phone. The app runs an exercise challenge to motivate users to do daily workouts. The challenge is for you to follow the app instructions and complete 30 minutes of workout every day for seven days. Now imagine that today was the first day of this week's challenge, and you successfully completed today's 30-minute workout. There are six more days left of this challenge.” Participants were then asked to imagine having received a message from the app. In the less SC condition, they read, “Congratulations! The first day of the workout challenge is over.” In the more SC condition, they read, “Congratulations! The first day of the workout challenge is over. Reflect on all the things that you could do instead of working out. For example, you could have watched a TV show, or browsed social media. Although it might have been difficult, you successfully exerted self-control and chose to complete your workouts.” We also manipulated perceived commitment to the goal. Participants in the high-commitment condition read, “As a means to further commit yourself to successfully complete the challenge, you decide to set up an arrangement with your friend. You give your friend $20. If you successfully complete the challenge, they will give you the $20 back. If you do not, your friend will keep the $20.” Participants in the low-commitment condition did not receive this prompt.
We then asked all participants, “How much progress do you think you have made on your goal of completing this challenge successfully?” Responses were recorded on a 101-point scale (0 = “No progress at all,” and 100 = “A lot of progress”).
Results
Four participants failed the attention check question. As preregistered, only the remaining 2,001 participants were included in the analyses (56% female; Mage = 38.00 years, age range = 18–85 years).
Commitment manipulation check
To ensure that the scenario about participants’ giving $20 to a friend to commit themselves to the challenge indeed manipulated commitment, we ran an independent pretest on a sample of 601 participants (44% female; Mage = 39.90 years, age range = 20–75 years). All participants read about the exercise app and the challenge it runs, as in the study. Half of participants read, “As a means to further commit yourself to successfully complete the challenge, you decide to set up an arrangement with your friend. You give your friend $20. If you successfully complete the challenge, they will give you the $20 back. If you do not, your friend will keep the $20.” The other half did not read this prompt. We asked all participants to rate their commitment to complete the challenge. Participants who received the commitment manipulation rated their perceived commitment higher than those who did not get the manipulation (Mmanipulation = 85.96, SD = 16.22; Mcontrol = 73.44, SD = 21.82; t(599) = −7.892, p < .001, d = .64), showing that the manipulation successfully manipulated perceived commitment.
We ran another pretest to ensure that this manipulation of commitment would not affect perceived effort to start or complete the workout task. A sample of 203 participants (52% female; Mage = 41.94 years, age range = 20–73 years) took part in this pretest. All read about the workout challenge on the app. Again, half of them received the money manipulation described previously, and the other half did not. All read that they successfully completed the first day of challenge and were asked to rate the effort they had to exert to start the workout and the effort they had to exert to complete the workout, both on seven-point scales (1 = “No effort at all,” and 7 = “A lot of effort”). There was no difference between the two groups in their ratings of effort to start (Mmanipulation = 4.78, SD = 1.59; Mcontrol = 5.03, SD = 1.36; t(201) = 1.179, p = .24) and effort to complete (Mmanipulation = 5.26, SD = 1.57; Mcontrol = 5.38, SD = 1.41; t(201) = .531, p = .60).
Perceived progress
We conducted a linear regression predicting perceived progress from a dummy variable representing the more (vs. less) SC conditions, the high-commitment (vs. control) conditions, and their interaction. A significant 2 (SC: more vs. less) × 2 (commitment: high vs. control) interaction effect on perceived progress emerged (B = −10.59, SE = 2.62, t(1,997) = −4.04, p < .001, 95% CI of the difference = [−15.72, −5.45]).
We then conducted an analysis of simple effects. In the control condition, participants who were reminded of their self-control perceived higher progress toward the goal than those who did not receive the self-control message (Mmore SC = 38.15, SD = 29.13 vs. Mless SC = 27.88, SD = 24.41; d = .38; SE = 1.85; F(1,997) = 30.77, p < .001, 95% CI of the difference = [−13.90, −6.64]), replicating our finding in the previous studies. However, in line with our predictions, when perceived commitment was manipulated to be high (participants imagined paying $20 to a friend), reminding participants of their self-control did not affect their perceived progress toward the goal (Mmore SC = 43.98, SD = 30.42 vs. Mless SC = 44.30, SD = 32.60; SE = 1.85; F(1,997) = .029, p = .86, 95% CI of the difference = [−3.32, 3.95]; see Figure 1).
Effect of Self-Control by Commitment Interaction on Progress Perceptions.
Discussion
In this study, we specifically focused on the second link in the self-control–commitment–progress causal chain by manipulating commitment. Our hypothesis was that exerting self-control increases perceptions of progress by signaling commitment. Therefore, if commitment is already high due to an external commitment strategy (e.g., wagering money to force oneself to complete the challenge), the effect of self-control on progress perceptions effectively goes away. We found evidence of this in Study 2.
In the next study, we aim to manipulate the second mediator: the perceived ease of future goal pursuit. If the effect of exerting self-control on perceived progress goes through consumers’ perceptions that overcoming future self-control hurdles will be easier, this effect should be mitigated in situations when future goal pursuit has been made easier due to an external factor.
Study 3: Manipulating the Perceived Ease of Future Goal Pursuit
Method
Fifteen hundred ninety-nine MTurk participants (58.8% female; Mage = 40.06 years, age range = 18–86 years) were randomly assigned to one of four conditions in a 2 (SC: more vs. less) × 2 (future: easier vs. control) between-subjects design. The stimuli we used in this study were similar to Study 2. All participants read about an app that runs an exercise challenge to motivate users to do daily workouts: “The challenge is for you to follow the app instructions and complete 30 minutes of workout every day for seven days.” We asked participants to imagine that today was the first day of this week's challenge, and that they successfully completed today's 30-minute workout. As in Study 2, participants were then asked to imagine having received a message from the app. In the more SC condition, the message reminded them of their self-control effort to complete the first day of the challenge, while in the less SC condition participants were simply congratulated on completing the first day.
We then manipulated the perceived ease of future goal pursuit by telling half of the participants that “unlike the first day of the challenge, the following days would require less self-control, as you will have no access to TV because there is maintenance going on in your building. Therefore, there will be no distraction to keep you from doing your workouts.” The other half of participants did not receive this prompt. In other words, for half of the participants, completing the future workouts was made easier due to an external factor (i.e., not having access to TV), which is expected to reduce the possibility of future self-control lapses. We then asked all participants to rate their progress by responding to “How much progress do you think you have made on your goal of completing this challenge successfully?” on a scale from 0 to 100.
Results
Sixteen participants failed the attention check question, leaving us with 1,583 participants (58.8% female; Mage = 40.14 years, age range = 18–86 years), as preregistered.
Perceived ease manipulation check
To ensure that the maintenance scenario that we used indeed manipulated perceived ease of completing the challenge, we ran an independent pretest on a sample of 601 participants (44% female; Mage = 39.90 years, age range = 20–75 years). All participants read about the same workout challenge as described in the study and that they completed the first day of the challenge successfully. Half of participants were randomly assigned to read, “Unlike the first day of the challenge, the following days would require less self-control, as you will have no access to TV because there is maintenance going on in your building. Therefore, there will be no distraction to keep you from doing your workouts.” The other half did not read this prompt. We asked all participants, “How easy do you think it will be for you to successfully complete this challenge?” Participants who saw the maintenance manipulation rated the perceived ease of completing the challenge significantly higher than those who did not get the manipulation (Mmanipulation = 75.97, SD = 21.32; Mcontrol = 61.82, SD = 25.69; t(599) = −7.355, p < .001, d = .60). This shows that the maintenance scenario successfully manipulated the perceived ease of completing the goal.
Perceived progress
We conducted a linear regression predicting perceived progress from a dummy variable representing the more (vs. less) SC conditions, the more (vs. less) easy conditions, and their interaction. We found a significant 2 (SC: more vs. less) × 2 (future: more easy vs. less easy) interaction effect on perceived progress (B = −7.58, SE = 3.13, t(1,583) = −2.42, p = .015, 95% CI of the difference = [−13.72, −1.44]).
We conducted an analysis of simple effects. In the control condition, participants in the more SC condition perceived to have made higher progress than those in the less SC condition (Mmore SC = 44.54, SD = 31.55 vs. Mless SC = 34.87, SD = 30.11; d = .31, SE = 2.21, F(1,583) = 19.154, p < .001, 95% CI of the difference = [−14.00, −5.433]). However, when the future workouts were perceived to feel easier because of lack of access to TV, we found no difference in perceived progress between the two conditions (Mmore SC = 53.01, SD = 31.72, vs. Mless SC = 50.93, SD = 31.05; SE = 2.21, F(1,583) = .885, p = .35, 95% CI of the difference = [−6.43, 2.26]).
Discussion
In this study, we focused on the second mediator by manipulating the perceived ease of future goal pursuit by reducing future self-control hurdles and examined the interaction effect of exerting self-control and how easy participants perceived completing the future goal-relevant tasks will be on their perceptions of progress. We found that when completing future tasks are felt to be easier due to an external factor (e.g., participants will not have access to TV, so they will not have to exert self-control to prioritize exercising over watching TV), the effect of exerting self-control will not be as essential in signaling the lower likelihood of giving in to self-control lapses in the future. Therefore, the effect of exerting self-control on perceived progress is mitigated.
In the next study, we aim to examine (1) if the effect persists even when self-control is applied in the middle of goal-relevant task completion and (2) if commitment and perceived ease of future goal pursuit serially mediate the relationship between exerting self-control and progress perceptions.
Study 4: Exerting Self-Control to Start Versus Continue the Task
So far in all our studies, the effort of choosing the goal-consistent option has been applied to the beginning of the task. But is the effect of exerting self-control to choose to complete a goal-consistent action on perceived progress specific to task initiation instances? In Study 4, we aim to examine if it matters where during the task completion process (beginning vs. middle) the self-control effort is applied to choose to complete the task. In addition, we aim to explore the underlying mechanism further. In particular, we examine the overall causal link: exerting self-control–inferred commitment–perceived ease of future goal pursuit–perceived progress. We propose that higher inferred commitment due to choosing a goal-consistent action over the goal-inconsistent alternatives will make consumers feel that overcoming self-control hurdles, and therefore pursuing the goal in the future, will be easier. We test this proposition in Study 4.
Method
We preregistered this study (https://aspredicted.org/blind.php?x=/TKT_5GK) for 900 participants on MTurk. Nine hundred twenty-six participants (40% female; Mage = 38.52 years, age range = 18–80 years) took part in this study and were randomly assigned to one of the three conditions: SC beginning, SC middle, and no SC. All participants were asked to imagine having the goal of becoming more fit, for which they exercise four days a week, for about 60 minutes each time. We also asked them to imagine doing their workouts at their home where they have a small home gym. Participants in the SC beginning condition then read, “Now imagine that today is a workout day. You are about to start your workout, but you realize that your favorite TV show is on exactly at the time of your workout. Despite wanting to watch the show, you decide to let it go and start doing your workout instead. You complete all your exercises.” Participants in the SC middle read, “Now imagine that today is a workout day. You have started your workout but in the middle of your workout, you realize that your favorite TV show is about to start. Despite wanting to stop your workout and watch the show, you decide to let it go and keep doing your workout instead. You complete all your exercises.” Therefore, the difference between these two scenarios was that in the SC beginning condition, the self-control dilemma had to be resolved before the goal-relevant task (i.e., exercising) started, whereas in the SC middle condition, this conflict needed to be resolved in the middle of the task. Participants in the no SC condition simply read, “Now imagine that today is a workout day. You start your workout and you complete all your exercises.” Notably, in the SC middle and SC beginning conditions, participants had to exert more self-control effort to choose to continue or complete the workout than in the no SC condition; however, the workout itself was equivalent and required an equal amount of effort.
Similar to the previous studies, we asked all participants to rate their progress by responding to “How much progress do you think you have made from this workout session on your goal of completing all your workouts in the week?” We also asked participants, “How committed do you think you are to your workout goal?” and “How easy do you think it will be to get yourself to complete the rest of your workouts in the week?” Responses were collected on a scale from 0 to 100. As a manipulation check we asked, “In that scenario, how much effort did you need to convince yourself to do your workout?” on a scale from 0 to 100.
Results
Forty-three participants failed the attention check question and were removed from the analyses. This left us with 883 participants (46.2% female; Mage = 39.08 years, age range = 18–83 years).
Manipulation check
Participants in both self-control conditions (SC beginning and SC middle) rated the effort to convince themselves to do the workout higher than participants in the no SC condition (MSC begin = 67.39, SD = 24.91 vs. Mno SC = 57.25, SD = 28.83; t(576) = 4.52, p < .001; MSC mid = 66.75, SD = 24.37 vs. Mno SC = 57.25, SD = 28.83; t(591) = 4.34, p < .001). The difference in perceived effort between the two SC conditions was not significant (t(593) = .316, p = .75).
Perceived progress
We conducted a linear regression predicting perceived progress from two dummy variables representing conditions, with the no SC condition as the reference group. Participants in the SC beginning condition rated their progress significantly higher than those in the no SC condition (MSC begin = 62.67, SD = 28.73 vs. Mno SC = 54.61, SD = 28.94; B = 4.03, SE = 1.20, t(576) = 3.36, p < .001, d = .28, 95% CI of the difference = [3.34, 12.77]), replicating our previous findings. Similarly, participants in the SC middle condition rated their progress higher than those in the no SC condition (MSC mid = 68.07, SD = 26.35 vs. Mno SC = 54.61, SD = 28.94; B = 13.46, SE = 2.27, t(576) = 5.93, p < .001, d = .49, 95% CI of the difference = [9.00, 17.92]). In fact, participants in the SC middle condition rated their progress higher than those in the SC beginning condition (B = 5.40, SE = 2.26, t(576) = −2.39, p = .009, d = .19, 95% CI of the difference = [−9.84, −.97]).
Commitment and perceived ease of completing future goal-consistent tasks
Participants in the SC beginning condition rated their commitment and perceived ease of completing future workouts significantly higher than those in the no SC condition (commitment: MSC begin = 84.18, SD = 16.92 vs. Mno SC = 79.42, SD = 18.03; t(576) = 3.27, p = .001, d = .27, 95% CI of the difference = [1.90, 7.61]; ease of completing future workouts: MSC begin = 72.63, SD = 20.91 vs. Mno SC = 67.66, SD = 21.93; t(576) = 2.79, p = .005, d = .23, 95% CI of the difference = [1.47, 8.47]). We find similar effects with the SC middle condition versus the no SC condition (commitment: MSC begin = 84.82, SD = 16.72 vs. Mno SC = 79.42, SD = 18.03; t(576) = 3.78, p < .001, d = .31, 95% CI of the difference = [2.60, 8.20]; ease of completing future workouts: MSC begin = 70.88, SD = 21.63 vs. Mno SC = 67.66, SD = 21.93; t(576) = 1.80, p = .072, d = .15, 95% CI of the difference = [−.29, 6.74]).
Serial mediation
We collapsed across the two self-control conditions. 2 We ran a serial mediation analysis using the PROCESS macro (Hayes 2017, Model 6) with condition (0 = No SC, 1 = SC beginning or SC middle) as the independent variable, commitment as the first mediator and perceived ease of completing future workouts as the second mediator, and progress perceptions as the dependent variable. Exerting self-control had a positive effect on commitment (b = 5.09, p < .001). Commitment had a positive and significant effect on ease of completing future workouts (b = .64, p < .001), which positively predicted perceived progress (b = .42, p < .001). The 95% CI of the indirect effect of condition on progress through the commitment–perceived ease of completing future workouts link did not include zero (bindirect = 1.36, SE = .39, 95% CI = [.68, 2.19]), in support of a serial mediation.
Discussion
Study 4 revealed that participants perceive greater goal progress after exerting more (vs. less) self-control, both when this self-control is exerted prior to and in the middle of completing a goal-relevant task. The results also provided support for our full causal model—that is, exerting self-control increases perceived commitment to the goal, which leads consumers’ to perceive that it will be easier for them to pursue the goal in the future, which in turn increases their perceptions of the progress contribution of the goal-relevant task.
We unexpectedly found in this study that participants in the SC middle condition rated their progress significantly higher than those in the SC beginning condition. It is possible that participants perceived the self-control conflict to be even greater when an appealing outside option appears in the midst of goal pursuit (while tired) compared with before goal pursuit (when less tired). Overcoming a greater perceived conflict in the SC middle condition could lead to this effect. However, future research should further explore this unexpected finding.
In the next two studies, we pursue two objectives. One, we change the context of our scenarios. So far, all our studies have employed the fitness context. In Studies 5 and 6, we will test our effects in the daily to-do list and financial goals contexts, respectively. Two, we examine the implications of our findings for marketers. We have already shown in Studies 2 and 3 that companies can leverage the self-control effect on progress perceptions by reminding their customers of their self-control efforts, which will signal commitment to them, make them feel that the future goal pursuit is easier, and boost their perceptions of progress. In the following studies, we focus on downstream behavioral consequences of exerting self-control and examine their pros and cons for consumer motivation. In Study 5, we examine the effect of encouraging consumers to start by completing a goal-consistent option that requires more (vs. less) self-control. In particular, we look at the impact of first completing a task that has been on consumers’ to-do lists for a longer (vs. shorter) amount of time on consumers’ perceptions of goal progress. In Study 6, we focus on a downstream consequence that is directly relevant to marketers.
Study 5: Completing Procrastinated Tasks on a To-Do List
When consumers find it difficult to choose to complete a goal-consistent task, they often procrastinate and delay completing it (Soman et al. 2005). That is, even though two tasks might be equivalent in terms of physical effort required to complete them, those tasks that consumers want to complete less may stay on their to-do list for a longer period of time. Thus, in Study 5, we examined how completing an item on one's to-do list that has been there for a longer (vs. shorter) period of time influences consumers’ perceptions of goal progress. Completing a task that has been on the to-do list for a longer (vs. shorter) period of time will require more self-control to choose to complete it. As a result, we suggest that completing a task that has been on the list for longer (vs. shorter) period of time, controlling for characteristics of the task itself, will lead consumers to perceive greater progress on completing tasks on their to-do list. Further, as prior research has found that consumers are more motivated when they perceive greater progress on a specific goal (Kivetz, Urminsky, and Zheng 2006), we suggest that consumers will be more motivated to complete more tasks after completing a task that has been on the list for a longer period of time.
Method
We preregistered this study (https://aspredicted.org/blind.php?x=ms4ii9), aiming to recruit until 1,000 people on MTurk signed up to participate, anticipating that there would be attrition for those who complete the task. Our preregistered sample size was estimated to be between 400 and 1,000 participants.
Stage 1
In the first stage (i.e., signup stage), 1,009 people signed up to participate. Participants were instructed to create a to-do list consisting of six tasks or activities. In constructing these to-do lists, they read, “Two of these tasks should be what you have been thinking of doing in the past 3 months but for some reason, you have not gotten around to doing them yet. The other four tasks should be what you have been thinking of doing today.” Participants were asked to make sure that the tasks they list will all take between 10 to 30 minutes. They were then shown six blank boxes to create their list. They read, “Please first list two tasks that you have wanted to do for the past 3 months or more, and then four tasks that you have wanted to do since today.”
On the next page, participants read that the experimenters will choose one of the tasks they listed. To receive a $.75 bonus, participants needed to complete their assigned task within 24 hours and complete a follow-up survey, in which they would be asked to upload a picture proving that they completed the task (“For example, if this task is to get a new gym membership, take a picture of your membership”). After reading the instructions, participants were asked if they wanted to participate in the bonus survey.
Six hundred sixteen participants (61% female; Mage = 36.78 years, age range = 16–81 years) indicated that they wanted to take part in the bonus survey. At this point, they were randomly assigned to one of the two conditions: more SC versus less SC. Participants in the more SC condition were asked to complete the first task they had listed as something they had wanted to do for the past three months but they had not done so. Participants in the less SC condition were asked to complete the first task they had listed as something they had intended to do in that day. After participants were told which task they were assigned to complete, they were given a link to a follow-up survey and were asked to save the link, complete the selected task within 24 hours, and then go to the link and upload a picture proving the completion of the task.
Stage 2
Three hundred ten participants (59.7% female; Mage = 37.86 years, age range = 18–81 years) ended up uploading pictures showing that they had completed the selected task. After uploading their pictures, participants were shown the to-do list they had created in the first stage. The task they had completed was scratched off. All participants then were asked about their perceived progress: “After completing this task, how much progress do you think you have made toward completing tasks on the to-do list listed above?” (0 = “No progress at all,” and 100 = “A lot of progress”). On the next page, participants were asked, “Are you planning on completing additional tasks on the to-do list above within the next 24 hours?” (yes/no). Participants who selected “Yes” were then asked to select which task(s) from their list they would like to complete. At this point, all participants had five tasks remaining on their lists.
Participants were then told that if they completed the task(s) they selected and uploaded the picture to prove it, they would receive an additional $.25 bonus (the link for uploading the picture was provided). Lastly, participants were asked to think about the task they completed after the first stage. They were asked to indicate how long (in minutes) the task took. They were also asked to answer “How important was completing this task to you?” (0 = “Not at all important,” and 100 = “Very important”) and “How pressing was it to complete this task?” (0 = “Not at all pressing,” and 100 = “Very pressing”).
Stage 3
In the third (i.e., final) stage, participants who had completed more tasks on their list uploaded the pictures proving the completion of tasks (N = 140; 59.9% female; Mage = 38.76 years, age range = 18–81 years) to the link that was given to them in Stage 2. They were asked to indicate which task(s) on the list they completed.
Results
As described in the “Method” subsection, we collected data for our manipulation checks and main dependent variables in Stages 2 and 3. There was no difference between the two conditions in terms of how important their assigned task was (Mmore SC = 71.58, SD = 25.30 vs. Mless SC = 71.92, SD = 24.69; t(285) = .11, p = .91, 95% CI of the difference = [−5.57, 6.24]) and how pressing it was to completed the assigned task (Mmore SC = 46.41, SD = 28.96 vs. Mless SC = 47.12, SD = 30.95; t(285) = .20, p = .84, 95% CI of the difference = [−6.42, 7.86]).
We asked participants in Stage 2 to rate their progress toward completing the tasks on their to-do list after completing their assigned task in Stage 1. As we predicted, participants in the more SC condition rated their progress as significantly higher than those in the less SC condition (Mmore SC = 39.57, SD = 24.91 vs. Mless SC = 31.90, SD = 25.76; t(288) = −2.52, p = .012, d = .30, 95% CI of the difference = [−13.66, −1.69]). The percentage of participants who indicated that they planned to do more tasks was not statistically different between the two conditions (%more SC = 91%, %less SC = 89%, p = .5). However, in Stage 3, participants in the more SC condition planned to complete more tasks (Mmore SC = 2.31, SD = 1.44 vs. Mless SC = 1.97, SD = 1.17; t(289) = −2.23, p = .027, d = .26, 95% CI of the difference = [−.645, −.040]) and actually completed more tasks (Mmore SC = 1.10, SD = 1.39 vs. Mless SC = .72, SD = 1.04; t(289) = −2.66, p = .008, d = .31, 95% CI of the difference = [−.675, −.083]).
Because there was a significant difference between the two groups in terms of how long the task took to complete (Mmore SC = 55.37 minutes, SD = 52.40 vs. Mless SC = 35.62 minutes, SD = 35.44; t(285) = −3.81, p < .001, 95% CI of the difference = [−29.95, −9.54]), we examined whether our effects would hold if we controlled for the time participants spent to complete their assigned task. We found that even when controlling for time, the effect of self-control on progress is still marginally significant (B = 5.61, SE = 3.12, t(284) = 1.80, p = .073, 95% CI = [−.52, 11.74]). Similarly, despite controlling for time, participants in the more SC condition planned to complete more tasks than those in the less SC condition (B = .38, SE = .16, t(284) = 2.38, p = .018, 95% CI = [.065, .689]).
Robustness check
We had asked participants to list tasks that would take between 10 to 30 minutes in both conditions. As a robustness check, we then looked at only the subset of participants who followed this instruction and reported the task to have taken between 10 to 30 minutes. Consistent with our previous analyses, participants in the more SC condition rated their progress significantly higher than those in the less SC condition (Mmore SC = 38.32, SD = 24.85 vs. Mless SC = 29.94, SD = 22.70; t(141) = −2.037, p = .044, d = .35; 95% CI of the difference = [−16.52, −.25]). However, time was no longer different between the two groups (Mmore SC = 22.00 minutes, SD = 7.31 vs. Mless SC = 19.96 minutes, SD = 7.67; t(141) = −1.54, p = .12, 95% CI of the difference = [−4.66, .573]). This further shows that our effect is not driven by time, because when we compare participants who actually selected and completed tasks that take 10–30 minutes (i.e., when time varies less between the two conditions), the effect is still robust.
Discussion
We find consistent evidence in a real-behavior study that consumers perceive they have made more progress toward their goals when they complete a goal-relevant task that requires relatively more (vs. less) self-control to choose to complete it. Although this study is beneficial in examining the real-world consequences of exerting more (vs. less) self-control, it suffers from some limitations, some of which are helpful in predicting the implications of our findings for marketers. First, we relied on self-generated tasks in this study to examine the real-world effects of completing a task that requires more (vs. less) self-control to choose to complete, finding that our effects hold even when controlling for differences in the task. To further confirm this effect holds when the task is identical, we ran a preregistered (https://aspredicted.org/blind.php?x=dm47mr) hypothetical study in which all participants imagined completing an identical task on their to-do list, and we merely manipulated how long participants imagined that item was on their to-do list (six months in the more SC condition and one week in the less SC condition). We find convergent evidence that participants in the more SC condition indeed rated their perceived progress toward completing the tasks on the to-do list, as well as their productivity, higher than participants in the less SC condition (perceived progress: Mmore SC = 48.77, SD = 27.05 vs. Mless SC = 29.35, SD = 22.91; t(599) = −9.50, p < .001, d = .77, 95% CI of the difference = [−23.43, −.15.40]; productivity: Mmore SC = 5.48, SD = 1.42 vs. Mless SC = 4.63, SD = 1.62; t(599) = −6.85, p < .001, d = .56, 95% CI of the difference = [−1.10, −.61]).
Second, this study suffered from self-selection. We found that participants in the more (vs. less) SC condition were significantly less likely to return to Stage 2 of the study (B = −.67, SE = .16, Wald = 16.92, p < .001). Prior research has found that consumers are susceptible to a phenomenon called the “starting problem” (Heath, Larrick, and Wu 1999). That is, it is often difficult for consumers to initiate goal-relevant tasks. It is likely that this attrition is due to the “starting problem” being exacerbated in the more SC condition. Thus, one downside of encouraging consumers to first start with a task that requires more self-control is that they may be less likely to start the task to begin with.
However, interestingly, we find that participants are actually more likely to return to Stage 3 of the survey and indicate whether they completed the next task on the to-do list if they are in the more (vs. less) SC condition. We found that a significantly higher number of participants overall returned to Stage 3 when we included all participants randomly assigned to a condition (B = 1.52, SE = .64, Wald = 5.58, p = .018), which is not susceptible to self-selection, and also when we included only those who returned to Stage 2 (B = 1.95, SE = .65, Wald = 8.92, p = .003). Indeed, as indicated previously, we found that participants in the more SC condition actually completed more of the tasks after completing a task that required more (vs. less) self-control to complete. This may suggest that, overall, there is still a net benefit in terms of motivation by encouraging consumers to first complete a task that requires more (vs. less) self-control. Indeed, Studies 2 and 5 show that marketers can get around the “starting problem” while still reaping the benefits of greater goal progress perceptions by having consumers merely retrospectively reflect on exerting more (vs. less) self-control to complete a goal-relevant task.
In our final study, we examine another important consumer domain in which self-control lapses can prevent consumers from achieving their objectives: the domain of financial goals.
Study 6: Nudge Tool Implications for Company Perceptions
Study 6 examined the possible positive downstream consequences for consumers’ perceptions of a company if it used a messaging manipulation similar to that used in Study 2. We preregistered Study 6 for 1,000 participants (https://aspredicted.org/DX3_WDM) on Prolific. Participants were randomly assigned to one of two conditions: more versus less SC. All participants were asked to imagine having downloaded a budgeting app to track their expenses and financial goals. We told them that the app runs a saving challenge in which participants were encouraged not to spend money on nonessential items for seven days (we added that nonessential items are things you want but do not need, such as takeout food, entertainment, or clothing). We then told them to imagine successfully completing the first day of the challenge by avoiding spending any money on nonessential items. Participants were then told that the app sent them a message notification at the end of the first day.
In the more SC condition, the message read, “Congratulations! The first day of the saving challenge is over. Reflect on all the things that you really wanted to buy today but chose not to. For example, you could have spent money on a new shirt that you wanted but didn’t need, takeout food you could make at home, or a chocolate bar while waiting in the grocery line. Although difficult, you successfully exerted self-control in these situations and didn’t buy any nonessential items.” In the less SC condition, the message was simply, “Congratulations! The first day of the saving challenge is over.” 3
We then asked participants to rate the progress they have made on their goal of completing this seven-day challenge successfully (0 = “No progress at all,” and 100 = “A lot of progress”). We also examined whether enhanced progress perceptions can have positive downstream consequences for perceptions of the app. Thus, after asking about progress perceptions, we asked, “How helpful would you consider this app to be in encouraging you to save?” (0 = “Not helpful at all,” and 100 = “Very helpful”).
Results
Two participants failed the attention check and were excluded, leaving 999 participants (71.3% female; Mage = 32.24 years, age range: 18–79 years).
Perceived progress
Replicating our previous results, participants in the more SC condition rated their progress toward completing the challenge higher than those in the less SC condition (Mmore SC = 35.01, SD = 27.93 vs. Mless SC = 29.27, SD = 25.3; B = 5.74, SE = 1.68, t(999) = 3.41, p < .001, d = .22, 95% CI of the difference = [2.43, 9.04]).
We also found that participants in the more SC condition found the app to be more helpful in encouraging them to save (Mmore SC = 58.18, SD = 25.75 vs. Mless SC = 54.26, SD = 25.99; B = 3.92, SE = 1.64, t(999) = 2.40, p = .017, d = .15, 95% CI of the difference = [.710, 7.13]). Further, we found that the effect of more (vs. less) self-control on helpfulness was mediated by goal progress perceptions (b = 2.35, 95% CI = [1.07, 4.56]).
Discussion
The results of this study replicated our main hypothesis in yet another context. Importantly, we showed that companies can elicit more interest in their users by reminding them of their own effort and encouraging them to engage in similar goal-related activities. In addition, the results revealed that when messaging focused on the difficulty of exerting self-control, consumers felt that they made more progress on the challenge, leading them to believe the app was more helpful in encouraging them to save.
General Discussion
In this research, we found converging evidence for a novel factor that influences consumers’ perceptions of their goal progress: the amount of self-control they have exerted to choose to complete a goal-consistent option. This effect was robust across various paradigms and means of manipulating self-control. Further, we demonstrate why consumers overestimate their goal progress when the goal-relevant task they completed required more (vs. less) self-control. We reveal that this is because consumers perceive that they are more committed to their goal and that their future goal pursuit will be easier. We also provide examples of how marketers can reap the benefits of our findings in their messaging.
This article contributes to the limited stream of research examining the factors that influence goal progress perceptions (e.g., Campbell and Warren 2015; Huang, Zhang, and Broniarczyk 2012; Sharif and Woolley 2020). Although consumers frequently must exert more (vs. less) self-control to complete a goal-consistent task, prior research has not considered how this might affect their perceptions of goal progress. We demonstrate that exerting more (vs. less) self-control to choose a goal-consistent option has an influence on perceptions of goal progress.
In doing so, we also contribute to research examining how effort can influence judgments (Kruger et al. 2004; Schrift, Kivetz, and Netzer 2016) and, relatedly, judgments of goal commitment. Prior research has focused on the objective effort that oneself or others exert (Zhang and Huang 2010). We build on this research by examining how the effort used to resolve a self-control dilemma can similarly influence perceptions of goal commitment and goal progress. We find that even though objective effort advances objective goal progress and self-control effort does not, both similarly influence perceptions of progress.
While we observed that objective effort and self-control effort can similarly affect perceptions of progress, future research should explore if and when these two types of effort have different consequences—for example, on motivation. In one study, Jin, Xu, and Zhang (2015) had participants complete an easy (vs. hard) subgoal first. They found that participants who completed the task that required higher objective effort (i.e., the hard subgoal) were less motivated to further pursue the goal compared with those who completed the easy task first. We find preliminary evidence to suggest that the opposite occurs for self-control effort, such that people are more motivated after completing a task that requires more (vs. less) self-control effort. We suggest that in this case the different type of effort that consumers exert to complete a task (e.g., objectively more effort for a harder task vs. mental effort to start a task that consumers have less desire to complete) may have different consequences for motivation. This indicates that marketers should ensure that the task that consumers are recommended to start first is not objectively more difficult than subsequent tasks, but there may be some benefits if it requires additional self-control to choose.
We also contribute to the research on self-control (Bartels and Urminsky 2011; Milkman, Rogers, and Bazerman 2008; Mukhopadhyay and Johar 2005; Urminsky 2017), which has primarily focused on the difficulty of exerting self-control. We build on this research by finding corroborating evidence that exerting more self-control can be difficult and, at times, can lead to a “starting problem” (Heath, Larrick, and Wu 1999). More importantly, we demonstrate the consequences of successfully engaging in more (vs. less) self-control to complete the task, revealing that this can influence consumers’ perceptions of goal progress. We also show that the heightened perceptions of progress can be beneficial for consumers by increasing subsequent motivation (Study 5), and for companies by increasing consumers’ perceived helpfulness of the company's initiatives in achieving their goals (Study 6).
We also build on research that has examined progress perceptions and commitment as two distinct drivers of motivation (Carver 2004; Fishbach and Dhar 2005; Zhang, Fishbach, and Dhar 2007). We add to the literature by showing that inferred commitment from exerting higher self-control to engage in a goal-consistent action can elevate progress perceptions by making the future goal pursuit feel easier. However, it is possible that commitment may not always influence goal progress perceptions positively; for example, it may depend on the source of the inferred greater commitment. Further, we find that while exerting more (vs. less) self-control to complete an action increases perceived goal commitment, it does not necessarily affect goal importance. As this is an underexplored area, future research should further examine the relationship between commitment, progress, goal importance, and motivation.
Marketing Implications and Future Research
Marketers of companies that help consumers reach their personal goals (e.g., Weight Watchers, Mint) can use the findings of this research to motivate their consumers. Completing a goal-relevant task when it requires more (vs. less) self-control to choose to complete it can lead consumers to feel more satisfied with their progress, increasing consumer retention and positive word of mouth. While companies that help consumers reach their personal goals could encourage consumers to make more progress on their goals by giving them objectively more effortful tasks, this strategy could be demotivating, especially if consumers experience small failures in their attempt to complete them (Cochran and Tesser 1996; Sharif and Shu 2021). After expending greater objective effort toward a task, consumers could also be physically less able to complete a goal-relevant task the next day (i.e., a difficult gym task on Wednesday will make consumers sore for Thursday workout). This resonates with research finding that consumers are less motivated after completing a hard (vs. easy) subgoal (Jin, Xu, and Zhang 2015). Instead, companies could have consumers reflect on how much self-control they had to exert to complete their goal-relevant task that day. As demonstrated in this research, having consumers merely reflect on how effortful it was to choose to complete a task in the past, or reminding them that they exerted a lot of self-control, affects consumers’ perceived goal progress as well as improves perceptions of the company itself.
Relatedly, this research has implications for the order in which companies should encourage consumers to schedule their goal-relevant tasks. Prior research has demonstrated that consumers focus on the attainability of reaching a goal at low absolute goal progress (Huang and Zhang 2011). Thus, having them complete the tasks that require more self-control could lead them to feel greater perceived progress earlier on, leading them to be more motivated to complete the remaining tasks. While we find preliminary evidence to suggest that this may lead to an overall positive net effect on motivation, marketers should also be cautious of how this may lead to a greater chance of a “starting problem.”
In addition to those mentioned previously, there are also several other areas open for future research. We investigated the effect of self-control on goal-progress in both specific- and nonspecific goal contexts but we did not fully examine motivation in these contexts. Regardless of the end state, we can measure the net movement people perceive they have made by choosing to complete a high (vs. low) self-control goal-consistent task. We believe that when the goal is specific, participants who perceive to have made more progress will feel closer to the end state compared with those who have made less progress. In these situations, the goal-gradient hypothesis predicts that higher progress should lead to higher motivation. When the goal is nonspecific, however, the end state (e.g., being fit) can be construed differently for different participants depending on their current and ideal goal levels. Therefore, higher progress perceptions from exerting self-control may have no effect or a negative effect on motivation (Wallace and Etkin 2017). However, due to the other consequences of exerting self-control, such as greater commitment, it is possible that even with nonspecific goals, consumers are more motivated after exerting more (vs. less) self-control. Future research should explore this possibility.
In this article, we demonstrate that the self-control a consumer exerts to complete a goal-relevant task can influence their progress perceptions. While a great deal of research has been conducted on goal progress and self-control independently, there has been less focus on their intersection, which is a fruitful area for future work.
Supplemental Material
sj-pdf-1-mrj-10.1177_00222437221123969 - Supplemental material for It’s the Effort That Counts: The Effect of Self-Control on Goal Progress Perceptions
Supplemental material, sj-pdf-1-mrj-10.1177_00222437221123969 for It’s the Effort That Counts: The Effect of Self-Control on Goal Progress Perceptions by Hoori Rafieian and Marissa A. Sharif in Journal of Marketing Research
Footnotes
Author Note
The authors contributed equally to this work, and authorship order was randomly determined.
Associate Editor
Keisha Cutright
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
