Delegating Decisions: Recruiting Others to Make Choices We Might Regret

Delegating Decisions: Recruiting Others to Make Choices We Might Regret Consumers typically prefer freedom of choice, but when faced with a choice they might regret, they may prefer freedom from choice. Eight experiments show that people delegate difficult decisions, regardless of the decision’s importance, and regardless of their potential surrogate’s expertise. Delegation stems from a desire to avoid responsibility for potentially making the wrong choice rather than simply the desire to avoid the possibility of a poor outcome: although anticipated disappointment with the outcome and anticipated regret about one’s decision both contribute to the decision to delegate, only anticipated regret directly leads people to delegate choices to others. Consequently, delegation is an appealing method for coping with difficult choices while allowing consumers to retain the benefits of choosing that they would forgo by opting out of the choice. Moreover, giving people the option to delegate makes them less prone to walk away from difficult choices empty-handed. choice delegation, choice deferral, responsibility, regret Having the freedom to make choices for oneself is considered by many to be an inalienable right. Some have suggested that having control of one’s own choices is a basic need (Ryan and Deci 2000). When choice autonomy is threatened, people assert their freedom; for example, when they encounter unsolicited advice, they often discount or reject that advice (Fitzsimons and Lehmann 2004). Perceiving a lack of choice autonomy can even threaten people’s happiness, self-esteem, and physical well-being (Langer and Rodin 1976; Seligman 1975; Steele 1988; Taylor and Brown 1988; Usta and Häubl 2011). It is perhaps no surprise then that, even when choosing exacts a toll, people often prefer to choose themselves rather than have another person choose on their behalf (Botti and Iyengar 2004; Botti and McGill 2006, 2011; Botti, Orfali, and Iyengar 2009). Yet it also seems likely that people might sometimes value freedom from choice, especially when they worry that they might choose poorly and thus regret their decision. Decision makers facing difficult choices often postpone them, take the “easy” way out by choosing defaults or status quo options, or even opt out of making a decision altogether (Anderson 2003). But there are times when choosing nothing is not an option, or when walking away empty-handed is unappealing. Thus far, research on choice avoidance has largely focused on the choice and the individual making it, rather than the broader social context in which that choice is situated. Consequently, it has given less consideration to the idea that other people can serve as a resource to a decision maker, and that delegating choices to others may be an often-utilized strategy for avoiding the burden of responsibility associated with making a difficult decision and an appealing alternative to walking away from a difficult choice without anything to show for it. Although little is known about when and why people might delegate choices to others, some qualitative research has explored who delegates shopping decisions, to whom, and with what benefits. This research has typically focused on formal transactions in which consumers hire professional agents to make shopping decisions on their behalf to get functional benefits (e.g., expertise) and symbolic benefits (e.g., status; Forsythe, Butler, and Schaefer 1990; Fuller and Blackwell 1992; Pratt 1981; Solomon 1987; Stern, Solomon, and Stinerock 1992; Stinerock, Stern, and Solomon 1991). The present research broadens the traditional conceptualization of delegation to reflect everyday choices and choice contexts, shows that surrogate usage is not predicated on professional expertise, and demonstrates that delegation is often motivated by factors aside from functional and symbolic benefits. Consumers readily recruit others in their immediate social context—salespeople, waiters, friends, even strangers—to make difficult decisions on their behalf so that they do not have to bear responsibility for choosing and feel at fault if the choice is suboptimal. This research examines when consumers delegate, why they do so, and what this means for purchases and sales when delegation is or is not an option. DETERMINANTS OF DELEGATION If the desire to make one’s own choices is often so strong, when might people prefer to delegate? Delegation can provide a number of benefits that may make it an appealing alternative to choosing oneself. For one, delegation enables people to avoid the effort of choosing. Perhaps people delegate choices when the consequences feel unimportant and putting a lot of time and energy into making a decision does not seem worthwhile. Another benefit of delegation is that it allows people to put choices in the hands of people who might have special expertise. Perhaps people delegate when they believe that the potential surrogate has knowledge or experience that would make him or her more capable of making a good choice. But even more fundamentally, delegation allows people to transfer some of the responsibility for the decision to someone else. As we test here, perhaps people delegate when they want to avoid feeling or being at fault for potentially making a less-than-ideal decision. This benefit may be particularly appealing when choices feel difficult and the chance of choosing poorly—and thus regretting one’s choice—seems high. If concerns about responsibility and regret are determinants of the decision to delegate, examining whether people delegate more when faced with a difficult choice is a good place to start, because choice difficulty is known to be tied to regret (Sugden 1985; Zeelenberg 1999). Choice difficulty undermines people’s confidence that they will select the best option (Dhar 1996, 1997a, 1997b; Dhar and Nowlis 1999; Steffel and Williams 2015) and prompts people to try to mitigate the bad feelings they might experience if they choose suboptimally (Zeelenberg 1999). In their conflict-theory model of decision making, Janis and Mann (1976) hypothesized that people who are facing a challenging decision are more likely to practice “defensive avoidance”—procrastinating, ignoring, or passing responsibility for a decision to someone else. Indeed, choice difficulty has been shown to lead to a variety of forms of choice avoidance (Anderson 2003). People avoid difficult choices by postponing them to a later point in time, either by gathering more information, seeking additional alternatives, or simply mulling it over (Dhar 1996, 1997b; Dhar and Nowlis 1999; Luce 1998; Novemsky et al. 2007; Tversky and Shafir 1992). People also avoid difficult decisions by retaining a status quo or default option (Luce 1998; Redelmeier and Shafir 1995; Tversky and Shafir 1992). Or, at times, they opt not to choose anything at all (Dhar 1997a; Iyengar and Lepper 2000). Although there has yet to be a direct test of whether difficulty drives delegation, there is correlational evidence to suggest that people might also be more likely to delegate when choices are challenging. Tetlock and Boettger (1994) found a correlation between participants’ ratings of choice difficulty and the likelihood that they would invite another party to review and potentially revise a decision they have made. Among consumers utilizing wardrobe consultants, those reporting the greatest difficulty in making buying decisions were most likely to let a wardrobe consultant make the final decision of what to purchase (Fuller and Blackwell 1992). Further, patients who report being afraid of making the “wrong” decision were more likely to prefer that their doctors make treatment decisions on their behalf (Charles et al. 1998; Kenny et al. 1999). Thus, we hypothesize that increasing the feeling of choice difficulty will lead people to delegate. H1: As the subjective difficulty of making a choice increases, people are more likely to delegate the decision to another person. AVOIDING RESPONSIBILITY AND REGRET We suspect that difficulty increases the extent to which people worry about being responsible for making a suboptimal decision, and encourages them to reduce or even evade that responsibility. Disappointment and regret both capture a person’s reaction to having made a bad choice. What differentiates them is that people feel disappointment when a choice does not turn out as well as they had hoped, regardless of the cause of the failure or their hand in the choice, but they feel regret if they bear responsibility for a less-than-optimal outcome (Zeelenberg et al. 2000). That people feel responsible for the outcomes of choices they actively make is well known, having been studied in contexts like personal control (DeCharms 1968), cognitive dissonance (Festinger 1964), and regret itself (Ordóñez and Connolly 2000). Indeed, the feeling of responsibility, rather than the act of choosing, is likely to determine one’s emotional reaction to a decision, including satisfaction and regret (Botti and McGill 2006). Further, one can feel these unpleasant emotions in prospect of a decision, not just in retrospect. Anticipated disappointment is felt when someone is worried that the choice outcome will not be as good as their expectations or hopes. Anticipated regret is felt when someone is worried that the option they choose will not be as good as the other options they choose to forgo, and it will be their fault for having chosen it. Not only do people feel responsible after having made a choice, they also worry about feeling responsible in anticipation of making it. Anticipated regret has been shown to contribute to other forms of choice avoidance: namely, it makes people more inclined to choose default or status quo options (Baron and Ritov 1994; Park, Jun, and MacInnis 2000; Simonson 1992) and increases inaction inertia (Tykocinski and Pittman 1998). Given what we know about the interplay between responsibility, regret, and choice avoidance, we suspect that when a decision is difficult and it is unclear whether one will make the best choice, decision makers will worry about feeling responsible for the outcome of the choice and thus regretting their role in it, which will make delegating that choice to someone else an appealing option. Note that it is possible that anticipated disappointment may also prompt choice avoidance, as it may lead to feelings of powerlessness and a need to escape or do nothing (Zeelenberg, van Dijk, and Manstead 1998), and lower expectations about how well the decision might turn out (van Dijk, Zeelenberg, and van der Pligt 2003). However, although both of these anticipated emotions could encourage delegation, anticipated regret is more likely to directly prompt delegation, as it may provide a means of escaping regret by reducing one’s own responsibility for a choice and passing it to a surrogate. Anticipated disappointment is less likely to directly prompt delegation, as resolving a decision via delegation does not eliminate (and could conceivably increase) the possibility that the choice might turn out badly. Note, too, that it is also possible that both may play a role, or that disappointment may even precede regret in prompting delegation; after all, one has to have a choice go badly in order to regret having made it. H2a: The more difficult a choice is to make, the more likely people are to anticipate that they might feel disappointment and regret if they were responsible for choosing a suboptimal option. H2b: Although anticipated disappointment and regret may both contribute to delegation, only anticipated regret directly leads people to delegate choices to others. One important assumption here is that the expected benefit of delegation is the reduction of responsibility. Do people expect to feel less responsible when they delegate? One could imagine that responsibility might remain with the original decision makers, or might even increase, once they have delegated a choice, because they are not only responsible for the outcome of the choice but also for asking someone else to decide for them. Alternatively, as we predict, delegation may enable people to avoid feeling responsible for choice outcomes by shifting who ultimately determines the outcome and allowing them to attribute that responsibility to another person. To understand the drivers of delegation, it is important to know whether people do expect to retain responsibility for choosing even when they delegate. PILOT STUDY This pilot study explored whether people believe that they can reduce their responsibility for a choice outcome by delegating that choice. We recruited 197 participants via Mechanical Turk, setting the target sample size such that we would have about 200 participants total. One participant did not complete all of the dependent measures, yielding a sample of 196. We presented participants with a difficult choice between two flavors of ice cream that were similarly appealing to them. We determined these flavors by asking participants to rank a set of 14 ice cream flavors, and then we told them that only their 4th and 5th ranked flavors were available. Participants next imagined having to decide between the two available flavors, when Jenny, the server, offered to help them choose. With this in mind, they were asked to: “Imagine that you had chosen an ice cream flavor yourself without any help from Jenny. Unfortunately you did not like the ice cream you chose.” They then rated, “How much would you feel that YOU are responsible for the fact that you didn’t enjoy the flavor?” and “How much would you feel that JENNY is responsible for the fact that you didn’t enjoy the flavor?” on scales from 1 = not at all to 7 = very much. Then, participants imagined that, instead, they had asked Jenny to give them the ice cream flavor she thought they would like best, that they did not like the ice cream she chose, and to ascribe responsibility for that outcome to themselves and Jenny. The order in which participants imagined choosing themselves or delegating and the order in which participants ascribed responsibility to themselves and Jenny were counterbalanced. This pilot showed that delegating does allow people to feel less personally responsible for choice outcomes by enabling them to attribute responsibility to another person. We found a significant interaction between choice method and target (F(1, 195) = 120.22, p < .001, ηp2 = .38), such that participants felt less personally responsible for the fact that they did not enjoy the flavor of ice cream they received when they delegated the choice to Jenny than when they chose themselves and attributed more responsibility for the choice outcome to Jenny when they delegated the choice to her than when they chose themselves. Additionally, participants ascribed more responsibility overall to themselves than to Jenny (F(1, 195) = 281.79, p < .001, ηp2 = .59). The effect of choice method (choosing oneself or delegating) on attributions of responsibility was not significant (F(1, 195) = 2.11, p = .15, ηp2 = .01). See table 1. Table 1 Attributions of Responsibility for Choice Outcomes When the Choice Was Made Oneself or Delegated to Another Person   Made choice oneself Mean (SD)  Delegated choice to Jenny Mean (SD)    How responsible I am  5.91 (1.66)  4.52 (1.88)  t(195) = 9.86, p < .001, d = 1.76  How responsible Jenny is  1.69 (1.47)  3.27 (1.97)  t(195) = 10.15, p < .001, d = 2.05    Made choice oneself Mean (SD)  Delegated choice to Jenny Mean (SD)    How responsible I am  5.91 (1.66)  4.52 (1.88)  t(195) = 9.86, p < .001, d = 1.76  How responsible Jenny is  1.69 (1.47)  3.27 (1.97)  t(195) = 10.15, p < .001, d = 2.05  Decision makers do seem to perceive delegation as an effective way to reduce the responsibility they feel for a choice by transferring some of that responsibility to another person. Participants who imagined delegating a decision to a salesperson anticipated feeling less responsible for the outcome of that decision and attributed more responsibility to the salesperson than those who imagined making the choice themselves. Given that delegating allows people to expect to feel less personally responsible for choice outcomes by enabling them to attribute some of that responsibility to another person, we can now turn to testing whether consumers do opt to delegate or whether they prefer other ways out of difficult decisions. DELEGATION VERSUS OPTING OUT Delegation is not the only means by which people can avoid making difficult decisions. Another common way in which people escape tough choices is by refusing to choose at all. Whereas delegation and opting out of choosing may both be appealing when either one (but not the other) is available as an alternative to making a difficult decision oneself, delegation may be a uniquely attractive alternative when choices are difficult but choosing is necessary or desirable and when both delegation and opting out are available alternatives. One potentially important way in which delegation differs from opting out is that it enables people to avoid feeling at fault if the choice outcome is suboptimal while still allowing them to get something out of the choice, whereas opting out necessitates that a decision maker walk away empty-handed. If people’s primary motive is to avoid responsibility and regret for a less-than-ideal outcome of a decision while still leaving open the possibility of a good outcome, then delegation may be uniquely appealing. If, on the other hand, people’s motive is to guarantee that they avoid an unappealing outcome entirely, then opting out may be preferred to delegation. H3a: Delegation (vs. opting out of choosing or choosing oneself) will be more attractive as a means of deciding as choices become more difficult. H3b: Opting out (vs. delegation or choosing oneself) will be more attractive as a means of deciding as the choice options become less appealing. In most contexts, choosing or walking away are not typically people’s only options. Rather, consumers often have the option of asking other people—friends, salespeople or other service professionals, even strangers—to help them choose. Providing consumers with surrogate decision makers to whom they can delegate may make them less likely to opt out of making a purchase in situations in which they might otherwise walk away empty-handed. For example, a consumer faced with a difficult decision between equally attractive product offerings may be less likely to leave without making a purchase if he or she can resolve his or her decision conflict by seeking decision support from a salesperson or shopping companion. Many choices are hard to make because the options are equally appealing (Williams, Gneezy, and Armor 2015), and choosing one will mean forgoing the appealing aspects of the other. Ironically, this should also mean that people should be happy with their chosen option no matter which one they choose, and yet this difficulty can lead them to put off choosing or to walk away entirely. For example, people are less likely to make a purchase from a large than a small array of options, despite the fact that more options should only increase the likelihood that one of those options should prove appealing (Iyengar and Lepper 2000), and they are more likely to opt out of a choice with options that are closer rather than more distant in appeal (Dhar 1997b), when they should be more indifferent between the options in the former case. People might be more likely to make a purchase in such cases if they have the option of ceding responsibility for choosing to another person. Thus, providing opportunities for consumers to delegate may allow consumers to get the benefits of choosing without the burdens and enable businesses to increase purchases by making it less likely that consumers will opt out of decisions and not make a purchase. H4: Consumers are less likely to walk away from a difficult decision empty-handed when they have the option to delegate than when they do not. THE PRESENT RESEARCH The present research demonstrates that consumers are prone to delegate decisions they find to be difficult, for which the chance of regretting the outcome seems higher and thus off-loading responsibility is more appealing. Delegation, like other methods of choice avoidance, allows people to avoid choosing between options and thus possibly regretting the option they chose, but unlike other methods of choice avoidance, provides the added benefit of also resolving the decision itself. Delegation can thus also benefit businesses, which may find that making a surrogate available can reduce the number of customers who find themselves unable to decide and who might otherwise walk away from a transaction empty-handed. We test whether people delegate decisions they might regret by manipulating how difficult and thus potentially regret-provoking choices are. Choice difficulty can sometimes arise directly from the objective features of the choice options. For example, choices between comparably attractive options produce more choice difficulty than choices for which one alternative is much more attractive than the other (Brehm 1956; Festinger 1964; Tversky and Shafir 1992). Likewise, choices that involve making tradeoffs between valued attributes tend to feel more difficult than those that do not (Luce 1998). Additionally, choices for which there are many available options tend to feel more difficult than choices for which there are few options (Iyengar and Lepper 2000). Choice difficulty can also arise independently of the content of the choice options, such as from the fluency by which the choice options are presented (Alter and Oppenheimer 2009), like a difficult-to-read versus easy-to-read font (Novemsky et al. 2007). To show that the feeling of difficulty drives choice delegation as opposed to other factors that might be associated with any particular instantiation of difficulty, we manipulate difficulty in a variety of ways—namely, via set size, relative attractiveness, tradeoffs, and linguistic fluency. In eight experiments, we examine the antecedents of delegation and the consequences of having the option to delegate. Experiments 1a, 1b, and 1c serve as initial demonstrations that consumers are more likely to delegate choices that feel difficult than choices that feel easy. Experiment 2 shows that this is the case for both highly important and less important decisions. Experiment 3 shows that people are more likely to delegate to experts than nonexperts, but the tendency to delegate difficult decisions is not predicated on expertise: consumers are more likely to delegate choices that feel difficult both when the potential surrogate is an expert and when he or she is not. Experiment 4 suggests that the desire to avoid responsibility for a poor outcome, rather than to simply avoid the poor outcome itself, underlies delegation. It also shows that anticipated regret, but not anticipated disappointment, predicts delegation of difficult choices. Finally, experiments 5 and 6 compare delegation to other forms of choice avoidance like opting out of choosing. Experiment 5 tests whether people prefer delegation over opting out as a means of handling a difficult decision, and whether the appeal of the choice set affects the tendency to delegate or opt out. Experiment 6 examines whether adding the option to delegate reduces opting out in situations in which people are otherwise prone to leaving empty-handed. EXPERIMENTS 1A, 1B, AND 1C: DOES DIFFICULTY PROMPT DELEGATION? Our initial studies examine whether, despite a general preference for making their own choices, people do at times prefer to have others choose on their behalf, and do so more often when choices feel difficult. Experiment 1a tests this prediction in an experiment with real consequences. Students chose one of two sets of earphones that were pretested to be close or distant in appeal to use in an experiment and keep as a gift. Participants could pick which set of earphones they wanted or have the experimenter choose a set for them. In experiment 1b, also in an experiment with real consequences, participants chose a jelly bean to sample in a taste test from a small or large array of options, another manipulation of choice difficulty. Participants could pick which flavor they wanted to taste or have the experimenter choose for them. In experiment 1c, participants considered the important decision of whether to undergo surgery to treat a life-threatening injury in a hypothetical medical scenario. Additionally, to examine whether the feeling of choice difficulty is a sufficient catalyst for choice delegation, we held constant the content of the target choice and manipulated difficulty superficially via the fluency of the language with which the options were described (Alter and Oppenheimer 2009). All three experiments show that participants are more likely to delegate when choices are difficult than easy, even when those choices have enduring, immediate, or important consequences. Experiment 1a Method Participants Undergraduates (N = 366) at the University of California, San Diego, participated in exchange for credit in marketing and other business classes. The sample size was set a priori to encompass one week’s worth of participants in the lab. In this and all subsequent studies, we report all measures, conditions, and data exclusions. Procedure Under the guise of an experiment about rating advertisements, participants learned that they would be receiving a free set of headphones they could use to listen to an ad. When they entered the lab, each participant had an order form at their desk with the two earphone options available to them. In the easy condition, one was a silver pair of earbuds and the other was a black pair of over-the-head headphones. In the difficult condition, one was a pair of silver earbuds and one was a pair of grey earbuds. Pretest participants at the same university (N = 320) rated the choice between the headphones and the earbuds to be easier than the choice between the two pairs of earbuds on a scale ranging from 1 = very easy to 7 = very difficult (M = 1.94, SD = 1.28 vs. M = 2.61, SD = 1.59, paired t(319) = 7.69, p < .001, d = 1.04). The conditions alternated between sessions so that everyone in one session would have the same two options. At the beginning of each session, the experimenter informed participants that: “In today’s session, you will be asked to listen to an advertisement and to evaluate its effectiveness. You will listen to the advertisement using a set of earphones, and as an incentive, you will get to keep the earphones that you use. You can either choose whichever earphones you prefer or you can opt for me to give you the pair of headphones that I think are best. Before you start today’s surveys, please make your selections using the earphone order form provided at your station.” Participants made their choices on the order form, indicating either that 1) they wanted to choose a set of earphones themselves and write down which set they wanted, or 2) they wanted to ask the experimenter to give them the earphones that the experimenter thought were best. The experimenter then went around the room, either distributing the earphones that participants chose or providing earphones when the participants delegated (the silver earbuds, in both conditions). Experiment 1a Results Participants were more likely to delegate the choice of earphones to the experimenter when presented with a difficult choice between a silver pair of earbuds and a grey pair of earbuds (28%) than when presented with an easy choice between a silver pair of earbuds and a black pair of over-the-head headphones (11%; χ2(1, N = 366) = 16.72, p < .001, ϕ = .21). Experiment 1b Method Participants Volunteers (N = 200) were recruited on campus at the University of Cincinnati to participate in exchange for the opportunity to taste a gourmet jelly bean. The target sample size was set a priori such that we would have about 100 participants per condition. Procedure Two research assistants who were blind to the hypothesis invited participants to engage in a taste test of gourmet jelly beans. Participants received a menu that contained five flavors (small set condition) or 25 flavors (large set condition). There were four versions of each menu: in the large set condition, the order of the flavors varied, and in the small set condition, the subset of flavors that was offered varied. Pretest participants at the same university (N = 141) were randomly assigned to see either the short or the long menu and rated how easy or difficult it would be to decide which flavor to taste, on a scale ranging from 1 = very easy to 10 = very difficult. Participants indicated it would be easier to choose a flavor from the small set (M = 3.52, SD = 2.78) than the large set (M = 5.64, SD = 3.14; t(139) = 4.23, p < .001, d = .71). Participants were asked, “Would you like to choose which flavor you will taste? Or would you like for me to choose a flavor for you?” Once they had tasted their jelly bean, all participants rated how much they liked the flavor they tasted on a scale ranging from 1 = not at all to 10 = very much. Experiment 1b Results Participants were marginally more likely to delegate the choice of which jelly bean flavor to taste to the experimenter when presented with a difficult choice between many options (28%) than when presented with a more manageable choice between fewer options (17%; χ2(1, N = 200) = 3.47, p = .06, ϕ = .13). This difference did not seem to be due to any variation in how appealing the flavor options in each set were: participants liked the jelly beans they tasted equally well in the large set condition (M = 7.57, SD = 1.69) and small set condition (M = 7.52, SD = 1.62; t(198) = –.21, p = .83, d = .03). Experiment 1c Method Participants Undergraduates (N = 296) at the University of California, San Diego, participated in exchange for course credit. The sample size was set a priori to reflect one week’s worth of participants in the lab. Procedure Participants imagined that they had sustained a neck injury in a car accident and had to decide whether to undergo surgery (see the appendix for full stimulus materials). We manipulated the subjective ease or difficulty of the decision by presenting the treatment options in either plain language (fluent condition) or medical jargon (disfluent condition). A SMOG analysis (McLaughlin 1969) indicated that the plain language version of the scenario was written at an eighth-grade reading level, the maximum reading level recommended by the International Patient Decision Aid Standards (Elwyn et al. 2006), and the medical jargon version was written at a 13th-grade reading level, the typical reading level for health-related materials (Rudd, Moeykens, and Colton 1999). Pretest participants from the same university (N = 47) rated the scenario as easier to read when it was in plain language (M = 2.09, SD = 1.41) than when it was in medical jargon (M = 5.23, SD = 1.55; paired t(46) = 10.25, p < 001, d = 2.12). Additionally, most participants in another pretest at the same university (N = 127) answered three comprehension questions correctly in both the medical jargon condition (86%) and plain language condition (91%; χ2(1, N = 127) = .63, p = .43), suggesting that participants were equally capable of understanding the scenario regardless of the language that was used. Participants indicated whether they would want to 1) choose a treatment themselves and identify which treatment they preferred, or 2) ask their doctor to choose a treatment for them. On the next page, participants responded to the same three multiple-choice questions from the pretest to determine how closely they were attending to the scenario. Experiment 1c Results Participants were more likely to delegate their choice of treatment to their doctor when we made the choice feel subjectively difficult by describing the same options in medical jargon (38%) versus plain language (20%; χ2(1, N = 296) = 12.03, p = .001, ϕ = .20). Interestingly, participants in the disfluent condition did not recall the details of the scenario as well as did those in the fluent condition (paired χ2(1, N = 296) = 30.35, p = .001, ϕ = .32): 72% of participants in the fluent condition answered all three comprehension questions correctly, but only 41% of those in the disfluent condition did so. We suspect that the difference in performance on the comprehension check between the pretest and experiment may be attributed to the fact that the comprehension check appeared on the same page as the scenario in the pretest and on a separate page from the scenario and choice in the experiment. We interpret the results of the comprehension check in the pretest to mean that participants were equally capable of comprehending the scenario regardless of whether it was in plain language or medical jargon and the results of the comprehension check in the experiment to mean that participants attended less to the scenario when it was disfluent and they had the option of delegating the decision to their doctor. Nonetheless, the influence of fluency on delegation is the same if the analysis is restricted only to those who accurately recalled all of the details of the scenario (χ2(1, N = 167) = 12.93, p = .001, ϕ = .28): 38% of participants delegated when the scenario was described in medical jargon, whereas only 14% delegated when the same information was in plain language, suggesting that the difference in delegation is unlikely to be due solely to comprehension or attention. Discussion Our initial experiments show that people do delegate decisions, and they are more likely to delegate when choices feel difficult. Choice difficulty increases choice delegation, and heightening the subjective feeling of difficulty is sufficient to increase delegation even when the objective content of the choices is the same. Moreover, people are willing to delegate difficult choices even when those choices have enduring, immediate, or important consequences. Although we tried to examine delegation in the context of decisions that are meaningful and consequential, an alternative interpretation of our findings thus far might be that people delegate difficult decisions because they do not wish to exert too much effort over choices that seem abstract or inconsequential to them. Perhaps the factors in studies 1a–c that made the choices feel difficult merely served to increase the necessary effort while simultaneously lowering the payoff of investing more effort to identify the best option. In the next experiment, we vary the importance of decisions along with their difficulty to explore the possibility that people delegate simply to avoid making effortful but relatively inconsequential decisions. EXPERIMENT 2: DOES THE IMPORTANCE OF THE DECISION MATTER? Perhaps people delegate difficult decisions not to avoid responsibility for potentially making a bad decision, but because they do not want to invest time and energy into decisions that are not worth the effort. In experiment 2, we used a scenario about choosing a new doctor to directly manipulate the importance of the choice consequences. For some participants, this choice was nonbinding and could be easily changed; for others, this choice was binding and changing was costly. We predict that choice difficulty, but not importance, prompts delegation, such that people delegate more when the choice of doctors feels difficult than when it feels easy, both when the choice seems important and when it seems unimportant. Method Participants Adults (N = 815) were recruited via Mechanical Turk to fill out an online survey for $0.25 Amazon.com credit. The target sample size was set a priori such that we would have about 200 participants per condition. Procedure Participants imagined that they were about to join a new health plan that required them to choose a new physician, using a scenario adapted from Schrift, Netzer, and Kivetz (2011) designed to test the effects of choice difficulty and importance on decision making. The physicians were described along three attributes: 1) office hours that did or did not include evenings and weekends, 2) a three- or 10-day wait time to get an appointment, and 3) home visits included or excluded. In the easy conditions, one physician had better office hours and wait times, the two attributes judged to be more important by participants in prior research. In the difficult conditions, each physician was better on only one of the important attributes, so that participants had to make a tradeoff. Participants in the high-importance conditions imagined that their choice was binding for a year and that switching physicians before the year ended would be difficult and require paying additional fees. Participants in the low-importance conditions imagined that their choice was nonbinding and they could easily switch doctors whenever they wanted without fees. Participants were told, “As you discuss this decision with your current physician, Dr. Brenner, he tells you that, if you are having trouble choosing a physician, he would be glad to choose one for you,” and were asked to indicate whether they would want to 1) choose which of the physicians they would prefer and name who that was, or 2) ask their current physician to choose a new physician for them. Additionally, participants rated how difficult the choice was on a scale ranging from 1 = very easy to 7 = very difficult and how important the decision was on a scale from 1 = not at all important to 7 = extremely important. Results Manipulation Checks As intended, participants rated choices to be more important when their choice was binding for a year and switching physicians would be difficult and costly (M = 4.93, SD = 1.29) than when their choice was not binding and they could easily switch physicians without penalty (M = 4.56, SD = 1.34; F(1, 803) = 15.41, p < .001, ηp2 = .02). Neither the difficulty manipulation (F(1, 803) = .13, p = .72, ηp2 < .001) nor the interaction between the difficulty and importance manipulations reliably affected rated importance (F(1, 803) = 2.92, p = .09, ηp2 = .004). Also as intended, participants rated the choice to be more difficult when they had to choose between physicians who were better on only one of the important attributes, meaning that they had to trade off the attributes (M = 3.69, SD = 1.65), than when one of the physicians was better on both attributes (M = 3.01, SD = 1.60; F(1, 803) = 34.93, p < .001, ηp2 = .04). Neither the importance manipulation (F(1, 803) = .08, p = .78, ηp2 < .001) nor the interaction between the difficulty and importance manipulations reliably affected rated choice difficulty (F(1, 803) = .02, p = .88, ηp2 < .001). Delegation Participants were more likely to delegate their choice of a new physician to their current physician when the choice felt difficult (44%) than when it felt easy (29%; Wald’s χ2 = 29.32, p < .001, odds ratio = 1.98), both when the choice was important (41% vs. 29%; χ2 = (1, N = 406) = 5.55, p = .02, ϕ = .12) and when it was less important (48% vs. 28%; χ2 = (1, N = 409) = 17.62, p < .001, ϕ = .21). Delegation did not vary as a function of decision importance (Wald’s χ2 = .44, p = .51, odds ratio = 1.10) or the interaction between difficulty and importance (Wald’s χ2 = 1.65, p = .20, odds ratio = 1.46). Mediation We next examined the process underlying the effect of our difficulty manipulation on delegation. A mediation model (PROCESS model 4) with our difficulty manipulation as the independent variable, rated difficulty as the mediator, and delegation as the dependent variable yielded significant mediation via rated difficulty (95% CI = –.65, –.30). Namely, our difficulty manipulation increased difficulty (β = –.68, SE = .11, t = –5.92, p < .001), and difficulty increased delegation (β = .68, SE = .06, Z = 11.76, p < .001). We did not include rated importance in the mediation model since it was unaffected by our difficulty manipulation; if included in the model as a parallel mediator, the pattern of results is the same, and rated importance does not mediate the relationship between the difficulty manipulation and delegation. Discussion Choice difficulty, but not importance, influences people’s likelihood to delegate. People are more likely to delegate when choices feel difficult than when they feel easy, and this does not vary depending on how important people feel the choices are. These findings are consistent with the notion that people delegate to avoid anticipated regret—conceptualized as the expectation that one will feel responsible for having chosen a suboptimal outcome and forgoing a better one (Zeelenberg et al. 2000)—and inconsistent with the alternate explanation that the degree to which people deem the choice worth deliberating about determines their likelihood of delegating. In our studies thus far, one could imagine that participants thought that the experimenter in experiments 1a and 1b or the doctor in experiments 1c and 2 might be an expert or at least know more about the options than they did. One might wonder whether people delegate not to avoid responsibility for potentially making a bad choice, but to put those decisions in more capable hands. Alternatively, if delegating difficult choices enables people to avoid feeling responsible for the choice outcome, then it might not matter whether the potential surrogate is an expert or not because delegation can still allow them to feel less responsible for the final outcome regardless of who else chooses. The next study explores whether consumers delegate even when the potential surrogate does not have special expertise pertinent to the decision. EXPERIMENT 3: DO SURROGATES NEED TO BE EXPERTS? Much of the past research on choice delegation in marketing has focused on cases in which consumers hire professionals with special expertise to make choices on their behalf, implicitly assuming that people delegate only to more knowledgeable others. Yet expertise may not be essential for delegation. Experiment 3 examines whether the expectation that the surrogate will have knowledge pertinent to the choice is a necessary determinant of delegation. We expected that participants would be more likely to delegate to surrogates with relevant expertise than to surrogates without expertise. But more importantly, we predicted that, regardless of the surrogate’s expertise, participants would be more likely to delegate when choices felt difficult than when they felt easy, choosing to transfer some of the responsibility for such choices to someone else even when the potential surrogate did not have special expertise pertinent to the judgment. Method Participants Adults (N = 403) were recruited via Mechanical Turk to fill out an online survey for $0.15 Amazon.com credit. The target sample size was set a priori such that we would have about 100 participants per condition. Procedure Participants imagined that they were out shopping with a friend and were interested in renting a movie from a movie rental kiosk featuring foreign films. Participants were told that their friend would not be able to watch the movie with them but was with them to help with another errand. Participants were presented with 10 foreign films, all in a foreign language with English subtitles (the original language was written in italics at the end of each movie’s description). Foreign films were chosen to minimize the likelihood that participants would be familiar with the movies. Indeed, 87% of participants indicated that they had not seen any of the movies. For each film, participants were provided with reformatted descriptions and images of the DVD covers that appeared on Netflix. Participants ranked the movies from 1 to 10, with the movie they were most interested in seeing at the top of the list (#1) and the movie they were least interested in watching at the bottom of the list (#10). Participants were able to click on each movie to drag and drop the movies into the appropriate order to indicate their rankings. On the next page, participants were told, “Once you have a chance to look over the movies, you realize that the kiosk is having a busy day, and the only two movies in the foreign film category that are currently available are these two movies.” Participants in the easy condition were presented with their fourth- and 10th-ranked movies, and those in the difficult condition were presented with their fourth- and fifth-ranked movies. Participants’ first-, second-, and third-ranked movies were not presented so as to avoid movies that might be more familiar to participants. In a pretest, 25 Mechanical Turk participants were shown both pairs of movies and, for each pair, rated how easy or difficult it would be to decide which of the two movies they would rather rent on a scale ranging from 1 = very easy to 10 = very difficult. Participants indicated that the decision would be easier to make when the movies were ranked further apart (M = 3.44, SD = 2.68) than closer together (M = 5.04, SD = 2.75; paired t(24) = 3.12, p = .005, d = 1.32). Once they were presented with the two movies they would be choosing between, participants in the expert condition were told, “When your friend sees you thinking about the decision, they tell you, ‘I have seen both of these movies, and I’d be glad to help you pick one if you need me to.’” Those in the nonexpert condition were instead told, “When your friend sees you thinking about the decision, they tell you, ‘I haven’t seen either of these movies, but I’d be glad to help you pick one if you need me to.’” In a pretest, 40 participants from Mechanical Turk were shown either the expert or nonexpert version of the scenario and were asked to rate, “To what extent do you think your friend has expertise that is relevant to this decision?” on a scale ranging from 1 = not at all to 7 = very much. Participants thought that their friend had greater expertise relevant to the decision in the expert condition (M = 5.70, SD = 1.17) than in the nonexpert condition (M = 3.75, SD = 1.62; t(38) = 4.36, p < .001, d = 1.41). Finally, participants were asked to indicate whether they would want to 1) choose which of the two movies they would rent on their own and name what it was or 2) ask their friend which of these two movies they should rent and rent that one. Results Participants were more likely to delegate to experts than nonexperts, but even so, delegation was not predicated on expertise. A logistic regression examining the effects of choice difficulty and surrogate expertise on preferences for choosing or delegating indicated that participants were more likely to delegate to a friend when that person was knowledgeable about the choice alternatives (68%) than when they were not (42%; Wald’s χ2 = 29.31, p < .001, odds ratio = 3.23). Additionally, as predicted, participants were more likely to delegate the choice to a friend in the difficult condition (66%) than in the easy condition (44%; Wald’s χ2 = 21.84, p < .001, odds ratio = 2.75). However, they did so regardless of the potential surrogate’s level of expertise (Wald’s χ2 = .01, p = .90, odds ratio = .95): participants were more likely to delegate the choice to a friend when the choice was difficult than when it was easy, both when that person had relevant expertise (79% vs. 57%; χ2 = (1, N = 201) = 11.37, p < .001, ϕ = .24) and when he or she did not (54% vs. 30%; χ2 = (1, N = 202) = 11.07, p < .001, ϕ = .23). Discussion People are more likely to delegate decisions to people with knowledge of the decision options than those without. However, delegation is not contingent upon the expectation that the surrogate decision maker has additional information or special expertise pertinent to the choice. People are more likely to delegate when choices feel difficult than when they feel easy, regardless of whether the potential surrogate has additional knowledge or experience regarding the choice options. This is consistent with the notion that what makes people more likely to delegate difficult than easy decisions is avoiding responsibility rather than seeking better insight. If people delegate to avoid responsibility for a decision, what might they perceive to be the benefits of avoiding responsibility for a decision? Perhaps people wish to avoid feeling disappointed if the choice outcome is less than ideal, so they delegate in hopes that the surrogate decision maker will make a better decision than they themselves would make. However, delegation may be driven less by the decision maker’s worry about the actual outcome, but instead more by their concern about their role in bringing it about. Although a nonexpert surrogate may not have special insight into a decision, they may have other qualities that make them an appealing alternative to choosing oneself—namely, the ability to take on some of the responsibility of making the choice. People may wish to avoid the regret that they might feel if they were responsible for making a poor choice, and they delegate in order to cede decision responsibility to another person. In the next study, to further probe what prompts people to delegate difficult decisions, we more directly assess the role of responsibility in delegation by examining the role of anticipated regret, in which responsibility for a decision outcome plays a role, and of anticipated disappointment, in which it does not, in the decision to delegate. EXPERIMENT 4: REGRET VERSUS DISAPPOINTMENT What prompts people to delegate—is it the desire to avoid the disappointment associated with a bad outcome, independent of whether they choose it or not, or is it the desire to avoid the regret associated with being responsible for bringing about that bad outcome? In experiment 4, we both manipulated choice difficulty by varying the difference in utility among the available options in the set (in this case, making them close or distant in appeal) and manipulated the overall attractiveness of the available options by varying the overall utility of the set of options, (making them all either appealing or unappealing). We also directly measured anticipated disappointment and regret in order to test whether they play a mediating role in prompting delegation. If people delegate merely to avoid disappointment, then they may be especially likely to delegate when the options are all unappealing and there is a good chance that the outcome of a decision will be unpleasant. If, as we predict, people delegate primarily to avoid regret, however, then they may delegate most when the options are close in appeal and it seems quite possible they might wish they had chosen the other option. Method Participants Adults (N = 823) were recruited via Mechanical Turk to complete an online survey for $.25 in Amazon.com credit. The target sample size was set a priori such that we would have about 200 participants per condition. Six participants did not complete the regret and disappointment measures. Procedure Participants were asked to imagine that they came across an ice cream truck selling 14 flavors of gourmet ice cream (The ice cream flavors, images, and descriptions came from Jeni’s Splendid Ice Cream [jenis.com].) Participants ranked the flavors from most appealing to least appealing by dragging and dropping the images into the appropriate order. Next, participants were told that only two of the 14 flavors were currently available. Participants in the easy attractive condition were shown their first and seventh ranked flavors, those in the easy unattractive condition were shown their eighth and 14th ranked flavors, those in the difficult attractive condition were shown their fourth and fifth ranked flavors, and those in the difficult unattractive condition were shown their eighth and ninth ranked flavors (note: pretesting indicated that participants found a choice between their two top-ranked flavors to be so easy that the difficult attractive condition was indistinguishable from the easy attractive condition, so to differentiate those conditions, we presented participants in the difficult attractive condition with slightly lower-ranked options). Participants were told, “As you think about the ice creams, you strike up a conversation with Jenny, the woman behind the counter. Jenny mentions that, if you are having a hard time choosing a flavor, that she has tried all of the flavors herself and would be glad to give you the flavor that she thinks you would like the best.” Participants then rated the extent to which they would prefer to choose themselves or delegate the decision to Jenny on a seven-point scale ranging from 1 = I would prefer to choose a flavor myself without any help from Jenny to 7 = I would prefer to have Jenny give me the flavor she thinks I would like the best. Next, to assess anticipated regret and disappointment, participants were asked, “If you had to choose by yourself which of the two available ice cream flavors to order, think about what would be going through your mind as you consider which flavor to choose. In particular, try to anticipate how you might feel if you are unhappy with the flavor that you choose.” Participants then responded to four regret items and four disappointment items, alternating between regret and disappointment, on a seven-point scale ranging from 1 = not at all to 7 = to a very great extent. Anticipated regret and disappointment items were adapted from Marcatto and Ferrante (2008) and Zeelenberg et al. (1998). See table 2 for all regret and disappointment items. Table 2 Structure Matrix for Regret and Disappointment, Experiment 4 Item:  Component 1: Disappointment  Component 2: Regret  How much might you NOT want to choose either of the two available flavors? (D)  .72  .37  How much might you feel powerless over getting a good flavor? (D)  .82  .53  How much might you feel like there is nothing you can do to ensure that you get a good flavor? (D)  .85  .50  How much might you feel that events that were beyond your control were the cause if you were unhappy with your flavor? (D)  .77  .27  How much might you feel that you should have known better than to choose the flavor that you did? (R)  .43  .76  How much might you think about what a mistake you made in choosing the flavor that you did? (R)  .59  .84  How much might you feel the tendency to kick yourself for choosing the flavor that you did? (R)  .60  .86  How much might you feel responsible if you are unhappy with your flavor? (R)  .11  .60  Item:  Component 1: Disappointment  Component 2: Regret  How much might you NOT want to choose either of the two available flavors? (D)  .72  .37  How much might you feel powerless over getting a good flavor? (D)  .82  .53  How much might you feel like there is nothing you can do to ensure that you get a good flavor? (D)  .85  .50  How much might you feel that events that were beyond your control were the cause if you were unhappy with your flavor? (D)  .77  .27  How much might you feel that you should have known better than to choose the flavor that you did? (R)  .43  .76  How much might you think about what a mistake you made in choosing the flavor that you did? (R)  .59  .84  How much might you feel the tendency to kick yourself for choosing the flavor that you did? (R)  .60  .86  How much might you feel responsible if you are unhappy with your flavor? (R)  .11  .60  NOTE.—Items marked D were intended to measure disappointment, and items marked R were intended to measure regret. Results Manipulation Checks As intended, participants rated choices to be more difficult when the options were similar in appeal (M = 3.38, SD = 1.68) than when one option was preferred much more than the other (M = 2.67, SD = 1.70; F(1, 813) = 37.17, p < .001, ηp2 = .04). They also rated choices to be more difficult when both options were poorly ranked (M = 3.22, SD = 1.74) than highly ranked (M = 2.83, SD = 1.68; F(1, 813) = 11.28, p = .001, ηp2 = .01). There was an interaction (F(1, 813) = 4.23, p = .04, ηp2 = .005), such that the choice was equally difficult regardless of the overall attractiveness of the set when the options were similar in appeal (Mattractive= 3.31, SD = 1.70 vs. Munattractive= 3.46, SD = 1.65; t(407) = .92, p = .36, d = .09), but when one option was preferred much more than the other, the choice was easier when the options were both highly ranked than when they were both poorly ranked (Mattractive= 2.35, SD = 1.51 vs. Munattractive= 2.99, SD = 1.81; t(392.41) = 3.83, p < .001, d = .38, equal variances not assumed). Also as intended, participants rated the choice options to be less attractive when they were both poorly ranked (M = 3.87, SD = 1.66) than highly ranked (M = 5.06, SD = 1.47; F(1, 813) = 121.36, p < .001, ηp2 = .13). There was an interaction (F(1, 813) = 30.71, p < .001, ηp2 = .04), such that the overall attractiveness of the choice set mattered more when participants preferred one option much more than the other (Mattractive= 5.37, SD = 1.37 vs. Munattractive= 3.59, SD = 1.65; t(390.72) = 11.87, p < .001, d = .17, equal variances not assumed) than when the options were similar in appeal (Mattractive= 4.74, SD = 1.50 vs. Munattractive= 4.15, SD = 1.62; t(407) = 3.82, p < .001, d = .38). Whether the options were similar in appeal (M = 4.45, SD = 1.59) or one option was preferred much more than the other (M = 4.48, SD = 1.76) did not influence rated attractiveness (F(1, 813) = .08, p = .78, ηp2 < .001). Disappointment and Regret A principle component analysis with Promax rotation on our anticipated disappointment and regret items yielded two distinct components (component 1 (disappointment) eigenvalue = 3.88; component 2 (regret) eigenvalue = 1.16; all other eigenvalues < .82). See table 2 for structure matrix. Moreover, our disappointment items (alpha = .77) and regret items (alpha = .81) hung together reliably. Therefore, we created composites for disappointment and regret by averaging each set of items together. Delegation Participants were more likely to delegate when the options were similar in appeal (M = 3.76, SD = 2.12) than when they preferred one option much more than the other (M = 3.37, SD = 2.13; F(1, 819) = 7.00, p = .008, ηp2 = .008)—in other words, when the choice was difficult rather than easy. There was also a marginal interaction (F(1, 819) = 3.57, p = .059, ηp2 = .004) such that our manipulation of choice difficulty was more influential when the options were both attractive (Mdifficult= 3.84, SD = 2.09 vs. Measy= 3.16, SD = 2.10; t(411) = 3.25, p = .001, d = .32) than when they were both unattractive (Mdifficult= 3.69, SD = 2.16 vs. Measy= 3.58, SD = 2.13; t(408) = .53, p = .60, d = .05). Overall, though, participants were no more likely to delegate when both options were appealing (M = 3.50, SD = 2.12) than when both options were unappealing (M = 3.64, SD = 2.14; F(1, 819) = .86, p = .36, ηp2 = .001). Mediation A serial mediation model (PROCESS model 6) with our choice difficulty manipulation as the independent variable; attractiveness, difficulty, disappointment, and regret as serial mediators; and delegation as the dependent variable yielded significant mediation via rated difficulty (95% CI = .18, .41), serially via rated difficulty and regret (95% CI = .003, .04), and serially via rated difficulty, disappointment, and regret (95% CI = .001, .03). None of the other indirect paths was significant. See figure 1 for the mediation model. Although we did not have an a priori prediction that disappointment would predict regret, having disappointment precede regret in the model is consistent with the idea that one must first have a bad outcome (one that is disappointing) before one has something to regret. FIGURE 1 View largeDownload slide MEDIATION MODEL, EXPERIMENT 4 *p < .05, ***p < .001 NOTES.—X = choice difficulty manipulation; Y = choice delegation; M1 = rated unattractiveness; M2 = rated difficulty; M3 = anticipated disappointment; M4 = anticipated regret Indirect effect of X on Y through M1 only: 95% CI = (–.03, .02) Indirect effect of X on Y through M2 only: 95% CI = (.18, .41)* Indirect effect of X on Y through M3 only: 95% CI = (–.04, .01) Indirect effect of X on Y through M4 only: 95% CI = (–.05, .01) Indirect effect of X on Y through M1 and M2 in serial: 95% CI = (–.02, .03) Indirect effect of X on Y through M1 and M3 in serial: 95% CI = (–.005, .01) Indirect effect of X on Y through M1 and M4 in serial: 95% CI = (–.006, .003) Indirect effect of X on Y through M2 and M3 in serial: 95% CI = (–.02, .04) Indirect effect of X on Y through M2 and M4 in serial: 95% CI = (.003, .04)* Indirect effect of X on Y through M3 and M4 in serial: 95% CI = (–.03, .01) Indirect effect of X on Y through M1, M2, and M3 in serial: 95% CI = (–.001, .002) Indirect effect of X on Y through M1, M2, and M4 in serial: 95% CI = (–.001, .002) Indirect effect of X on Y through M1, M3, and M4 in serial: 95% CI = (–.005, .01) Indirect effect of X on Y through M2, M3, and M4 in serial: 95% CI = (.001, .03)* Indirect effect of X on Y through M1, M2, M3, and M4 in serial: 95% CI = (–.001, .002) Direct effect of X on Y: 95% CI = (–.20, .36) FIGURE 1 View largeDownload slide MEDIATION MODEL, EXPERIMENT 4 *p < .05, ***p < .001 NOTES.—X = choice difficulty manipulation; Y = choice delegation; M1 = rated unattractiveness; M2 = rated difficulty; M3 = anticipated disappointment; M4 = anticipated regret Indirect effect of X on Y through M1 only: 95% CI = (–.03, .02) Indirect effect of X on Y through M2 only: 95% CI = (.18, .41)* Indirect effect of X on Y through M3 only: 95% CI = (–.04, .01) Indirect effect of X on Y through M4 only: 95% CI = (–.05, .01) Indirect effect of X on Y through M1 and M2 in serial: 95% CI = (–.02, .03) Indirect effect of X on Y through M1 and M3 in serial: 95% CI = (–.005, .01) Indirect effect of X on Y through M1 and M4 in serial: 95% CI = (–.006, .003) Indirect effect of X on Y through M2 and M3 in serial: 95% CI = (–.02, .04) Indirect effect of X on Y through M2 and M4 in serial: 95% CI = (.003, .04)* Indirect effect of X on Y through M3 and M4 in serial: 95% CI = (–.03, .01) Indirect effect of X on Y through M1, M2, and M3 in serial: 95% CI = (–.001, .002) Indirect effect of X on Y through M1, M2, and M4 in serial: 95% CI = (–.001, .002) Indirect effect of X on Y through M1, M3, and M4 in serial: 95% CI = (–.005, .01) Indirect effect of X on Y through M2, M3, and M4 in serial: 95% CI = (.001, .03)* Indirect effect of X on Y through M1, M2, M3, and M4 in serial: 95% CI = (–.001, .002) Direct effect of X on Y: 95% CI = (–.20, .36) Discussion Although both anticipated disappointment and regret may contribute to delegation, only anticipated regret directly leads people to delegate choices to others. This is consistent with the notion that people delegate choices when they are concerned that they might feel responsible for a less-than-optimal outcome. This key benefit of delegation—enabling people to avoid feeling responsible if a choice is suboptimal while still allowing them to get something out of the choice—may make it a uniquely appealing method for resolving difficult decisions, even when other options for resolution are available. Another option that people often have available to them to escape tough choices is to opt out of decisions and refuse to choose anything at all. While both delegation and opting out may be commonly used methods of avoiding difficult decisions when either option is the sole alternative to choosing, when delegation and opting out are both options, delegation may often be a more appealing means of handling difficult decisions than opting out, because it allows people to avoid responsibility and regret while still getting something out of the decision. Opting out, on the other hand, may be preferred to delegation when the options are all unappealing, since it allows people to guarantee that they avoid an unappealing outcome. We explore these predictions in the next experiment. EXPERIMENT 5: DELEGATION VERSUS OPTING OUT In this study, we varied choice difficulty and the overall appeal of the choice options and gave participants the option to choose themselves, delegate, or opt out. Since delegation enables people to avoid responsibility while still allowing them to realize the benefits of a chosen option, people may prefer delegation when the choice is difficult so long as the set of options is somewhat appealing, and especially if those options are highly appealing. And because opting out enables people to avoid the consequences of a choice altogether, people may uniquely prefer to opt out of choosing entirely when all options are unappealing rather than appealing. Method Participants Adults (N = 590) were recruited via Mechanical Turk to complete an online survey for $.25 in Amazon.com credit. The target sample size was set a priori such that we would have about 150 participants per condition. Procedure This study followed the same procedure as in experiment 4, except the regret and disappointment items were removed and participants had three options: 1) “I would prefer to choose a flavor myself without any help from Jenny”; 2) “I would prefer to have Jenny give me the flavor she thinks I would like the best”; and 3) “I would prefer NOT to get any ice cream.” Results A multinomial logistic regression with our choice difficulty and overall attractiveness manipulations as predictors and choice method (choose, delegate, opt out, with choice as the reference category) as the outcome indicated that participants were more likely to delegate than choose themselves when the choice was difficult than when it was easy (42% vs. 32%; Wald’s χ2 = 8.88, p = .003, odds ratio = .48), but not when the options were both appealing versus when they were both unappealing (38% vs. 36%; Wald’s χ2 < .001, p = .99, odds ratio = 1.00). Although the interaction between choice difficulty and overall appeal did not significantly influence participants’ preference to delegate rather than choose themselves (Wald’s χ2 = 2.61, p = .11, odds ratio = 1.76), an examination of the simple effects revealed that participants were more likely to prefer delegation when the options were similar in appeal and choices were difficult (44%) than when they were distant in appeal and choices were easy (27%) when both available options were appealing (Wald’s χ2 = 8.88, p = .003, odds ratio = .48), but participants were equally likely to prefer delegation of easy and difficult choices when both options were unappealing (41% vs. 36%; Wald’s χ2 = .50, p = .48, odds ratio = .83). Perhaps unsurprisingly, participants were more likely to opt out of the decision and not pick an ice cream at all when the options were both unappealing. Because so few people opted out, we ran a Firth penalized maximum likelihood estimation, which showed that opting out was more common in the unappealing conditions than the appealing conditions (9% vs. 1%; Wald’s χ2 = 18.97, p < .001, odds ratio = .11), but was just as common when the choice was difficult than when it was easy (4% vs. 5%; Wald’s χ2 = .49, p = .48, odds ratio = .74). The interaction between choice difficulty and the overall appeal did not significantly influence the likelihood with which participants preferred to opt out rather than choose themselves (Wald’s χ2 = .06, p = .81). Participants opted out more when the options were all unappealing versus when they were all appealing, both when the options were similar in appeal and the choice was difficult (8% vs. 1%; Wald’s χ2 = 9.15, p = .002, odds ratio < .001) and when they were distant in appeal and the choice was easy (10% vs. 1%; Wald’s χ2 = 9.97, p = .002, odds ratio = .13). Discussion People prefer to delegate rather than opt out of difficult decisions, so long as the available options are somewhat appealing. Participants in this experiment were more likely to delegate when the options were similar in appeal rather than when one option was much more appealing than the other, especially when the options were both appealing. And, as one might expect, people were more likely to opt out of choosing when the available options were unappealing. These findings hint at a possible solution for retailers looking to encourage consumers to make a purchase when faced with challenging decisions, a situation in which consumers might otherwise be inclined to walk away empty-handed. If consumers often prefer to delegate rather than opt out of difficult decisions, then giving consumers the option to delegate might reduce or even reverse the tendency for them to opt out of difficult decisions. In the next study, we examine whether consumers are less likely to walk away from difficult decisions empty-handed when given the option to delegate than when their only options are to choose or walk away. EXPERIMENT 6: DELEGATION AS A MEANS OF DECREASING OPTING OUT Shoppers may sometimes be less likely to make a purchase from a large array of options than from a small array, despite the fact that a greater number of options should only increase the likelihood that one of those options is appealing (Iyengar and Lepper 2000). This experiment examines how giving shoppers the option to delegate affects their purchase likelihood when they are presented with choices between many or few products. Participants were presented with a small or large selection of teas. Some participants were asked whether they would like to purchase a tea or opt out of the decision, and other participants were asked whether they would like to purchase a tea, opt out, or ask the salesperson to choose for them. We predicted that participants would be equally likely to make a purchase from a small array regardless of whether they had the option to delegate, but that participants with the option to delegate would be more likely to make a purchase from a large array than those without the option to delegate. Method Participants Undergraduates (N = 151) at the University of Florida participated in exchange for extra course credit. The sample size was set a priori to encompass one week’s worth of participants in the lab. Procedure Participants imagined shopping at a gourmet tea shop. They were randomly assigned to see an array of 30 teas (large set condition) presented in a randomized order or six teas (small set condition) randomly selected from the large set. They saw an image of each tea and its ingredients, price, and popularity rating, with a checkbox alongside. In the salesperson-not-present conditions, participants chose whether they preferred to: 1) “Choose a package of tea yourself, and buy that one (If you select this option, check the box next to the ONE tea you would like to buy),” or 2) “Pass for the time being (If you select this option, leave all boxes EMPTY and continue to the next page).” Others assigned to the salesperson-present condition were told, “A store employee approaches you and asks if you would like any help in choosing a tea,” and were given the option to 1) choose, 2) pass for the time being, or 3) “Tell the store employee which teas you are considering, and buy whichever one he/she recommends (If you select this option, check boxes next to ANY teas that you are considering).” Lastly, participants indicated whether they liked tea and how often they drank it. Nineteen participants were excluded because they did not select the appropriate number of teas given their stated preference to choose themselves, delegate, or opt out, leaving 132 participants. The pattern of results remains the same when these participants were included, as well as when we control for tea liking and drinking. Results We examined how set size and the presence of a salesperson affected whether or not participants opted to purchase tea. A purchase in this study refers to when a participant would leave the shop with a tea, either because they chose a tea on their own or because they delegated the choice to the salesperson. A logistic regression testing the effects of set size and salesperson presence on likelihood of purchase indicated that, overall, participants were equally likely to make a purchase from a large array (59%) and a small array (63%; Wald’s χ2 = .01, p = .92, odds ratio = 1.04), and a greater percentage of participants purchased tea when a salesperson offered to help them choose (74%) than when a salesperson was not present (46%; Wald’s χ2 = 10.22, p = .001, odds ratio = 3.34). Importantly, as predicted, there was a significant interaction between set size and salesperson presence on purchases (Wald’s χ2 = 8.90, p = .003, odds ratio = 11.27). See figure 2. FIGURE 2 View largeDownload slide PERCENTAGE OF PARTICIPANTS WHO PURCHASED TEA BY SET SIZE AND PRESENCE OF THE OPTION TO DELEGATE TO A SALESPERSON, EXPERIMENT 6 FIGURE 2 View largeDownload slide PERCENTAGE OF PARTICIPANTS WHO PURCHASED TEA BY SET SIZE AND PRESENCE OF THE OPTION TO DELEGATE TO A SALESPERSON, EXPERIMENT 6 We can examine this interaction by breaking down the results by set size. Although purchases did not differ based on whether a salesperson was present (61%) or absent (64%) when there were few options (χ2(1, N = 56) = .02, p = .89, ϕ = .02), more participants purchased tea when a salesperson was present (86%) than when a salesperson was not present when there were many options (37%; χ2(1, N = 76) = 18.87, p < .001, ϕ = .50). Or, if we instead break down the interaction by the presence of a salesperson, fewer participants purchased tea when there were many available tea varieties (37%) than when there were few (64%) when a salesperson was not present (χ2(1, N = 63) = 4.22, p = .04, ϕ = .26), but more participants purchased tea when there were many available tea varieties (86%) than when there were few (61%) when a salesperson was present (χ2(1, N = 69) = 5.13, p = .02, ϕ = .27). Discussion Ensuring that consumers have access to salespeople or other surrogates to whom they can delegate difficult decisions can reduce and even reverse the tendency to walk away from choices empty-handed. Participants were more likely to make a purchase from a large array of options when they could delegate the choice to another person than when they could not. And, whereas fewer participants made a purchase from a large array of options than a small array when delegation was not an option, more participants made a purchase from a large array when it was. GENERAL DISCUSSION This research shows that consumers cope with difficult decisions by recruiting others to choose for them. Across eight experiments, participants were more likely to ask others to choose on their behalf when choices felt difficult than when they felt easy. Delegation increased when choices felt difficult regardless of whether that feeling was because the choices themselves were more difficult (e.g., with a larger number of alternatives, with difficult tradeoffs, or with a smaller difference in relative attractiveness between the alternatives), or because the choices were processed less fluently for superficial reasons (e.g., the options were presented in jargon). Delegation increased when choices felt difficult regardless of whether the consequences were real or hypothetical and regardless of the importance of those consequences. Our findings suggest that the decision to delegate a difficult choice is rooted in the desire to avoid the potential regret associated with feeling responsible for a bad choice, rather than the desire to simply avoid the possibility of a bad outcome. Delegation increased when choices felt difficult regardless of whether the potential surrogate was an expert or nonexpert, consistent with the notion that people delegate to avoid responsibility for potentially making a suboptimal decision rather than to put decisions in more capable hands. Moreover, the degree to which participants anticipated regretting a poor choice, but not the degree to which they expected to be disappointed by a poor outcome, directly predicted the tendency to delegate. Consequently, delegation is often a more attractive option for avoiding a difficult choice than opting out of choosing altogether, and giving consumers the option to delegate can reduce the tendency for consumers to walk away from difficult decisions empty-handed. One might wonder why participants in our studies were willing to delegate difficult decisions when past research on the preference for choosing has shown that people overwhelmingly prefer to make decisions themselves, even when doing so is cognitively or emotionally costly (Botti et al. 2009). A potentially important distinction between how preference for choice is typically studied and our everyday examples of choice delegation is that, unlike in our studies, whether participants choose themselves or another person chooses for them is typically externally imposed in past research. The fact that participants assigned to have another person choose on their behalf were deprived of their freedom to choose could explain why they indicated that they would have preferred to choose themselves even when having another person choose on their behalf led them to feel more satisfied with the choice outcomes. Thus, although the preference for choice literature shows that people prefer choosing themselves when who gets to make the final choice is externally dictated, even when faced with difficult decisions between options with negative or tragic consequences (Botti et al. 2009), people may prefer to voluntarily cede decision control when considering difficult decisions in prospect. Directions for Future Research Our research is among the first to examine delegation in everyday contexts, leaving many open questions for future research about the extent and impact of delegation as a form of decision support. For one, to what extent might the present findings extend to advice seeking? Advice seeking can provide many of the same benefits as delegation for mitigating uncertainty and reducing the risk of making a bad choice: for example, it can provide additional information or validation of one’s inclinations. Thus, many factors that prompt delegation are also likely to prompt advice seeking. In fact, when avoiding being disappointed by a bad outcome is of primary concern, advice seeking may be more attractive than delegation in that it provides many of the same benefits while still allowing people to maintain control over the final decision. In our work, we focused on the antecedents of the decision to delegate, but decision makers are not the only ones with emotional and other stakes in the process. Future work might investigate how surrogates react to being asked to make a choice on someone else’s behalf. People are often flattered by being asked for advice (Brooks, Gino, and Schweitzer, 2015), for instance. But people can find making a choice on behalf of others to be stressful, and are willing to seek out second-order surrogates to make choices for others for them (Steffel, Williams, and Perrmann-Graham, 2016). This work shows that people are most likely to “pass the buck,” so to speak, when there is the potential for feeling responsible and being blamed for someone else’s negative outcome (Steffel et al. 2016). What are other factors that influence a potential surrogate’s willingness to make a choice for someone else? Surrogate expertise did not differentially affect our participants’ likelihood of delegating, but might it affect the surrogates’ willingness to choose? Further, there are many people who are expected to make decisions on behalf of others like our surrogates here do, like parents, spouses, doctors, and bosses. How do they react to this responsibility, and what makes them take on or abdicate this responsibility? Practical and Theoretical Implications Understanding the conditions under which consumers are more likely to delegate choices can help marketers to better target marketing campaigns to the people who will ultimately be making the purchasing decision. Depending on who the primary decision makers are, marketers may allocate resources differently across different elements of the marketing mix. For example, a pharmaceutical company might allocate fewer resources to advertising and more resources to personal selling in situations where consumers are more prone to delegate the choice of what drug to take to their doctors. Additionally, marketers may position their products differently when professional surrogates are the decision makers as opposed to consumers, because those professionals may have a different level of expertise or a different set of considerations than consumers. For example, a pharmaceutical company may market a drug to consumers by emphasizing lifestyle benefits but market the same drug to doctors by emphasizing performance in clinical trials. This research can also help retailers and service providers identify and make the most of situations in which consumers are likely to delegate choices to salespeople. Unsolicited product recommendations are discounted relative to solicited recommendations and may even evoke reactance (Fitzsimons and Lehmann 2004). Knowing the conditions under which consumers are more likely to delegate choices can help businesses improve the quality of their customer service and the effectiveness of their salespeople by helping them lend decision support when and where it would be most desired. For example, retailers may better position salespeople by concentrating them around products that are difficult to differentiate or for which there are many alternatives. It also can help retailers identify opportunities in which salespeople are likely to have more influence in the decision process and can be used more effectively to move merchandise. Despite the prevalence of choice delegation in today’s marketplace and the tremendous value in being able to predict when consumers are likely to delegate, the antecedents and implications of delegation are not well understood. This present research provides new insight into the preference for choice determination by identifying regret and responsibility as determinants of whether people prefer freedom of choice or freedom from choice, suggesting that the anticipated regret associated with feeling responsible for a bad choice can overwhelm people’s general preference for making their own choices. The present research contributes to an understanding of surrogate usage by identifying when consumers are most likely to employ others to make decisions on their behalf and by showing that expertise is a desirable but nonessential feature of a potential surrogate. Finally, this research contributes to the choice avoidance literature in particular, and the judgment and decision making literature more broadly, by showing that individual decision making is only a small part of a larger picture and that decisions, including the decision whether or not to decide in the first place, often hinge upon the social context and whether people have the option to delegate those choices to others. DATA COLLECTION INFORMATION The second author supervised the collection of data by research assistants at the University of California, San Diego, for experiment 1a in spring 2016 and for experiment 1c in spring 2014. The first author supervised the collection of data for experiment 1b by research assistants at the University of Cincinnati in spring 2015. The first and second authors jointly managed the collection of data using MTurk for experiment 3 in fall 2015, and the pilot experiment and experiments 2, 4, and 5 in fall 2016. The first and second authors jointly managed the collection of data for experiment 6 by research assistants at the University of Florida in fall 2011. The first and second authors jointly analyzed the data. The authors thank Paige Bausch, Mark Havens, Mercedes Castro, Peter Chami, Ana Figueras, Bernardo Guzman, Danielle King, Gia Nardini, Shane Strunk, and David Vitt for assistance with data collection; Lauren McCarthy and Megan Parin for assistance with information gathering; Julia Le, Matteo Prandi, Isabella Mozdzierz-Monico, and Apoorva Siddhi Sudhir for assistance with survey creation; and Chris Janiszewski, Nathan Novemsky, Aner Sela, and three anonymous reviewers for helpful comments APPENDIX MEDICAL TREATMENT CHOICE SCENARIO, EXPERIMENT 1C Plain language version: Imagine waking up to this scenario and being faced with an important life decision: Hello, my name is Dr. Bunnalai, and I am a doctor with the critical care team here at the hospital. You were in a car accident. One of the bones in your neck was damaged, and the spinal cord in your neck is severely bruised and bleeding. You have two options. Option 1: You can choose to undergo surgery to remove the blood that has collected around your spinal cord and to stop the bleeding. After the surgery, for the rest of your life, you may experience some constant neck pain and a tingling or burning sensation that will sometimes spread through your arms and legs. Option 2: Alternatively, you can choose NOT to undergo surgery. Instead, you will receive drugs to control your pain as needed. However, if the blood that has collected is not removed and the bleeding is not stopped, the blood will continue to press on your spinal cord and may permanently damage your spinal cord and cause paralysis. Medical jargon version: Imagine waking up to this scenario and being faced with an important life decision: Hello, my name is Dr. Bunnalai, and I am an Attending Intensivist with the trauma critical care medicine team here at the Level One Trauma Center. You were involved in a MVA, a motor vehicle accident. You sustained blunt force trauma injuries to your C4 vertebrae, part of the cervical spinal column in your neck, and you have a hematomyelia, an intramedullary spinal cord hemorrhage. You have two options. Option 1: You can choose to undergo surgical evacuation of the hematoma, the extravasated blood that has collected around your C4 vertebrae, and endovascular embolization to selectively occlude the affected blood vessels and arrest the bleeding. Post-operation, for the duration of your life, you may experience some chronic cervicalgia, a painful sensation that is localized to your neck, and dysesthesia, a painful tingling or burning sensation that will intermittently radiate though your extremities. Option 2: Alternatively, you can choose NOT to undergo surgery. Instead, you will receive palliative care consisting of PRN intravenous analgesics to control your pain. However, if the hematoma remains unextracted and blood vessels remain patent, the hematoma will continue to impinge on your spinal cord and may permanently damage your spinal cord and cause paralysis. REFERENCES Alter Adam L., Oppenheimer Daniel M. ( 2009), “Uniting the Tribes of Fluency to Form a Metacognitive Nation,” Personality and Social Psychology Review , 13 3, 219– 35. Google Scholar CrossRef Search ADS PubMed  Anderson Christopher J. ( 2003), “The Psychology of Doing Nothing: Forms of Decision Avoidance Result from Reason and Emotion,” Psychological Bulletin , 129 1, 129– 67. Google Scholar CrossRef Search ADS   Baron Jonathan, Ritov Ilana ( 1994), “Reference Points and Omission Bias,” Organizational Behavior and Human Decision Processes , 59 3, 475– 98. Google Scholar CrossRef Search ADS   Botti Simona, Iyengar Sheena S. ( 2004), “The Psychological Pleasure and Pain of Choosing: When People Prefer Choosing at the Cost of Subsequent Outcome Satisfaction,” Journal of Personality and Social Psychology , 87 3, 312– 26. Google Scholar CrossRef Search ADS PubMed  Botti Simona, McGill Ann L. ( 2006), “When Choosing Is Not Deciding: The Effect of Perceived Responsibility on Satisfaction,” Journal of Consumer Research , 33 2, 211– 9. Google Scholar CrossRef Search ADS   Botti Simona, McGill Ann L. ( 2011), “The Locus of Choice: Personal Causality and Satisfaction with Hedonic and Utilitarian Decisions,” Journal of Consumer Research , 37 6, 1065– 78. Google Scholar CrossRef Search ADS   Botti Simona, Orfali Kristina, Iyengar Sheena S. ( 2009), “Tragic Choices: Autonomy and Emotional Responses to Medical Decisions,” Journal of Consumer Research , 36 3, 337– 52. Google Scholar CrossRef Search ADS   Brehm Jack W. ( 1956), “Postdecision Changes in the Desirability of Alternatives,” Journal of Abnormal and Social Psychology , 52 3, 384– 9. Google Scholar CrossRef Search ADS   Brooks Alison Wood, Gino Francesca, Schweitzer Maurice E. ( 2015), “Smart People Ask for (My) Advice: Seeking Advice Boosts Perceptions of Competence,” Management Science , 61 6, 1421– 35. Google Scholar CrossRef Search ADS   Charles Cathy, Whelan Tim, Gafni Amiram, Reyno Leonard, Redko Cristina ( 1998), “Doing Nothing Is No Choice: Lay Constructions of Treatment Decision-Making among Women with Early-Stage Breast Cancer,” Sociology of Health and Illness , 20 1, 71– 95. Google Scholar CrossRef Search ADS   DeCharms Richard ( 1968), Personal Causation , New York: Academic Press. Dhar Ravi ( 1996), “The Effect of Decision Strategy on Deciding to Defer Choice,” Journal of Behavioral Decision Making , 9 4, 265– 81. Google Scholar CrossRef Search ADS   Dhar Ravi ( 1997a), “Consumer Preference for a No-Choice Option,” Journal of Consumer Research , 24 2, 215– 31. Google Scholar CrossRef Search ADS   Dhar Ravi ( 1997b), “Context and Task Effects on Choice Deferral,” Marketing Letters , 8 1, 119– 30. Google Scholar CrossRef Search ADS   Dhar Ravi, Nowlis Stephen M. ( 1999), “The Effect of Time Pressure on Consumer Choice Deferral,” Journal of Consumer Research , 25 4, 369– 84. Google Scholar CrossRef Search ADS   Elwyn Glyn, O’Connor Annette, Stacey Dawn, Volk Robert, Edwards Adrian, Coulter Angela, Thomson Richard, Barratt Alexandra, Barry Michael, Bernstein Steven, Butow Phyllis, Clarke Aileen, Entwistle Vikki, Feldman-Stewart Deb, Holmes-Rovner Margaret, Llewellyn-Thomas Hilary, Moumjid Nora, Mulley Al, Ruland Cornelia, Sepucha Karen, Sykes Alan, Whelan Tim ( 2006), “International Patient Decision Aids Standards (IPDAS) Collaboration: Developing A Quality Criteria Framework For Patient Decision Aids: Online International Delphi Consensus Process,” British Medical Journal , 333 7565: 417. Google Scholar CrossRef Search ADS PubMed  Festinger Leon ( 1964), Conflict, Decision, and Dissonance , Stanford, CA: Stanford University Press. Fitzsimons Gavan J., Lehmann Donald R. ( 2004), “Reactance to Recommendations: When Unsolicited Advice Yields Contrary Responses,” Marketing Science , 23 1, 82– 94. Google Scholar CrossRef Search ADS   Forsythe Sandra, Butler Sara, Schaefer Robert ( 1990), “Surrogate Usage in the Acquisition of Women’s Business Apparel,” Journal of Retailing , 66 4, 446– 69. Fuller Barbara K., Blackwell Suzannah C. ( 1992), “Wardrobe Consultant Clientele: Identifying and Describing Three Market Segments,” Clothing and Textiles Research Journal , 10 2, 11– 7. Google Scholar CrossRef Search ADS   Iyengar Sheena S., Lepper Mark R. ( 2000), “When Choice Is Demotivating: Can One Desire Too Much of a Good Thing?” Journal of Personality and Social Psychology , 79 6, 995– 1006. Google Scholar CrossRef Search ADS PubMed  Janis Irving L., Mann Leon ( 1976), “Coping with Decision Conflict: An Analysis of How Stress Affects Decision-Making Suggests Interventions to Improve the Process,” American Scientist , 64 6, 657– 67. Kenny Patricia, Quine Susan, Shiell Alan, Cameron Sue ( 1999), “Participation in Treatment Decision-Making by Women with Early Stage Breast Cancer,” Health Expectations , 2 3, 159– 68. Google Scholar CrossRef Search ADS PubMed  Langer Ellen J., Rodin Judith ( 1976), “The Effects of Choice and Enhanced Personal Responsibility for the Aged: A Field Experiment in an Institutional Setting,” Journal of Personality and Social Psychology , 34 2, 191– 8. Google Scholar CrossRef Search ADS PubMed  Luce Mary Frances ( 1998), “Choosing to Avoid: Coping with Negatively Emotion-Laden Consumer Decisions,” Journal of Consumer Research , 24 4, 409– 33. Google Scholar CrossRef Search ADS   Marcatto Francesco, Ferrante Donatella ( 2008), “The Regret and Disappointment Scale: An Instrument for Assessing Regret and Disappointment in Decision Making,” Judgment and Decision Making , 3 1, 87– 99. McLaughlin G. Harry ( 1969), “SMOG Grading – A New Readability Formula,” Journal of Reading , 12 8, 639– 46. Novemsky Nathan, Dhar Ravi, Schwarz Norbert, Simonson Itamar ( 2007), “Preference Fluency in Choice,” Journal of Marketing Research , 44 3, 347– 56. Google Scholar CrossRef Search ADS   Ordóñez Lisa D., Connolly Terry ( 2000), “Regret and Responsibility: A Reply to Zeelenberg et al.  ,” Organizational Behavior and Human Decision Processes , 81 1, 132– 42. Google Scholar CrossRef Search ADS PubMed  Park C. Whan, Jun Sung Youl, MacInnis Deborah J. ( 2000), “Choosing What I Want Versus Rejecting What I Don’t Want: An Application of Decision Framing to Product Option Choice Decisions,” Journal of Marketing Research , 37 2, 187– 202. Google Scholar CrossRef Search ADS   Pratt Gerry ( 1981), “The House as an Expression of Social Worlds,” in Housing and Identity: Cross-Cultural Perspectives , ed. Duncan James S., London: Croon Helm, 135– 79. Preacher Kris J., Hayes Andrew F. ( 2008) “Asymptotic and Resampling Strategies for Assessing and Comparing Indirect Effects in Multiple Mediator Models,” Behavior Research Methods , 40 3, 879– 91. Google Scholar CrossRef Search ADS PubMed  Redelmeier Donald A., Shafir Eldar ( 1995), “Medical Decision Making in Situations that Offer Multiple Alternatives,” Journal of the American Medical Association , 273 4, 302– 5. Google Scholar CrossRef Search ADS PubMed  Rudd Rima E., Moeykens Barbara A., Colton Tayla C. ( 1999), “Health and Literacy: A Review of Medical and Public Health Literature,” in Annual Review of Adult Learning and Literacy.  ed. Comings John, Garners Barbara, Smith Christine, New York: Jossey-Bass, 175– 203. Ryan Richard M., Deci Edward L. ( 2000), “Self-Determination Theory and the Facilitation of Intrinsic Motivation, Social Development, and Well-Being,” American Psychologist , 55 1, 68– 78. Google Scholar CrossRef Search ADS PubMed  Schrift Rom, Netzer Oded, Kivetz Ran ( 2011), “Complicating Choice,” Journal of Marketing Research , 48 2, 308– 26. Google Scholar CrossRef Search ADS   Seligman Martin E. ( 1975), Helplessness: On Depression, Development, and Death , San Francisco: Freeman. Simonson Itamar ( 1992), “The Influence of Anticipating Regret and Responsibility on Purchase Decisions,” Journal of Consumer Research , 19 1, 105– 18. Google Scholar CrossRef Search ADS   Solomon Michael R. ( 1987), “The Wardrobe Consultant: Exploring the Role of a New Retailing Partner,” Journal of Retailing , 63 2, 110– 28. Sugden Robert ( 1985), “Regret, Recrimination, and Rationality,” Theory and Decision , 19 1, 77– 99. Google Scholar CrossRef Search ADS   Steele Claude M. ( 1988), “The Psychology of Self-Affirmation: Sustaining the Integrity of the Self,” in Advances in Experimental Social Psychology , Vol. 21, ed. Berkowitz Leonard, New York: Academic Press, 261– 2. Steffel Mary, Williams Elanor F. ( 2015), “Empowering Patients to Be Consumers of Their Own Medical Care: Increasing the Subjective Ease of Medical Decisions Increases Patient Participation,” working paper, Northeastern University, Boston, MA 02115. Steffel Mary, Williams Elanor F., Perrmann-Graham Jaclyn ( 2016), “Passing the Buck: Delegating Choices to Others to Avoid Blame,” Organizational Behavior and Human Decision Processes , 135 ( July), 32– 44. Google Scholar CrossRef Search ADS   Stern Barbara B., Solomon Michael R., Stinerock Robert ( 1992), “Surrogate Money Managers in the 1990s: Marketing Strategy for Financial Services Retailers,” Service Industries Journal , 12 1, 78– 96. Google Scholar CrossRef Search ADS   Stinerock Robert, Stern Barbara B., Solomon Michael R. ( 1991), “Sex and Money: Gender Differences in the Use of Surrogate Consumers for Financial Decision Making,” Journal of Professional Services Marketing , 7 2, 167– 82. Google Scholar CrossRef Search ADS   Taylor Shelley E., Brown Jonathon D. ( 1988), “Illusion and Well-Being: A Social Psychological Perspective on Mental Health,” Psychological Bulletin , 103 2, 193– 210. Google Scholar CrossRef Search ADS PubMed  Tetlock Philip E., Boettger Richard ( 1994), “Accountability Amplifies the Status Quo Effect When Change Creates Victims,” Journal of Behavioral Decision Making , 7 1, 1– 23. Google Scholar CrossRef Search ADS   Tversky Amos, Shafir Eldar ( 1992), “Choice under Conflict: The Dynamics of Deferred Decision,” Psychological Science , 3 6, 358– 61. Google Scholar CrossRef Search ADS   Tykocinski Orit E., Pittman Thane S. ( 1998), “The Consequences of Doing Nothing: Inaction Inertia as Avoidance of Anticipated Counterfactual Regret,” Journal of Personality and Social Psychology , 75 3, 607– 16. Google Scholar CrossRef Search ADS   Usta Murat, Häubl Gerald ( 2011), “Self-Regulatory Strength and Consumers’ Relinquishment of Decision Control: When Less Effortful Decisions Are More Resource Depleting,” Journal of Consumer Research , 48 2 403– 12. van Dijk Wilco W., Zeelenberg Marcel, van der Pligt Joop ( 2003), “Blessed Are Those Who Expect Nothing: Lowering Expectations as a Way of Avoiding Disappointment,” Journal of Economic Psychology , 24 4, 505– 16. Google Scholar CrossRef Search ADS   Williams Elanor F., Gneezy Ayelet, Armor David A. ( 2015), “Between a Rock and Another Rock: Ironic Persistence at Choices with Interchangeable Options,” working paper, Kelley School of Business, Indiana University, Bloomington, IN 47405. Zeelenberg Marcel ( 1999), “Anticipated Regret, Expected Feedback, and Behavioral Decision Making,” Journal of Behavioral Decision Making , 12 7, 93– 106. Google Scholar CrossRef Search ADS   Zeelenberg Marcel, van Dijk Wilco W., Manstead Antony S. R. ( 1998), “Reconsidering the Relation between Regret and Responsibility,” Organizational Behavior and Human Decision Processes , 74 3, 254– 72. Google Scholar CrossRef Search ADS PubMed  Zeelenberg Marcel, van Dijk Wilco W., Manstead Antony S. R., van der Pligt Joop ( 2000), “On Bad Decisions and Disconfirmed Expectancies: The Psychology of Regret and Disappointment,” Cognition and Emotion , 14 4, 521– 41. Google Scholar CrossRef Search ADS   © The Author 2017. Published by Oxford University Press on behalf of Journal of Consumer Research, Inc. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Consumer Research Oxford University Press

Delegating Decisions: Recruiting Others to Make Choices We Might Regret

Loading next page...
 
/lp/ou_press/delegating-decisions-recruiting-others-to-make-choices-we-might-regret-2qZx248q3T
Publisher
University of Chicago Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of Journal of Consumer Research, Inc. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0093-5301
eISSN
1537-5277
D.O.I.
10.1093/jcr/ucx080
Publisher site
See Article on Publisher Site

Abstract

Consumers typically prefer freedom of choice, but when faced with a choice they might regret, they may prefer freedom from choice. Eight experiments show that people delegate difficult decisions, regardless of the decision’s importance, and regardless of their potential surrogate’s expertise. Delegation stems from a desire to avoid responsibility for potentially making the wrong choice rather than simply the desire to avoid the possibility of a poor outcome: although anticipated disappointment with the outcome and anticipated regret about one’s decision both contribute to the decision to delegate, only anticipated regret directly leads people to delegate choices to others. Consequently, delegation is an appealing method for coping with difficult choices while allowing consumers to retain the benefits of choosing that they would forgo by opting out of the choice. Moreover, giving people the option to delegate makes them less prone to walk away from difficult choices empty-handed. choice delegation, choice deferral, responsibility, regret Having the freedom to make choices for oneself is considered by many to be an inalienable right. Some have suggested that having control of one’s own choices is a basic need (Ryan and Deci 2000). When choice autonomy is threatened, people assert their freedom; for example, when they encounter unsolicited advice, they often discount or reject that advice (Fitzsimons and Lehmann 2004). Perceiving a lack of choice autonomy can even threaten people’s happiness, self-esteem, and physical well-being (Langer and Rodin 1976; Seligman 1975; Steele 1988; Taylor and Brown 1988; Usta and Häubl 2011). It is perhaps no surprise then that, even when choosing exacts a toll, people often prefer to choose themselves rather than have another person choose on their behalf (Botti and Iyengar 2004; Botti and McGill 2006, 2011; Botti, Orfali, and Iyengar 2009). Yet it also seems likely that people might sometimes value freedom from choice, especially when they worry that they might choose poorly and thus regret their decision. Decision makers facing difficult choices often postpone them, take the “easy” way out by choosing defaults or status quo options, or even opt out of making a decision altogether (Anderson 2003). But there are times when choosing nothing is not an option, or when walking away empty-handed is unappealing. Thus far, research on choice avoidance has largely focused on the choice and the individual making it, rather than the broader social context in which that choice is situated. Consequently, it has given less consideration to the idea that other people can serve as a resource to a decision maker, and that delegating choices to others may be an often-utilized strategy for avoiding the burden of responsibility associated with making a difficult decision and an appealing alternative to walking away from a difficult choice without anything to show for it. Although little is known about when and why people might delegate choices to others, some qualitative research has explored who delegates shopping decisions, to whom, and with what benefits. This research has typically focused on formal transactions in which consumers hire professional agents to make shopping decisions on their behalf to get functional benefits (e.g., expertise) and symbolic benefits (e.g., status; Forsythe, Butler, and Schaefer 1990; Fuller and Blackwell 1992; Pratt 1981; Solomon 1987; Stern, Solomon, and Stinerock 1992; Stinerock, Stern, and Solomon 1991). The present research broadens the traditional conceptualization of delegation to reflect everyday choices and choice contexts, shows that surrogate usage is not predicated on professional expertise, and demonstrates that delegation is often motivated by factors aside from functional and symbolic benefits. Consumers readily recruit others in their immediate social context—salespeople, waiters, friends, even strangers—to make difficult decisions on their behalf so that they do not have to bear responsibility for choosing and feel at fault if the choice is suboptimal. This research examines when consumers delegate, why they do so, and what this means for purchases and sales when delegation is or is not an option. DETERMINANTS OF DELEGATION If the desire to make one’s own choices is often so strong, when might people prefer to delegate? Delegation can provide a number of benefits that may make it an appealing alternative to choosing oneself. For one, delegation enables people to avoid the effort of choosing. Perhaps people delegate choices when the consequences feel unimportant and putting a lot of time and energy into making a decision does not seem worthwhile. Another benefit of delegation is that it allows people to put choices in the hands of people who might have special expertise. Perhaps people delegate when they believe that the potential surrogate has knowledge or experience that would make him or her more capable of making a good choice. But even more fundamentally, delegation allows people to transfer some of the responsibility for the decision to someone else. As we test here, perhaps people delegate when they want to avoid feeling or being at fault for potentially making a less-than-ideal decision. This benefit may be particularly appealing when choices feel difficult and the chance of choosing poorly—and thus regretting one’s choice—seems high. If concerns about responsibility and regret are determinants of the decision to delegate, examining whether people delegate more when faced with a difficult choice is a good place to start, because choice difficulty is known to be tied to regret (Sugden 1985; Zeelenberg 1999). Choice difficulty undermines people’s confidence that they will select the best option (Dhar 1996, 1997a, 1997b; Dhar and Nowlis 1999; Steffel and Williams 2015) and prompts people to try to mitigate the bad feelings they might experience if they choose suboptimally (Zeelenberg 1999). In their conflict-theory model of decision making, Janis and Mann (1976) hypothesized that people who are facing a challenging decision are more likely to practice “defensive avoidance”—procrastinating, ignoring, or passing responsibility for a decision to someone else. Indeed, choice difficulty has been shown to lead to a variety of forms of choice avoidance (Anderson 2003). People avoid difficult choices by postponing them to a later point in time, either by gathering more information, seeking additional alternatives, or simply mulling it over (Dhar 1996, 1997b; Dhar and Nowlis 1999; Luce 1998; Novemsky et al. 2007; Tversky and Shafir 1992). People also avoid difficult decisions by retaining a status quo or default option (Luce 1998; Redelmeier and Shafir 1995; Tversky and Shafir 1992). Or, at times, they opt not to choose anything at all (Dhar 1997a; Iyengar and Lepper 2000). Although there has yet to be a direct test of whether difficulty drives delegation, there is correlational evidence to suggest that people might also be more likely to delegate when choices are challenging. Tetlock and Boettger (1994) found a correlation between participants’ ratings of choice difficulty and the likelihood that they would invite another party to review and potentially revise a decision they have made. Among consumers utilizing wardrobe consultants, those reporting the greatest difficulty in making buying decisions were most likely to let a wardrobe consultant make the final decision of what to purchase (Fuller and Blackwell 1992). Further, patients who report being afraid of making the “wrong” decision were more likely to prefer that their doctors make treatment decisions on their behalf (Charles et al. 1998; Kenny et al. 1999). Thus, we hypothesize that increasing the feeling of choice difficulty will lead people to delegate. H1: As the subjective difficulty of making a choice increases, people are more likely to delegate the decision to another person. AVOIDING RESPONSIBILITY AND REGRET We suspect that difficulty increases the extent to which people worry about being responsible for making a suboptimal decision, and encourages them to reduce or even evade that responsibility. Disappointment and regret both capture a person’s reaction to having made a bad choice. What differentiates them is that people feel disappointment when a choice does not turn out as well as they had hoped, regardless of the cause of the failure or their hand in the choice, but they feel regret if they bear responsibility for a less-than-optimal outcome (Zeelenberg et al. 2000). That people feel responsible for the outcomes of choices they actively make is well known, having been studied in contexts like personal control (DeCharms 1968), cognitive dissonance (Festinger 1964), and regret itself (Ordóñez and Connolly 2000). Indeed, the feeling of responsibility, rather than the act of choosing, is likely to determine one’s emotional reaction to a decision, including satisfaction and regret (Botti and McGill 2006). Further, one can feel these unpleasant emotions in prospect of a decision, not just in retrospect. Anticipated disappointment is felt when someone is worried that the choice outcome will not be as good as their expectations or hopes. Anticipated regret is felt when someone is worried that the option they choose will not be as good as the other options they choose to forgo, and it will be their fault for having chosen it. Not only do people feel responsible after having made a choice, they also worry about feeling responsible in anticipation of making it. Anticipated regret has been shown to contribute to other forms of choice avoidance: namely, it makes people more inclined to choose default or status quo options (Baron and Ritov 1994; Park, Jun, and MacInnis 2000; Simonson 1992) and increases inaction inertia (Tykocinski and Pittman 1998). Given what we know about the interplay between responsibility, regret, and choice avoidance, we suspect that when a decision is difficult and it is unclear whether one will make the best choice, decision makers will worry about feeling responsible for the outcome of the choice and thus regretting their role in it, which will make delegating that choice to someone else an appealing option. Note that it is possible that anticipated disappointment may also prompt choice avoidance, as it may lead to feelings of powerlessness and a need to escape or do nothing (Zeelenberg, van Dijk, and Manstead 1998), and lower expectations about how well the decision might turn out (van Dijk, Zeelenberg, and van der Pligt 2003). However, although both of these anticipated emotions could encourage delegation, anticipated regret is more likely to directly prompt delegation, as it may provide a means of escaping regret by reducing one’s own responsibility for a choice and passing it to a surrogate. Anticipated disappointment is less likely to directly prompt delegation, as resolving a decision via delegation does not eliminate (and could conceivably increase) the possibility that the choice might turn out badly. Note, too, that it is also possible that both may play a role, or that disappointment may even precede regret in prompting delegation; after all, one has to have a choice go badly in order to regret having made it. H2a: The more difficult a choice is to make, the more likely people are to anticipate that they might feel disappointment and regret if they were responsible for choosing a suboptimal option. H2b: Although anticipated disappointment and regret may both contribute to delegation, only anticipated regret directly leads people to delegate choices to others. One important assumption here is that the expected benefit of delegation is the reduction of responsibility. Do people expect to feel less responsible when they delegate? One could imagine that responsibility might remain with the original decision makers, or might even increase, once they have delegated a choice, because they are not only responsible for the outcome of the choice but also for asking someone else to decide for them. Alternatively, as we predict, delegation may enable people to avoid feeling responsible for choice outcomes by shifting who ultimately determines the outcome and allowing them to attribute that responsibility to another person. To understand the drivers of delegation, it is important to know whether people do expect to retain responsibility for choosing even when they delegate. PILOT STUDY This pilot study explored whether people believe that they can reduce their responsibility for a choice outcome by delegating that choice. We recruited 197 participants via Mechanical Turk, setting the target sample size such that we would have about 200 participants total. One participant did not complete all of the dependent measures, yielding a sample of 196. We presented participants with a difficult choice between two flavors of ice cream that were similarly appealing to them. We determined these flavors by asking participants to rank a set of 14 ice cream flavors, and then we told them that only their 4th and 5th ranked flavors were available. Participants next imagined having to decide between the two available flavors, when Jenny, the server, offered to help them choose. With this in mind, they were asked to: “Imagine that you had chosen an ice cream flavor yourself without any help from Jenny. Unfortunately you did not like the ice cream you chose.” They then rated, “How much would you feel that YOU are responsible for the fact that you didn’t enjoy the flavor?” and “How much would you feel that JENNY is responsible for the fact that you didn’t enjoy the flavor?” on scales from 1 = not at all to 7 = very much. Then, participants imagined that, instead, they had asked Jenny to give them the ice cream flavor she thought they would like best, that they did not like the ice cream she chose, and to ascribe responsibility for that outcome to themselves and Jenny. The order in which participants imagined choosing themselves or delegating and the order in which participants ascribed responsibility to themselves and Jenny were counterbalanced. This pilot showed that delegating does allow people to feel less personally responsible for choice outcomes by enabling them to attribute responsibility to another person. We found a significant interaction between choice method and target (F(1, 195) = 120.22, p < .001, ηp2 = .38), such that participants felt less personally responsible for the fact that they did not enjoy the flavor of ice cream they received when they delegated the choice to Jenny than when they chose themselves and attributed more responsibility for the choice outcome to Jenny when they delegated the choice to her than when they chose themselves. Additionally, participants ascribed more responsibility overall to themselves than to Jenny (F(1, 195) = 281.79, p < .001, ηp2 = .59). The effect of choice method (choosing oneself or delegating) on attributions of responsibility was not significant (F(1, 195) = 2.11, p = .15, ηp2 = .01). See table 1. Table 1 Attributions of Responsibility for Choice Outcomes When the Choice Was Made Oneself or Delegated to Another Person   Made choice oneself Mean (SD)  Delegated choice to Jenny Mean (SD)    How responsible I am  5.91 (1.66)  4.52 (1.88)  t(195) = 9.86, p < .001, d = 1.76  How responsible Jenny is  1.69 (1.47)  3.27 (1.97)  t(195) = 10.15, p < .001, d = 2.05    Made choice oneself Mean (SD)  Delegated choice to Jenny Mean (SD)    How responsible I am  5.91 (1.66)  4.52 (1.88)  t(195) = 9.86, p < .001, d = 1.76  How responsible Jenny is  1.69 (1.47)  3.27 (1.97)  t(195) = 10.15, p < .001, d = 2.05  Decision makers do seem to perceive delegation as an effective way to reduce the responsibility they feel for a choice by transferring some of that responsibility to another person. Participants who imagined delegating a decision to a salesperson anticipated feeling less responsible for the outcome of that decision and attributed more responsibility to the salesperson than those who imagined making the choice themselves. Given that delegating allows people to expect to feel less personally responsible for choice outcomes by enabling them to attribute some of that responsibility to another person, we can now turn to testing whether consumers do opt to delegate or whether they prefer other ways out of difficult decisions. DELEGATION VERSUS OPTING OUT Delegation is not the only means by which people can avoid making difficult decisions. Another common way in which people escape tough choices is by refusing to choose at all. Whereas delegation and opting out of choosing may both be appealing when either one (but not the other) is available as an alternative to making a difficult decision oneself, delegation may be a uniquely attractive alternative when choices are difficult but choosing is necessary or desirable and when both delegation and opting out are available alternatives. One potentially important way in which delegation differs from opting out is that it enables people to avoid feeling at fault if the choice outcome is suboptimal while still allowing them to get something out of the choice, whereas opting out necessitates that a decision maker walk away empty-handed. If people’s primary motive is to avoid responsibility and regret for a less-than-ideal outcome of a decision while still leaving open the possibility of a good outcome, then delegation may be uniquely appealing. If, on the other hand, people’s motive is to guarantee that they avoid an unappealing outcome entirely, then opting out may be preferred to delegation. H3a: Delegation (vs. opting out of choosing or choosing oneself) will be more attractive as a means of deciding as choices become more difficult. H3b: Opting out (vs. delegation or choosing oneself) will be more attractive as a means of deciding as the choice options become less appealing. In most contexts, choosing or walking away are not typically people’s only options. Rather, consumers often have the option of asking other people—friends, salespeople or other service professionals, even strangers—to help them choose. Providing consumers with surrogate decision makers to whom they can delegate may make them less likely to opt out of making a purchase in situations in which they might otherwise walk away empty-handed. For example, a consumer faced with a difficult decision between equally attractive product offerings may be less likely to leave without making a purchase if he or she can resolve his or her decision conflict by seeking decision support from a salesperson or shopping companion. Many choices are hard to make because the options are equally appealing (Williams, Gneezy, and Armor 2015), and choosing one will mean forgoing the appealing aspects of the other. Ironically, this should also mean that people should be happy with their chosen option no matter which one they choose, and yet this difficulty can lead them to put off choosing or to walk away entirely. For example, people are less likely to make a purchase from a large than a small array of options, despite the fact that more options should only increase the likelihood that one of those options should prove appealing (Iyengar and Lepper 2000), and they are more likely to opt out of a choice with options that are closer rather than more distant in appeal (Dhar 1997b), when they should be more indifferent between the options in the former case. People might be more likely to make a purchase in such cases if they have the option of ceding responsibility for choosing to another person. Thus, providing opportunities for consumers to delegate may allow consumers to get the benefits of choosing without the burdens and enable businesses to increase purchases by making it less likely that consumers will opt out of decisions and not make a purchase. H4: Consumers are less likely to walk away from a difficult decision empty-handed when they have the option to delegate than when they do not. THE PRESENT RESEARCH The present research demonstrates that consumers are prone to delegate decisions they find to be difficult, for which the chance of regretting the outcome seems higher and thus off-loading responsibility is more appealing. Delegation, like other methods of choice avoidance, allows people to avoid choosing between options and thus possibly regretting the option they chose, but unlike other methods of choice avoidance, provides the added benefit of also resolving the decision itself. Delegation can thus also benefit businesses, which may find that making a surrogate available can reduce the number of customers who find themselves unable to decide and who might otherwise walk away from a transaction empty-handed. We test whether people delegate decisions they might regret by manipulating how difficult and thus potentially regret-provoking choices are. Choice difficulty can sometimes arise directly from the objective features of the choice options. For example, choices between comparably attractive options produce more choice difficulty than choices for which one alternative is much more attractive than the other (Brehm 1956; Festinger 1964; Tversky and Shafir 1992). Likewise, choices that involve making tradeoffs between valued attributes tend to feel more difficult than those that do not (Luce 1998). Additionally, choices for which there are many available options tend to feel more difficult than choices for which there are few options (Iyengar and Lepper 2000). Choice difficulty can also arise independently of the content of the choice options, such as from the fluency by which the choice options are presented (Alter and Oppenheimer 2009), like a difficult-to-read versus easy-to-read font (Novemsky et al. 2007). To show that the feeling of difficulty drives choice delegation as opposed to other factors that might be associated with any particular instantiation of difficulty, we manipulate difficulty in a variety of ways—namely, via set size, relative attractiveness, tradeoffs, and linguistic fluency. In eight experiments, we examine the antecedents of delegation and the consequences of having the option to delegate. Experiments 1a, 1b, and 1c serve as initial demonstrations that consumers are more likely to delegate choices that feel difficult than choices that feel easy. Experiment 2 shows that this is the case for both highly important and less important decisions. Experiment 3 shows that people are more likely to delegate to experts than nonexperts, but the tendency to delegate difficult decisions is not predicated on expertise: consumers are more likely to delegate choices that feel difficult both when the potential surrogate is an expert and when he or she is not. Experiment 4 suggests that the desire to avoid responsibility for a poor outcome, rather than to simply avoid the poor outcome itself, underlies delegation. It also shows that anticipated regret, but not anticipated disappointment, predicts delegation of difficult choices. Finally, experiments 5 and 6 compare delegation to other forms of choice avoidance like opting out of choosing. Experiment 5 tests whether people prefer delegation over opting out as a means of handling a difficult decision, and whether the appeal of the choice set affects the tendency to delegate or opt out. Experiment 6 examines whether adding the option to delegate reduces opting out in situations in which people are otherwise prone to leaving empty-handed. EXPERIMENTS 1A, 1B, AND 1C: DOES DIFFICULTY PROMPT DELEGATION? Our initial studies examine whether, despite a general preference for making their own choices, people do at times prefer to have others choose on their behalf, and do so more often when choices feel difficult. Experiment 1a tests this prediction in an experiment with real consequences. Students chose one of two sets of earphones that were pretested to be close or distant in appeal to use in an experiment and keep as a gift. Participants could pick which set of earphones they wanted or have the experimenter choose a set for them. In experiment 1b, also in an experiment with real consequences, participants chose a jelly bean to sample in a taste test from a small or large array of options, another manipulation of choice difficulty. Participants could pick which flavor they wanted to taste or have the experimenter choose for them. In experiment 1c, participants considered the important decision of whether to undergo surgery to treat a life-threatening injury in a hypothetical medical scenario. Additionally, to examine whether the feeling of choice difficulty is a sufficient catalyst for choice delegation, we held constant the content of the target choice and manipulated difficulty superficially via the fluency of the language with which the options were described (Alter and Oppenheimer 2009). All three experiments show that participants are more likely to delegate when choices are difficult than easy, even when those choices have enduring, immediate, or important consequences. Experiment 1a Method Participants Undergraduates (N = 366) at the University of California, San Diego, participated in exchange for credit in marketing and other business classes. The sample size was set a priori to encompass one week’s worth of participants in the lab. In this and all subsequent studies, we report all measures, conditions, and data exclusions. Procedure Under the guise of an experiment about rating advertisements, participants learned that they would be receiving a free set of headphones they could use to listen to an ad. When they entered the lab, each participant had an order form at their desk with the two earphone options available to them. In the easy condition, one was a silver pair of earbuds and the other was a black pair of over-the-head headphones. In the difficult condition, one was a pair of silver earbuds and one was a pair of grey earbuds. Pretest participants at the same university (N = 320) rated the choice between the headphones and the earbuds to be easier than the choice between the two pairs of earbuds on a scale ranging from 1 = very easy to 7 = very difficult (M = 1.94, SD = 1.28 vs. M = 2.61, SD = 1.59, paired t(319) = 7.69, p < .001, d = 1.04). The conditions alternated between sessions so that everyone in one session would have the same two options. At the beginning of each session, the experimenter informed participants that: “In today’s session, you will be asked to listen to an advertisement and to evaluate its effectiveness. You will listen to the advertisement using a set of earphones, and as an incentive, you will get to keep the earphones that you use. You can either choose whichever earphones you prefer or you can opt for me to give you the pair of headphones that I think are best. Before you start today’s surveys, please make your selections using the earphone order form provided at your station.” Participants made their choices on the order form, indicating either that 1) they wanted to choose a set of earphones themselves and write down which set they wanted, or 2) they wanted to ask the experimenter to give them the earphones that the experimenter thought were best. The experimenter then went around the room, either distributing the earphones that participants chose or providing earphones when the participants delegated (the silver earbuds, in both conditions). Experiment 1a Results Participants were more likely to delegate the choice of earphones to the experimenter when presented with a difficult choice between a silver pair of earbuds and a grey pair of earbuds (28%) than when presented with an easy choice between a silver pair of earbuds and a black pair of over-the-head headphones (11%; χ2(1, N = 366) = 16.72, p < .001, ϕ = .21). Experiment 1b Method Participants Volunteers (N = 200) were recruited on campus at the University of Cincinnati to participate in exchange for the opportunity to taste a gourmet jelly bean. The target sample size was set a priori such that we would have about 100 participants per condition. Procedure Two research assistants who were blind to the hypothesis invited participants to engage in a taste test of gourmet jelly beans. Participants received a menu that contained five flavors (small set condition) or 25 flavors (large set condition). There were four versions of each menu: in the large set condition, the order of the flavors varied, and in the small set condition, the subset of flavors that was offered varied. Pretest participants at the same university (N = 141) were randomly assigned to see either the short or the long menu and rated how easy or difficult it would be to decide which flavor to taste, on a scale ranging from 1 = very easy to 10 = very difficult. Participants indicated it would be easier to choose a flavor from the small set (M = 3.52, SD = 2.78) than the large set (M = 5.64, SD = 3.14; t(139) = 4.23, p < .001, d = .71). Participants were asked, “Would you like to choose which flavor you will taste? Or would you like for me to choose a flavor for you?” Once they had tasted their jelly bean, all participants rated how much they liked the flavor they tasted on a scale ranging from 1 = not at all to 10 = very much. Experiment 1b Results Participants were marginally more likely to delegate the choice of which jelly bean flavor to taste to the experimenter when presented with a difficult choice between many options (28%) than when presented with a more manageable choice between fewer options (17%; χ2(1, N = 200) = 3.47, p = .06, ϕ = .13). This difference did not seem to be due to any variation in how appealing the flavor options in each set were: participants liked the jelly beans they tasted equally well in the large set condition (M = 7.57, SD = 1.69) and small set condition (M = 7.52, SD = 1.62; t(198) = –.21, p = .83, d = .03). Experiment 1c Method Participants Undergraduates (N = 296) at the University of California, San Diego, participated in exchange for course credit. The sample size was set a priori to reflect one week’s worth of participants in the lab. Procedure Participants imagined that they had sustained a neck injury in a car accident and had to decide whether to undergo surgery (see the appendix for full stimulus materials). We manipulated the subjective ease or difficulty of the decision by presenting the treatment options in either plain language (fluent condition) or medical jargon (disfluent condition). A SMOG analysis (McLaughlin 1969) indicated that the plain language version of the scenario was written at an eighth-grade reading level, the maximum reading level recommended by the International Patient Decision Aid Standards (Elwyn et al. 2006), and the medical jargon version was written at a 13th-grade reading level, the typical reading level for health-related materials (Rudd, Moeykens, and Colton 1999). Pretest participants from the same university (N = 47) rated the scenario as easier to read when it was in plain language (M = 2.09, SD = 1.41) than when it was in medical jargon (M = 5.23, SD = 1.55; paired t(46) = 10.25, p < 001, d = 2.12). Additionally, most participants in another pretest at the same university (N = 127) answered three comprehension questions correctly in both the medical jargon condition (86%) and plain language condition (91%; χ2(1, N = 127) = .63, p = .43), suggesting that participants were equally capable of understanding the scenario regardless of the language that was used. Participants indicated whether they would want to 1) choose a treatment themselves and identify which treatment they preferred, or 2) ask their doctor to choose a treatment for them. On the next page, participants responded to the same three multiple-choice questions from the pretest to determine how closely they were attending to the scenario. Experiment 1c Results Participants were more likely to delegate their choice of treatment to their doctor when we made the choice feel subjectively difficult by describing the same options in medical jargon (38%) versus plain language (20%; χ2(1, N = 296) = 12.03, p = .001, ϕ = .20). Interestingly, participants in the disfluent condition did not recall the details of the scenario as well as did those in the fluent condition (paired χ2(1, N = 296) = 30.35, p = .001, ϕ = .32): 72% of participants in the fluent condition answered all three comprehension questions correctly, but only 41% of those in the disfluent condition did so. We suspect that the difference in performance on the comprehension check between the pretest and experiment may be attributed to the fact that the comprehension check appeared on the same page as the scenario in the pretest and on a separate page from the scenario and choice in the experiment. We interpret the results of the comprehension check in the pretest to mean that participants were equally capable of comprehending the scenario regardless of whether it was in plain language or medical jargon and the results of the comprehension check in the experiment to mean that participants attended less to the scenario when it was disfluent and they had the option of delegating the decision to their doctor. Nonetheless, the influence of fluency on delegation is the same if the analysis is restricted only to those who accurately recalled all of the details of the scenario (χ2(1, N = 167) = 12.93, p = .001, ϕ = .28): 38% of participants delegated when the scenario was described in medical jargon, whereas only 14% delegated when the same information was in plain language, suggesting that the difference in delegation is unlikely to be due solely to comprehension or attention. Discussion Our initial experiments show that people do delegate decisions, and they are more likely to delegate when choices feel difficult. Choice difficulty increases choice delegation, and heightening the subjective feeling of difficulty is sufficient to increase delegation even when the objective content of the choices is the same. Moreover, people are willing to delegate difficult choices even when those choices have enduring, immediate, or important consequences. Although we tried to examine delegation in the context of decisions that are meaningful and consequential, an alternative interpretation of our findings thus far might be that people delegate difficult decisions because they do not wish to exert too much effort over choices that seem abstract or inconsequential to them. Perhaps the factors in studies 1a–c that made the choices feel difficult merely served to increase the necessary effort while simultaneously lowering the payoff of investing more effort to identify the best option. In the next experiment, we vary the importance of decisions along with their difficulty to explore the possibility that people delegate simply to avoid making effortful but relatively inconsequential decisions. EXPERIMENT 2: DOES THE IMPORTANCE OF THE DECISION MATTER? Perhaps people delegate difficult decisions not to avoid responsibility for potentially making a bad decision, but because they do not want to invest time and energy into decisions that are not worth the effort. In experiment 2, we used a scenario about choosing a new doctor to directly manipulate the importance of the choice consequences. For some participants, this choice was nonbinding and could be easily changed; for others, this choice was binding and changing was costly. We predict that choice difficulty, but not importance, prompts delegation, such that people delegate more when the choice of doctors feels difficult than when it feels easy, both when the choice seems important and when it seems unimportant. Method Participants Adults (N = 815) were recruited via Mechanical Turk to fill out an online survey for $0.25 Amazon.com credit. The target sample size was set a priori such that we would have about 200 participants per condition. Procedure Participants imagined that they were about to join a new health plan that required them to choose a new physician, using a scenario adapted from Schrift, Netzer, and Kivetz (2011) designed to test the effects of choice difficulty and importance on decision making. The physicians were described along three attributes: 1) office hours that did or did not include evenings and weekends, 2) a three- or 10-day wait time to get an appointment, and 3) home visits included or excluded. In the easy conditions, one physician had better office hours and wait times, the two attributes judged to be more important by participants in prior research. In the difficult conditions, each physician was better on only one of the important attributes, so that participants had to make a tradeoff. Participants in the high-importance conditions imagined that their choice was binding for a year and that switching physicians before the year ended would be difficult and require paying additional fees. Participants in the low-importance conditions imagined that their choice was nonbinding and they could easily switch doctors whenever they wanted without fees. Participants were told, “As you discuss this decision with your current physician, Dr. Brenner, he tells you that, if you are having trouble choosing a physician, he would be glad to choose one for you,” and were asked to indicate whether they would want to 1) choose which of the physicians they would prefer and name who that was, or 2) ask their current physician to choose a new physician for them. Additionally, participants rated how difficult the choice was on a scale ranging from 1 = very easy to 7 = very difficult and how important the decision was on a scale from 1 = not at all important to 7 = extremely important. Results Manipulation Checks As intended, participants rated choices to be more important when their choice was binding for a year and switching physicians would be difficult and costly (M = 4.93, SD = 1.29) than when their choice was not binding and they could easily switch physicians without penalty (M = 4.56, SD = 1.34; F(1, 803) = 15.41, p < .001, ηp2 = .02). Neither the difficulty manipulation (F(1, 803) = .13, p = .72, ηp2 < .001) nor the interaction between the difficulty and importance manipulations reliably affected rated importance (F(1, 803) = 2.92, p = .09, ηp2 = .004). Also as intended, participants rated the choice to be more difficult when they had to choose between physicians who were better on only one of the important attributes, meaning that they had to trade off the attributes (M = 3.69, SD = 1.65), than when one of the physicians was better on both attributes (M = 3.01, SD = 1.60; F(1, 803) = 34.93, p < .001, ηp2 = .04). Neither the importance manipulation (F(1, 803) = .08, p = .78, ηp2 < .001) nor the interaction between the difficulty and importance manipulations reliably affected rated choice difficulty (F(1, 803) = .02, p = .88, ηp2 < .001). Delegation Participants were more likely to delegate their choice of a new physician to their current physician when the choice felt difficult (44%) than when it felt easy (29%; Wald’s χ2 = 29.32, p < .001, odds ratio = 1.98), both when the choice was important (41% vs. 29%; χ2 = (1, N = 406) = 5.55, p = .02, ϕ = .12) and when it was less important (48% vs. 28%; χ2 = (1, N = 409) = 17.62, p < .001, ϕ = .21). Delegation did not vary as a function of decision importance (Wald’s χ2 = .44, p = .51, odds ratio = 1.10) or the interaction between difficulty and importance (Wald’s χ2 = 1.65, p = .20, odds ratio = 1.46). Mediation We next examined the process underlying the effect of our difficulty manipulation on delegation. A mediation model (PROCESS model 4) with our difficulty manipulation as the independent variable, rated difficulty as the mediator, and delegation as the dependent variable yielded significant mediation via rated difficulty (95% CI = –.65, –.30). Namely, our difficulty manipulation increased difficulty (β = –.68, SE = .11, t = –5.92, p < .001), and difficulty increased delegation (β = .68, SE = .06, Z = 11.76, p < .001). We did not include rated importance in the mediation model since it was unaffected by our difficulty manipulation; if included in the model as a parallel mediator, the pattern of results is the same, and rated importance does not mediate the relationship between the difficulty manipulation and delegation. Discussion Choice difficulty, but not importance, influences people’s likelihood to delegate. People are more likely to delegate when choices feel difficult than when they feel easy, and this does not vary depending on how important people feel the choices are. These findings are consistent with the notion that people delegate to avoid anticipated regret—conceptualized as the expectation that one will feel responsible for having chosen a suboptimal outcome and forgoing a better one (Zeelenberg et al. 2000)—and inconsistent with the alternate explanation that the degree to which people deem the choice worth deliberating about determines their likelihood of delegating. In our studies thus far, one could imagine that participants thought that the experimenter in experiments 1a and 1b or the doctor in experiments 1c and 2 might be an expert or at least know more about the options than they did. One might wonder whether people delegate not to avoid responsibility for potentially making a bad choice, but to put those decisions in more capable hands. Alternatively, if delegating difficult choices enables people to avoid feeling responsible for the choice outcome, then it might not matter whether the potential surrogate is an expert or not because delegation can still allow them to feel less responsible for the final outcome regardless of who else chooses. The next study explores whether consumers delegate even when the potential surrogate does not have special expertise pertinent to the decision. EXPERIMENT 3: DO SURROGATES NEED TO BE EXPERTS? Much of the past research on choice delegation in marketing has focused on cases in which consumers hire professionals with special expertise to make choices on their behalf, implicitly assuming that people delegate only to more knowledgeable others. Yet expertise may not be essential for delegation. Experiment 3 examines whether the expectation that the surrogate will have knowledge pertinent to the choice is a necessary determinant of delegation. We expected that participants would be more likely to delegate to surrogates with relevant expertise than to surrogates without expertise. But more importantly, we predicted that, regardless of the surrogate’s expertise, participants would be more likely to delegate when choices felt difficult than when they felt easy, choosing to transfer some of the responsibility for such choices to someone else even when the potential surrogate did not have special expertise pertinent to the judgment. Method Participants Adults (N = 403) were recruited via Mechanical Turk to fill out an online survey for $0.15 Amazon.com credit. The target sample size was set a priori such that we would have about 100 participants per condition. Procedure Participants imagined that they were out shopping with a friend and were interested in renting a movie from a movie rental kiosk featuring foreign films. Participants were told that their friend would not be able to watch the movie with them but was with them to help with another errand. Participants were presented with 10 foreign films, all in a foreign language with English subtitles (the original language was written in italics at the end of each movie’s description). Foreign films were chosen to minimize the likelihood that participants would be familiar with the movies. Indeed, 87% of participants indicated that they had not seen any of the movies. For each film, participants were provided with reformatted descriptions and images of the DVD covers that appeared on Netflix. Participants ranked the movies from 1 to 10, with the movie they were most interested in seeing at the top of the list (#1) and the movie they were least interested in watching at the bottom of the list (#10). Participants were able to click on each movie to drag and drop the movies into the appropriate order to indicate their rankings. On the next page, participants were told, “Once you have a chance to look over the movies, you realize that the kiosk is having a busy day, and the only two movies in the foreign film category that are currently available are these two movies.” Participants in the easy condition were presented with their fourth- and 10th-ranked movies, and those in the difficult condition were presented with their fourth- and fifth-ranked movies. Participants’ first-, second-, and third-ranked movies were not presented so as to avoid movies that might be more familiar to participants. In a pretest, 25 Mechanical Turk participants were shown both pairs of movies and, for each pair, rated how easy or difficult it would be to decide which of the two movies they would rather rent on a scale ranging from 1 = very easy to 10 = very difficult. Participants indicated that the decision would be easier to make when the movies were ranked further apart (M = 3.44, SD = 2.68) than closer together (M = 5.04, SD = 2.75; paired t(24) = 3.12, p = .005, d = 1.32). Once they were presented with the two movies they would be choosing between, participants in the expert condition were told, “When your friend sees you thinking about the decision, they tell you, ‘I have seen both of these movies, and I’d be glad to help you pick one if you need me to.’” Those in the nonexpert condition were instead told, “When your friend sees you thinking about the decision, they tell you, ‘I haven’t seen either of these movies, but I’d be glad to help you pick one if you need me to.’” In a pretest, 40 participants from Mechanical Turk were shown either the expert or nonexpert version of the scenario and were asked to rate, “To what extent do you think your friend has expertise that is relevant to this decision?” on a scale ranging from 1 = not at all to 7 = very much. Participants thought that their friend had greater expertise relevant to the decision in the expert condition (M = 5.70, SD = 1.17) than in the nonexpert condition (M = 3.75, SD = 1.62; t(38) = 4.36, p < .001, d = 1.41). Finally, participants were asked to indicate whether they would want to 1) choose which of the two movies they would rent on their own and name what it was or 2) ask their friend which of these two movies they should rent and rent that one. Results Participants were more likely to delegate to experts than nonexperts, but even so, delegation was not predicated on expertise. A logistic regression examining the effects of choice difficulty and surrogate expertise on preferences for choosing or delegating indicated that participants were more likely to delegate to a friend when that person was knowledgeable about the choice alternatives (68%) than when they were not (42%; Wald’s χ2 = 29.31, p < .001, odds ratio = 3.23). Additionally, as predicted, participants were more likely to delegate the choice to a friend in the difficult condition (66%) than in the easy condition (44%; Wald’s χ2 = 21.84, p < .001, odds ratio = 2.75). However, they did so regardless of the potential surrogate’s level of expertise (Wald’s χ2 = .01, p = .90, odds ratio = .95): participants were more likely to delegate the choice to a friend when the choice was difficult than when it was easy, both when that person had relevant expertise (79% vs. 57%; χ2 = (1, N = 201) = 11.37, p < .001, ϕ = .24) and when he or she did not (54% vs. 30%; χ2 = (1, N = 202) = 11.07, p < .001, ϕ = .23). Discussion People are more likely to delegate decisions to people with knowledge of the decision options than those without. However, delegation is not contingent upon the expectation that the surrogate decision maker has additional information or special expertise pertinent to the choice. People are more likely to delegate when choices feel difficult than when they feel easy, regardless of whether the potential surrogate has additional knowledge or experience regarding the choice options. This is consistent with the notion that what makes people more likely to delegate difficult than easy decisions is avoiding responsibility rather than seeking better insight. If people delegate to avoid responsibility for a decision, what might they perceive to be the benefits of avoiding responsibility for a decision? Perhaps people wish to avoid feeling disappointed if the choice outcome is less than ideal, so they delegate in hopes that the surrogate decision maker will make a better decision than they themselves would make. However, delegation may be driven less by the decision maker’s worry about the actual outcome, but instead more by their concern about their role in bringing it about. Although a nonexpert surrogate may not have special insight into a decision, they may have other qualities that make them an appealing alternative to choosing oneself—namely, the ability to take on some of the responsibility of making the choice. People may wish to avoid the regret that they might feel if they were responsible for making a poor choice, and they delegate in order to cede decision responsibility to another person. In the next study, to further probe what prompts people to delegate difficult decisions, we more directly assess the role of responsibility in delegation by examining the role of anticipated regret, in which responsibility for a decision outcome plays a role, and of anticipated disappointment, in which it does not, in the decision to delegate. EXPERIMENT 4: REGRET VERSUS DISAPPOINTMENT What prompts people to delegate—is it the desire to avoid the disappointment associated with a bad outcome, independent of whether they choose it or not, or is it the desire to avoid the regret associated with being responsible for bringing about that bad outcome? In experiment 4, we both manipulated choice difficulty by varying the difference in utility among the available options in the set (in this case, making them close or distant in appeal) and manipulated the overall attractiveness of the available options by varying the overall utility of the set of options, (making them all either appealing or unappealing). We also directly measured anticipated disappointment and regret in order to test whether they play a mediating role in prompting delegation. If people delegate merely to avoid disappointment, then they may be especially likely to delegate when the options are all unappealing and there is a good chance that the outcome of a decision will be unpleasant. If, as we predict, people delegate primarily to avoid regret, however, then they may delegate most when the options are close in appeal and it seems quite possible they might wish they had chosen the other option. Method Participants Adults (N = 823) were recruited via Mechanical Turk to complete an online survey for $.25 in Amazon.com credit. The target sample size was set a priori such that we would have about 200 participants per condition. Six participants did not complete the regret and disappointment measures. Procedure Participants were asked to imagine that they came across an ice cream truck selling 14 flavors of gourmet ice cream (The ice cream flavors, images, and descriptions came from Jeni’s Splendid Ice Cream [jenis.com].) Participants ranked the flavors from most appealing to least appealing by dragging and dropping the images into the appropriate order. Next, participants were told that only two of the 14 flavors were currently available. Participants in the easy attractive condition were shown their first and seventh ranked flavors, those in the easy unattractive condition were shown their eighth and 14th ranked flavors, those in the difficult attractive condition were shown their fourth and fifth ranked flavors, and those in the difficult unattractive condition were shown their eighth and ninth ranked flavors (note: pretesting indicated that participants found a choice between their two top-ranked flavors to be so easy that the difficult attractive condition was indistinguishable from the easy attractive condition, so to differentiate those conditions, we presented participants in the difficult attractive condition with slightly lower-ranked options). Participants were told, “As you think about the ice creams, you strike up a conversation with Jenny, the woman behind the counter. Jenny mentions that, if you are having a hard time choosing a flavor, that she has tried all of the flavors herself and would be glad to give you the flavor that she thinks you would like the best.” Participants then rated the extent to which they would prefer to choose themselves or delegate the decision to Jenny on a seven-point scale ranging from 1 = I would prefer to choose a flavor myself without any help from Jenny to 7 = I would prefer to have Jenny give me the flavor she thinks I would like the best. Next, to assess anticipated regret and disappointment, participants were asked, “If you had to choose by yourself which of the two available ice cream flavors to order, think about what would be going through your mind as you consider which flavor to choose. In particular, try to anticipate how you might feel if you are unhappy with the flavor that you choose.” Participants then responded to four regret items and four disappointment items, alternating between regret and disappointment, on a seven-point scale ranging from 1 = not at all to 7 = to a very great extent. Anticipated regret and disappointment items were adapted from Marcatto and Ferrante (2008) and Zeelenberg et al. (1998). See table 2 for all regret and disappointment items. Table 2 Structure Matrix for Regret and Disappointment, Experiment 4 Item:  Component 1: Disappointment  Component 2: Regret  How much might you NOT want to choose either of the two available flavors? (D)  .72  .37  How much might you feel powerless over getting a good flavor? (D)  .82  .53  How much might you feel like there is nothing you can do to ensure that you get a good flavor? (D)  .85  .50  How much might you feel that events that were beyond your control were the cause if you were unhappy with your flavor? (D)  .77  .27  How much might you feel that you should have known better than to choose the flavor that you did? (R)  .43  .76  How much might you think about what a mistake you made in choosing the flavor that you did? (R)  .59  .84  How much might you feel the tendency to kick yourself for choosing the flavor that you did? (R)  .60  .86  How much might you feel responsible if you are unhappy with your flavor? (R)  .11  .60  Item:  Component 1: Disappointment  Component 2: Regret  How much might you NOT want to choose either of the two available flavors? (D)  .72  .37  How much might you feel powerless over getting a good flavor? (D)  .82  .53  How much might you feel like there is nothing you can do to ensure that you get a good flavor? (D)  .85  .50  How much might you feel that events that were beyond your control were the cause if you were unhappy with your flavor? (D)  .77  .27  How much might you feel that you should have known better than to choose the flavor that you did? (R)  .43  .76  How much might you think about what a mistake you made in choosing the flavor that you did? (R)  .59  .84  How much might you feel the tendency to kick yourself for choosing the flavor that you did? (R)  .60  .86  How much might you feel responsible if you are unhappy with your flavor? (R)  .11  .60  NOTE.—Items marked D were intended to measure disappointment, and items marked R were intended to measure regret. Results Manipulation Checks As intended, participants rated choices to be more difficult when the options were similar in appeal (M = 3.38, SD = 1.68) than when one option was preferred much more than the other (M = 2.67, SD = 1.70; F(1, 813) = 37.17, p < .001, ηp2 = .04). They also rated choices to be more difficult when both options were poorly ranked (M = 3.22, SD = 1.74) than highly ranked (M = 2.83, SD = 1.68; F(1, 813) = 11.28, p = .001, ηp2 = .01). There was an interaction (F(1, 813) = 4.23, p = .04, ηp2 = .005), such that the choice was equally difficult regardless of the overall attractiveness of the set when the options were similar in appeal (Mattractive= 3.31, SD = 1.70 vs. Munattractive= 3.46, SD = 1.65; t(407) = .92, p = .36, d = .09), but when one option was preferred much more than the other, the choice was easier when the options were both highly ranked than when they were both poorly ranked (Mattractive= 2.35, SD = 1.51 vs. Munattractive= 2.99, SD = 1.81; t(392.41) = 3.83, p < .001, d = .38, equal variances not assumed). Also as intended, participants rated the choice options to be less attractive when they were both poorly ranked (M = 3.87, SD = 1.66) than highly ranked (M = 5.06, SD = 1.47; F(1, 813) = 121.36, p < .001, ηp2 = .13). There was an interaction (F(1, 813) = 30.71, p < .001, ηp2 = .04), such that the overall attractiveness of the choice set mattered more when participants preferred one option much more than the other (Mattractive= 5.37, SD = 1.37 vs. Munattractive= 3.59, SD = 1.65; t(390.72) = 11.87, p < .001, d = .17, equal variances not assumed) than when the options were similar in appeal (Mattractive= 4.74, SD = 1.50 vs. Munattractive= 4.15, SD = 1.62; t(407) = 3.82, p < .001, d = .38). Whether the options were similar in appeal (M = 4.45, SD = 1.59) or one option was preferred much more than the other (M = 4.48, SD = 1.76) did not influence rated attractiveness (F(1, 813) = .08, p = .78, ηp2 < .001). Disappointment and Regret A principle component analysis with Promax rotation on our anticipated disappointment and regret items yielded two distinct components (component 1 (disappointment) eigenvalue = 3.88; component 2 (regret) eigenvalue = 1.16; all other eigenvalues < .82). See table 2 for structure matrix. Moreover, our disappointment items (alpha = .77) and regret items (alpha = .81) hung together reliably. Therefore, we created composites for disappointment and regret by averaging each set of items together. Delegation Participants were more likely to delegate when the options were similar in appeal (M = 3.76, SD = 2.12) than when they preferred one option much more than the other (M = 3.37, SD = 2.13; F(1, 819) = 7.00, p = .008, ηp2 = .008)—in other words, when the choice was difficult rather than easy. There was also a marginal interaction (F(1, 819) = 3.57, p = .059, ηp2 = .004) such that our manipulation of choice difficulty was more influential when the options were both attractive (Mdifficult= 3.84, SD = 2.09 vs. Measy= 3.16, SD = 2.10; t(411) = 3.25, p = .001, d = .32) than when they were both unattractive (Mdifficult= 3.69, SD = 2.16 vs. Measy= 3.58, SD = 2.13; t(408) = .53, p = .60, d = .05). Overall, though, participants were no more likely to delegate when both options were appealing (M = 3.50, SD = 2.12) than when both options were unappealing (M = 3.64, SD = 2.14; F(1, 819) = .86, p = .36, ηp2 = .001). Mediation A serial mediation model (PROCESS model 6) with our choice difficulty manipulation as the independent variable; attractiveness, difficulty, disappointment, and regret as serial mediators; and delegation as the dependent variable yielded significant mediation via rated difficulty (95% CI = .18, .41), serially via rated difficulty and regret (95% CI = .003, .04), and serially via rated difficulty, disappointment, and regret (95% CI = .001, .03). None of the other indirect paths was significant. See figure 1 for the mediation model. Although we did not have an a priori prediction that disappointment would predict regret, having disappointment precede regret in the model is consistent with the idea that one must first have a bad outcome (one that is disappointing) before one has something to regret. FIGURE 1 View largeDownload slide MEDIATION MODEL, EXPERIMENT 4 *p < .05, ***p < .001 NOTES.—X = choice difficulty manipulation; Y = choice delegation; M1 = rated unattractiveness; M2 = rated difficulty; M3 = anticipated disappointment; M4 = anticipated regret Indirect effect of X on Y through M1 only: 95% CI = (–.03, .02) Indirect effect of X on Y through M2 only: 95% CI = (.18, .41)* Indirect effect of X on Y through M3 only: 95% CI = (–.04, .01) Indirect effect of X on Y through M4 only: 95% CI = (–.05, .01) Indirect effect of X on Y through M1 and M2 in serial: 95% CI = (–.02, .03) Indirect effect of X on Y through M1 and M3 in serial: 95% CI = (–.005, .01) Indirect effect of X on Y through M1 and M4 in serial: 95% CI = (–.006, .003) Indirect effect of X on Y through M2 and M3 in serial: 95% CI = (–.02, .04) Indirect effect of X on Y through M2 and M4 in serial: 95% CI = (.003, .04)* Indirect effect of X on Y through M3 and M4 in serial: 95% CI = (–.03, .01) Indirect effect of X on Y through M1, M2, and M3 in serial: 95% CI = (–.001, .002) Indirect effect of X on Y through M1, M2, and M4 in serial: 95% CI = (–.001, .002) Indirect effect of X on Y through M1, M3, and M4 in serial: 95% CI = (–.005, .01) Indirect effect of X on Y through M2, M3, and M4 in serial: 95% CI = (.001, .03)* Indirect effect of X on Y through M1, M2, M3, and M4 in serial: 95% CI = (–.001, .002) Direct effect of X on Y: 95% CI = (–.20, .36) FIGURE 1 View largeDownload slide MEDIATION MODEL, EXPERIMENT 4 *p < .05, ***p < .001 NOTES.—X = choice difficulty manipulation; Y = choice delegation; M1 = rated unattractiveness; M2 = rated difficulty; M3 = anticipated disappointment; M4 = anticipated regret Indirect effect of X on Y through M1 only: 95% CI = (–.03, .02) Indirect effect of X on Y through M2 only: 95% CI = (.18, .41)* Indirect effect of X on Y through M3 only: 95% CI = (–.04, .01) Indirect effect of X on Y through M4 only: 95% CI = (–.05, .01) Indirect effect of X on Y through M1 and M2 in serial: 95% CI = (–.02, .03) Indirect effect of X on Y through M1 and M3 in serial: 95% CI = (–.005, .01) Indirect effect of X on Y through M1 and M4 in serial: 95% CI = (–.006, .003) Indirect effect of X on Y through M2 and M3 in serial: 95% CI = (–.02, .04) Indirect effect of X on Y through M2 and M4 in serial: 95% CI = (.003, .04)* Indirect effect of X on Y through M3 and M4 in serial: 95% CI = (–.03, .01) Indirect effect of X on Y through M1, M2, and M3 in serial: 95% CI = (–.001, .002) Indirect effect of X on Y through M1, M2, and M4 in serial: 95% CI = (–.001, .002) Indirect effect of X on Y through M1, M3, and M4 in serial: 95% CI = (–.005, .01) Indirect effect of X on Y through M2, M3, and M4 in serial: 95% CI = (.001, .03)* Indirect effect of X on Y through M1, M2, M3, and M4 in serial: 95% CI = (–.001, .002) Direct effect of X on Y: 95% CI = (–.20, .36) Discussion Although both anticipated disappointment and regret may contribute to delegation, only anticipated regret directly leads people to delegate choices to others. This is consistent with the notion that people delegate choices when they are concerned that they might feel responsible for a less-than-optimal outcome. This key benefit of delegation—enabling people to avoid feeling responsible if a choice is suboptimal while still allowing them to get something out of the choice—may make it a uniquely appealing method for resolving difficult decisions, even when other options for resolution are available. Another option that people often have available to them to escape tough choices is to opt out of decisions and refuse to choose anything at all. While both delegation and opting out may be commonly used methods of avoiding difficult decisions when either option is the sole alternative to choosing, when delegation and opting out are both options, delegation may often be a more appealing means of handling difficult decisions than opting out, because it allows people to avoid responsibility and regret while still getting something out of the decision. Opting out, on the other hand, may be preferred to delegation when the options are all unappealing, since it allows people to guarantee that they avoid an unappealing outcome. We explore these predictions in the next experiment. EXPERIMENT 5: DELEGATION VERSUS OPTING OUT In this study, we varied choice difficulty and the overall appeal of the choice options and gave participants the option to choose themselves, delegate, or opt out. Since delegation enables people to avoid responsibility while still allowing them to realize the benefits of a chosen option, people may prefer delegation when the choice is difficult so long as the set of options is somewhat appealing, and especially if those options are highly appealing. And because opting out enables people to avoid the consequences of a choice altogether, people may uniquely prefer to opt out of choosing entirely when all options are unappealing rather than appealing. Method Participants Adults (N = 590) were recruited via Mechanical Turk to complete an online survey for $.25 in Amazon.com credit. The target sample size was set a priori such that we would have about 150 participants per condition. Procedure This study followed the same procedure as in experiment 4, except the regret and disappointment items were removed and participants had three options: 1) “I would prefer to choose a flavor myself without any help from Jenny”; 2) “I would prefer to have Jenny give me the flavor she thinks I would like the best”; and 3) “I would prefer NOT to get any ice cream.” Results A multinomial logistic regression with our choice difficulty and overall attractiveness manipulations as predictors and choice method (choose, delegate, opt out, with choice as the reference category) as the outcome indicated that participants were more likely to delegate than choose themselves when the choice was difficult than when it was easy (42% vs. 32%; Wald’s χ2 = 8.88, p = .003, odds ratio = .48), but not when the options were both appealing versus when they were both unappealing (38% vs. 36%; Wald’s χ2 < .001, p = .99, odds ratio = 1.00). Although the interaction between choice difficulty and overall appeal did not significantly influence participants’ preference to delegate rather than choose themselves (Wald’s χ2 = 2.61, p = .11, odds ratio = 1.76), an examination of the simple effects revealed that participants were more likely to prefer delegation when the options were similar in appeal and choices were difficult (44%) than when they were distant in appeal and choices were easy (27%) when both available options were appealing (Wald’s χ2 = 8.88, p = .003, odds ratio = .48), but participants were equally likely to prefer delegation of easy and difficult choices when both options were unappealing (41% vs. 36%; Wald’s χ2 = .50, p = .48, odds ratio = .83). Perhaps unsurprisingly, participants were more likely to opt out of the decision and not pick an ice cream at all when the options were both unappealing. Because so few people opted out, we ran a Firth penalized maximum likelihood estimation, which showed that opting out was more common in the unappealing conditions than the appealing conditions (9% vs. 1%; Wald’s χ2 = 18.97, p < .001, odds ratio = .11), but was just as common when the choice was difficult than when it was easy (4% vs. 5%; Wald’s χ2 = .49, p = .48, odds ratio = .74). The interaction between choice difficulty and the overall appeal did not significantly influence the likelihood with which participants preferred to opt out rather than choose themselves (Wald’s χ2 = .06, p = .81). Participants opted out more when the options were all unappealing versus when they were all appealing, both when the options were similar in appeal and the choice was difficult (8% vs. 1%; Wald’s χ2 = 9.15, p = .002, odds ratio < .001) and when they were distant in appeal and the choice was easy (10% vs. 1%; Wald’s χ2 = 9.97, p = .002, odds ratio = .13). Discussion People prefer to delegate rather than opt out of difficult decisions, so long as the available options are somewhat appealing. Participants in this experiment were more likely to delegate when the options were similar in appeal rather than when one option was much more appealing than the other, especially when the options were both appealing. And, as one might expect, people were more likely to opt out of choosing when the available options were unappealing. These findings hint at a possible solution for retailers looking to encourage consumers to make a purchase when faced with challenging decisions, a situation in which consumers might otherwise be inclined to walk away empty-handed. If consumers often prefer to delegate rather than opt out of difficult decisions, then giving consumers the option to delegate might reduce or even reverse the tendency for them to opt out of difficult decisions. In the next study, we examine whether consumers are less likely to walk away from difficult decisions empty-handed when given the option to delegate than when their only options are to choose or walk away. EXPERIMENT 6: DELEGATION AS A MEANS OF DECREASING OPTING OUT Shoppers may sometimes be less likely to make a purchase from a large array of options than from a small array, despite the fact that a greater number of options should only increase the likelihood that one of those options is appealing (Iyengar and Lepper 2000). This experiment examines how giving shoppers the option to delegate affects their purchase likelihood when they are presented with choices between many or few products. Participants were presented with a small or large selection of teas. Some participants were asked whether they would like to purchase a tea or opt out of the decision, and other participants were asked whether they would like to purchase a tea, opt out, or ask the salesperson to choose for them. We predicted that participants would be equally likely to make a purchase from a small array regardless of whether they had the option to delegate, but that participants with the option to delegate would be more likely to make a purchase from a large array than those without the option to delegate. Method Participants Undergraduates (N = 151) at the University of Florida participated in exchange for extra course credit. The sample size was set a priori to encompass one week’s worth of participants in the lab. Procedure Participants imagined shopping at a gourmet tea shop. They were randomly assigned to see an array of 30 teas (large set condition) presented in a randomized order or six teas (small set condition) randomly selected from the large set. They saw an image of each tea and its ingredients, price, and popularity rating, with a checkbox alongside. In the salesperson-not-present conditions, participants chose whether they preferred to: 1) “Choose a package of tea yourself, and buy that one (If you select this option, check the box next to the ONE tea you would like to buy),” or 2) “Pass for the time being (If you select this option, leave all boxes EMPTY and continue to the next page).” Others assigned to the salesperson-present condition were told, “A store employee approaches you and asks if you would like any help in choosing a tea,” and were given the option to 1) choose, 2) pass for the time being, or 3) “Tell the store employee which teas you are considering, and buy whichever one he/she recommends (If you select this option, check boxes next to ANY teas that you are considering).” Lastly, participants indicated whether they liked tea and how often they drank it. Nineteen participants were excluded because they did not select the appropriate number of teas given their stated preference to choose themselves, delegate, or opt out, leaving 132 participants. The pattern of results remains the same when these participants were included, as well as when we control for tea liking and drinking. Results We examined how set size and the presence of a salesperson affected whether or not participants opted to purchase tea. A purchase in this study refers to when a participant would leave the shop with a tea, either because they chose a tea on their own or because they delegated the choice to the salesperson. A logistic regression testing the effects of set size and salesperson presence on likelihood of purchase indicated that, overall, participants were equally likely to make a purchase from a large array (59%) and a small array (63%; Wald’s χ2 = .01, p = .92, odds ratio = 1.04), and a greater percentage of participants purchased tea when a salesperson offered to help them choose (74%) than when a salesperson was not present (46%; Wald’s χ2 = 10.22, p = .001, odds ratio = 3.34). Importantly, as predicted, there was a significant interaction between set size and salesperson presence on purchases (Wald’s χ2 = 8.90, p = .003, odds ratio = 11.27). See figure 2. FIGURE 2 View largeDownload slide PERCENTAGE OF PARTICIPANTS WHO PURCHASED TEA BY SET SIZE AND PRESENCE OF THE OPTION TO DELEGATE TO A SALESPERSON, EXPERIMENT 6 FIGURE 2 View largeDownload slide PERCENTAGE OF PARTICIPANTS WHO PURCHASED TEA BY SET SIZE AND PRESENCE OF THE OPTION TO DELEGATE TO A SALESPERSON, EXPERIMENT 6 We can examine this interaction by breaking down the results by set size. Although purchases did not differ based on whether a salesperson was present (61%) or absent (64%) when there were few options (χ2(1, N = 56) = .02, p = .89, ϕ = .02), more participants purchased tea when a salesperson was present (86%) than when a salesperson was not present when there were many options (37%; χ2(1, N = 76) = 18.87, p < .001, ϕ = .50). Or, if we instead break down the interaction by the presence of a salesperson, fewer participants purchased tea when there were many available tea varieties (37%) than when there were few (64%) when a salesperson was not present (χ2(1, N = 63) = 4.22, p = .04, ϕ = .26), but more participants purchased tea when there were many available tea varieties (86%) than when there were few (61%) when a salesperson was present (χ2(1, N = 69) = 5.13, p = .02, ϕ = .27). Discussion Ensuring that consumers have access to salespeople or other surrogates to whom they can delegate difficult decisions can reduce and even reverse the tendency to walk away from choices empty-handed. Participants were more likely to make a purchase from a large array of options when they could delegate the choice to another person than when they could not. And, whereas fewer participants made a purchase from a large array of options than a small array when delegation was not an option, more participants made a purchase from a large array when it was. GENERAL DISCUSSION This research shows that consumers cope with difficult decisions by recruiting others to choose for them. Across eight experiments, participants were more likely to ask others to choose on their behalf when choices felt difficult than when they felt easy. Delegation increased when choices felt difficult regardless of whether that feeling was because the choices themselves were more difficult (e.g., with a larger number of alternatives, with difficult tradeoffs, or with a smaller difference in relative attractiveness between the alternatives), or because the choices were processed less fluently for superficial reasons (e.g., the options were presented in jargon). Delegation increased when choices felt difficult regardless of whether the consequences were real or hypothetical and regardless of the importance of those consequences. Our findings suggest that the decision to delegate a difficult choice is rooted in the desire to avoid the potential regret associated with feeling responsible for a bad choice, rather than the desire to simply avoid the possibility of a bad outcome. Delegation increased when choices felt difficult regardless of whether the potential surrogate was an expert or nonexpert, consistent with the notion that people delegate to avoid responsibility for potentially making a suboptimal decision rather than to put decisions in more capable hands. Moreover, the degree to which participants anticipated regretting a poor choice, but not the degree to which they expected to be disappointed by a poor outcome, directly predicted the tendency to delegate. Consequently, delegation is often a more attractive option for avoiding a difficult choice than opting out of choosing altogether, and giving consumers the option to delegate can reduce the tendency for consumers to walk away from difficult decisions empty-handed. One might wonder why participants in our studies were willing to delegate difficult decisions when past research on the preference for choosing has shown that people overwhelmingly prefer to make decisions themselves, even when doing so is cognitively or emotionally costly (Botti et al. 2009). A potentially important distinction between how preference for choice is typically studied and our everyday examples of choice delegation is that, unlike in our studies, whether participants choose themselves or another person chooses for them is typically externally imposed in past research. The fact that participants assigned to have another person choose on their behalf were deprived of their freedom to choose could explain why they indicated that they would have preferred to choose themselves even when having another person choose on their behalf led them to feel more satisfied with the choice outcomes. Thus, although the preference for choice literature shows that people prefer choosing themselves when who gets to make the final choice is externally dictated, even when faced with difficult decisions between options with negative or tragic consequences (Botti et al. 2009), people may prefer to voluntarily cede decision control when considering difficult decisions in prospect. Directions for Future Research Our research is among the first to examine delegation in everyday contexts, leaving many open questions for future research about the extent and impact of delegation as a form of decision support. For one, to what extent might the present findings extend to advice seeking? Advice seeking can provide many of the same benefits as delegation for mitigating uncertainty and reducing the risk of making a bad choice: for example, it can provide additional information or validation of one’s inclinations. Thus, many factors that prompt delegation are also likely to prompt advice seeking. In fact, when avoiding being disappointed by a bad outcome is of primary concern, advice seeking may be more attractive than delegation in that it provides many of the same benefits while still allowing people to maintain control over the final decision. In our work, we focused on the antecedents of the decision to delegate, but decision makers are not the only ones with emotional and other stakes in the process. Future work might investigate how surrogates react to being asked to make a choice on someone else’s behalf. People are often flattered by being asked for advice (Brooks, Gino, and Schweitzer, 2015), for instance. But people can find making a choice on behalf of others to be stressful, and are willing to seek out second-order surrogates to make choices for others for them (Steffel, Williams, and Perrmann-Graham, 2016). This work shows that people are most likely to “pass the buck,” so to speak, when there is the potential for feeling responsible and being blamed for someone else’s negative outcome (Steffel et al. 2016). What are other factors that influence a potential surrogate’s willingness to make a choice for someone else? Surrogate expertise did not differentially affect our participants’ likelihood of delegating, but might it affect the surrogates’ willingness to choose? Further, there are many people who are expected to make decisions on behalf of others like our surrogates here do, like parents, spouses, doctors, and bosses. How do they react to this responsibility, and what makes them take on or abdicate this responsibility? Practical and Theoretical Implications Understanding the conditions under which consumers are more likely to delegate choices can help marketers to better target marketing campaigns to the people who will ultimately be making the purchasing decision. Depending on who the primary decision makers are, marketers may allocate resources differently across different elements of the marketing mix. For example, a pharmaceutical company might allocate fewer resources to advertising and more resources to personal selling in situations where consumers are more prone to delegate the choice of what drug to take to their doctors. Additionally, marketers may position their products differently when professional surrogates are the decision makers as opposed to consumers, because those professionals may have a different level of expertise or a different set of considerations than consumers. For example, a pharmaceutical company may market a drug to consumers by emphasizing lifestyle benefits but market the same drug to doctors by emphasizing performance in clinical trials. This research can also help retailers and service providers identify and make the most of situations in which consumers are likely to delegate choices to salespeople. Unsolicited product recommendations are discounted relative to solicited recommendations and may even evoke reactance (Fitzsimons and Lehmann 2004). Knowing the conditions under which consumers are more likely to delegate choices can help businesses improve the quality of their customer service and the effectiveness of their salespeople by helping them lend decision support when and where it would be most desired. For example, retailers may better position salespeople by concentrating them around products that are difficult to differentiate or for which there are many alternatives. It also can help retailers identify opportunities in which salespeople are likely to have more influence in the decision process and can be used more effectively to move merchandise. Despite the prevalence of choice delegation in today’s marketplace and the tremendous value in being able to predict when consumers are likely to delegate, the antecedents and implications of delegation are not well understood. This present research provides new insight into the preference for choice determination by identifying regret and responsibility as determinants of whether people prefer freedom of choice or freedom from choice, suggesting that the anticipated regret associated with feeling responsible for a bad choice can overwhelm people’s general preference for making their own choices. The present research contributes to an understanding of surrogate usage by identifying when consumers are most likely to employ others to make decisions on their behalf and by showing that expertise is a desirable but nonessential feature of a potential surrogate. Finally, this research contributes to the choice avoidance literature in particular, and the judgment and decision making literature more broadly, by showing that individual decision making is only a small part of a larger picture and that decisions, including the decision whether or not to decide in the first place, often hinge upon the social context and whether people have the option to delegate those choices to others. DATA COLLECTION INFORMATION The second author supervised the collection of data by research assistants at the University of California, San Diego, for experiment 1a in spring 2016 and for experiment 1c in spring 2014. The first author supervised the collection of data for experiment 1b by research assistants at the University of Cincinnati in spring 2015. The first and second authors jointly managed the collection of data using MTurk for experiment 3 in fall 2015, and the pilot experiment and experiments 2, 4, and 5 in fall 2016. The first and second authors jointly managed the collection of data for experiment 6 by research assistants at the University of Florida in fall 2011. The first and second authors jointly analyzed the data. The authors thank Paige Bausch, Mark Havens, Mercedes Castro, Peter Chami, Ana Figueras, Bernardo Guzman, Danielle King, Gia Nardini, Shane Strunk, and David Vitt for assistance with data collection; Lauren McCarthy and Megan Parin for assistance with information gathering; Julia Le, Matteo Prandi, Isabella Mozdzierz-Monico, and Apoorva Siddhi Sudhir for assistance with survey creation; and Chris Janiszewski, Nathan Novemsky, Aner Sela, and three anonymous reviewers for helpful comments APPENDIX MEDICAL TREATMENT CHOICE SCENARIO, EXPERIMENT 1C Plain language version: Imagine waking up to this scenario and being faced with an important life decision: Hello, my name is Dr. Bunnalai, and I am a doctor with the critical care team here at the hospital. You were in a car accident. One of the bones in your neck was damaged, and the spinal cord in your neck is severely bruised and bleeding. You have two options. Option 1: You can choose to undergo surgery to remove the blood that has collected around your spinal cord and to stop the bleeding. After the surgery, for the rest of your life, you may experience some constant neck pain and a tingling or burning sensation that will sometimes spread through your arms and legs. Option 2: Alternatively, you can choose NOT to undergo surgery. Instead, you will receive drugs to control your pain as needed. However, if the blood that has collected is not removed and the bleeding is not stopped, the blood will continue to press on your spinal cord and may permanently damage your spinal cord and cause paralysis. Medical jargon version: Imagine waking up to this scenario and being faced with an important life decision: Hello, my name is Dr. Bunnalai, and I am an Attending Intensivist with the trauma critical care medicine team here at the Level One Trauma Center. You were involved in a MVA, a motor vehicle accident. You sustained blunt force trauma injuries to your C4 vertebrae, part of the cervical spinal column in your neck, and you have a hematomyelia, an intramedullary spinal cord hemorrhage. You have two options. Option 1: You can choose to undergo surgical evacuation of the hematoma, the extravasated blood that has collected around your C4 vertebrae, and endovascular embolization to selectively occlude the affected blood vessels and arrest the bleeding. Post-operation, for the duration of your life, you may experience some chronic cervicalgia, a painful sensation that is localized to your neck, and dysesthesia, a painful tingling or burning sensation that will intermittently radiate though your extremities. Option 2: Alternatively, you can choose NOT to undergo surgery. Instead, you will receive palliative care consisting of PRN intravenous analgesics to control your pain. However, if the hematoma remains unextracted and blood vessels remain patent, the hematoma will continue to impinge on your spinal cord and may permanently damage your spinal cord and cause paralysis. REFERENCES Alter Adam L., Oppenheimer Daniel M. ( 2009), “Uniting the Tribes of Fluency to Form a Metacognitive Nation,” Personality and Social Psychology Review , 13 3, 219– 35. Google Scholar CrossRef Search ADS PubMed  Anderson Christopher J. ( 2003), “The Psychology of Doing Nothing: Forms of Decision Avoidance Result from Reason and Emotion,” Psychological Bulletin , 129 1, 129– 67. Google Scholar CrossRef Search ADS   Baron Jonathan, Ritov Ilana ( 1994), “Reference Points and Omission Bias,” Organizational Behavior and Human Decision Processes , 59 3, 475– 98. Google Scholar CrossRef Search ADS   Botti Simona, Iyengar Sheena S. ( 2004), “The Psychological Pleasure and Pain of Choosing: When People Prefer Choosing at the Cost of Subsequent Outcome Satisfaction,” Journal of Personality and Social Psychology , 87 3, 312– 26. Google Scholar CrossRef Search ADS PubMed  Botti Simona, McGill Ann L. ( 2006), “When Choosing Is Not Deciding: The Effect of Perceived Responsibility on Satisfaction,” Journal of Consumer Research , 33 2, 211– 9. Google Scholar CrossRef Search ADS   Botti Simona, McGill Ann L. ( 2011), “The Locus of Choice: Personal Causality and Satisfaction with Hedonic and Utilitarian Decisions,” Journal of Consumer Research , 37 6, 1065– 78. Google Scholar CrossRef Search ADS   Botti Simona, Orfali Kristina, Iyengar Sheena S. ( 2009), “Tragic Choices: Autonomy and Emotional Responses to Medical Decisions,” Journal of Consumer Research , 36 3, 337– 52. Google Scholar CrossRef Search ADS   Brehm Jack W. ( 1956), “Postdecision Changes in the Desirability of Alternatives,” Journal of Abnormal and Social Psychology , 52 3, 384– 9. Google Scholar CrossRef Search ADS   Brooks Alison Wood, Gino Francesca, Schweitzer Maurice E. ( 2015), “Smart People Ask for (My) Advice: Seeking Advice Boosts Perceptions of Competence,” Management Science , 61 6, 1421– 35. Google Scholar CrossRef Search ADS   Charles Cathy, Whelan Tim, Gafni Amiram, Reyno Leonard, Redko Cristina ( 1998), “Doing Nothing Is No Choice: Lay Constructions of Treatment Decision-Making among Women with Early-Stage Breast Cancer,” Sociology of Health and Illness , 20 1, 71– 95. Google Scholar CrossRef Search ADS   DeCharms Richard ( 1968), Personal Causation , New York: Academic Press. Dhar Ravi ( 1996), “The Effect of Decision Strategy on Deciding to Defer Choice,” Journal of Behavioral Decision Making , 9 4, 265– 81. Google Scholar CrossRef Search ADS   Dhar Ravi ( 1997a), “Consumer Preference for a No-Choice Option,” Journal of Consumer Research , 24 2, 215– 31. Google Scholar CrossRef Search ADS   Dhar Ravi ( 1997b), “Context and Task Effects on Choice Deferral,” Marketing Letters , 8 1, 119– 30. Google Scholar CrossRef Search ADS   Dhar Ravi, Nowlis Stephen M. ( 1999), “The Effect of Time Pressure on Consumer Choice Deferral,” Journal of Consumer Research , 25 4, 369– 84. Google Scholar CrossRef Search ADS   Elwyn Glyn, O’Connor Annette, Stacey Dawn, Volk Robert, Edwards Adrian, Coulter Angela, Thomson Richard, Barratt Alexandra, Barry Michael, Bernstein Steven, Butow Phyllis, Clarke Aileen, Entwistle Vikki, Feldman-Stewart Deb, Holmes-Rovner Margaret, Llewellyn-Thomas Hilary, Moumjid Nora, Mulley Al, Ruland Cornelia, Sepucha Karen, Sykes Alan, Whelan Tim ( 2006), “International Patient Decision Aids Standards (IPDAS) Collaboration: Developing A Quality Criteria Framework For Patient Decision Aids: Online International Delphi Consensus Process,” British Medical Journal , 333 7565: 417. Google Scholar CrossRef Search ADS PubMed  Festinger Leon ( 1964), Conflict, Decision, and Dissonance , Stanford, CA: Stanford University Press. Fitzsimons Gavan J., Lehmann Donald R. ( 2004), “Reactance to Recommendations: When Unsolicited Advice Yields Contrary Responses,” Marketing Science , 23 1, 82– 94. Google Scholar CrossRef Search ADS   Forsythe Sandra, Butler Sara, Schaefer Robert ( 1990), “Surrogate Usage in the Acquisition of Women’s Business Apparel,” Journal of Retailing , 66 4, 446– 69. Fuller Barbara K., Blackwell Suzannah C. ( 1992), “Wardrobe Consultant Clientele: Identifying and Describing Three Market Segments,” Clothing and Textiles Research Journal , 10 2, 11– 7. Google Scholar CrossRef Search ADS   Iyengar Sheena S., Lepper Mark R. ( 2000), “When Choice Is Demotivating: Can One Desire Too Much of a Good Thing?” Journal of Personality and Social Psychology , 79 6, 995– 1006. Google Scholar CrossRef Search ADS PubMed  Janis Irving L., Mann Leon ( 1976), “Coping with Decision Conflict: An Analysis of How Stress Affects Decision-Making Suggests Interventions to Improve the Process,” American Scientist , 64 6, 657– 67. Kenny Patricia, Quine Susan, Shiell Alan, Cameron Sue ( 1999), “Participation in Treatment Decision-Making by Women with Early Stage Breast Cancer,” Health Expectations , 2 3, 159– 68. Google Scholar CrossRef Search ADS PubMed  Langer Ellen J., Rodin Judith ( 1976), “The Effects of Choice and Enhanced Personal Responsibility for the Aged: A Field Experiment in an Institutional Setting,” Journal of Personality and Social Psychology , 34 2, 191– 8. Google Scholar CrossRef Search ADS PubMed  Luce Mary Frances ( 1998), “Choosing to Avoid: Coping with Negatively Emotion-Laden Consumer Decisions,” Journal of Consumer Research , 24 4, 409– 33. Google Scholar CrossRef Search ADS   Marcatto Francesco, Ferrante Donatella ( 2008), “The Regret and Disappointment Scale: An Instrument for Assessing Regret and Disappointment in Decision Making,” Judgment and Decision Making , 3 1, 87– 99. McLaughlin G. Harry ( 1969), “SMOG Grading – A New Readability Formula,” Journal of Reading , 12 8, 639– 46. Novemsky Nathan, Dhar Ravi, Schwarz Norbert, Simonson Itamar ( 2007), “Preference Fluency in Choice,” Journal of Marketing Research , 44 3, 347– 56. Google Scholar CrossRef Search ADS   Ordóñez Lisa D., Connolly Terry ( 2000), “Regret and Responsibility: A Reply to Zeelenberg et al.  ,” Organizational Behavior and Human Decision Processes , 81 1, 132– 42. Google Scholar CrossRef Search ADS PubMed  Park C. Whan, Jun Sung Youl, MacInnis Deborah J. ( 2000), “Choosing What I Want Versus Rejecting What I Don’t Want: An Application of Decision Framing to Product Option Choice Decisions,” Journal of Marketing Research , 37 2, 187– 202. Google Scholar CrossRef Search ADS   Pratt Gerry ( 1981), “The House as an Expression of Social Worlds,” in Housing and Identity: Cross-Cultural Perspectives , ed. Duncan James S., London: Croon Helm, 135– 79. Preacher Kris J., Hayes Andrew F. ( 2008) “Asymptotic and Resampling Strategies for Assessing and Comparing Indirect Effects in Multiple Mediator Models,” Behavior Research Methods , 40 3, 879– 91. Google Scholar CrossRef Search ADS PubMed  Redelmeier Donald A., Shafir Eldar ( 1995), “Medical Decision Making in Situations that Offer Multiple Alternatives,” Journal of the American Medical Association , 273 4, 302– 5. Google Scholar CrossRef Search ADS PubMed  Rudd Rima E., Moeykens Barbara A., Colton Tayla C. ( 1999), “Health and Literacy: A Review of Medical and Public Health Literature,” in Annual Review of Adult Learning and Literacy.  ed. Comings John, Garners Barbara, Smith Christine, New York: Jossey-Bass, 175– 203. Ryan Richard M., Deci Edward L. ( 2000), “Self-Determination Theory and the Facilitation of Intrinsic Motivation, Social Development, and Well-Being,” American Psychologist , 55 1, 68– 78. Google Scholar CrossRef Search ADS PubMed  Schrift Rom, Netzer Oded, Kivetz Ran ( 2011), “Complicating Choice,” Journal of Marketing Research , 48 2, 308– 26. Google Scholar CrossRef Search ADS   Seligman Martin E. ( 1975), Helplessness: On Depression, Development, and Death , San Francisco: Freeman. Simonson Itamar ( 1992), “The Influence of Anticipating Regret and Responsibility on Purchase Decisions,” Journal of Consumer Research , 19 1, 105– 18. Google Scholar CrossRef Search ADS   Solomon Michael R. ( 1987), “The Wardrobe Consultant: Exploring the Role of a New Retailing Partner,” Journal of Retailing , 63 2, 110– 28. Sugden Robert ( 1985), “Regret, Recrimination, and Rationality,” Theory and Decision , 19 1, 77– 99. Google Scholar CrossRef Search ADS   Steele Claude M. ( 1988), “The Psychology of Self-Affirmation: Sustaining the Integrity of the Self,” in Advances in Experimental Social Psychology , Vol. 21, ed. Berkowitz Leonard, New York: Academic Press, 261– 2. Steffel Mary, Williams Elanor F. ( 2015), “Empowering Patients to Be Consumers of Their Own Medical Care: Increasing the Subjective Ease of Medical Decisions Increases Patient Participation,” working paper, Northeastern University, Boston, MA 02115. Steffel Mary, Williams Elanor F., Perrmann-Graham Jaclyn ( 2016), “Passing the Buck: Delegating Choices to Others to Avoid Blame,” Organizational Behavior and Human Decision Processes , 135 ( July), 32– 44. Google Scholar CrossRef Search ADS   Stern Barbara B., Solomon Michael R., Stinerock Robert ( 1992), “Surrogate Money Managers in the 1990s: Marketing Strategy for Financial Services Retailers,” Service Industries Journal , 12 1, 78– 96. Google Scholar CrossRef Search ADS   Stinerock Robert, Stern Barbara B., Solomon Michael R. ( 1991), “Sex and Money: Gender Differences in the Use of Surrogate Consumers for Financial Decision Making,” Journal of Professional Services Marketing , 7 2, 167– 82. Google Scholar CrossRef Search ADS   Taylor Shelley E., Brown Jonathon D. ( 1988), “Illusion and Well-Being: A Social Psychological Perspective on Mental Health,” Psychological Bulletin , 103 2, 193– 210. Google Scholar CrossRef Search ADS PubMed  Tetlock Philip E., Boettger Richard ( 1994), “Accountability Amplifies the Status Quo Effect When Change Creates Victims,” Journal of Behavioral Decision Making , 7 1, 1– 23. Google Scholar CrossRef Search ADS   Tversky Amos, Shafir Eldar ( 1992), “Choice under Conflict: The Dynamics of Deferred Decision,” Psychological Science , 3 6, 358– 61. Google Scholar CrossRef Search ADS   Tykocinski Orit E., Pittman Thane S. ( 1998), “The Consequences of Doing Nothing: Inaction Inertia as Avoidance of Anticipated Counterfactual Regret,” Journal of Personality and Social Psychology , 75 3, 607– 16. Google Scholar CrossRef Search ADS   Usta Murat, Häubl Gerald ( 2011), “Self-Regulatory Strength and Consumers’ Relinquishment of Decision Control: When Less Effortful Decisions Are More Resource Depleting,” Journal of Consumer Research , 48 2 403– 12. van Dijk Wilco W., Zeelenberg Marcel, van der Pligt Joop ( 2003), “Blessed Are Those Who Expect Nothing: Lowering Expectations as a Way of Avoiding Disappointment,” Journal of Economic Psychology , 24 4, 505– 16. Google Scholar CrossRef Search ADS   Williams Elanor F., Gneezy Ayelet, Armor David A. ( 2015), “Between a Rock and Another Rock: Ironic Persistence at Choices with Interchangeable Options,” working paper, Kelley School of Business, Indiana University, Bloomington, IN 47405. Zeelenberg Marcel ( 1999), “Anticipated Regret, Expected Feedback, and Behavioral Decision Making,” Journal of Behavioral Decision Making , 12 7, 93– 106. Google Scholar CrossRef Search ADS   Zeelenberg Marcel, van Dijk Wilco W., Manstead Antony S. R. ( 1998), “Reconsidering the Relation between Regret and Responsibility,” Organizational Behavior and Human Decision Processes , 74 3, 254– 72. Google Scholar CrossRef Search ADS PubMed  Zeelenberg Marcel, van Dijk Wilco W., Manstead Antony S. R., van der Pligt Joop ( 2000), “On Bad Decisions and Disconfirmed Expectancies: The Psychology of Regret and Disappointment,” Cognition and Emotion , 14 4, 521– 41. Google Scholar CrossRef Search ADS   © The Author 2017. Published by Oxford University Press on behalf of Journal of Consumer Research, Inc. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

Journal of Consumer ResearchOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial