When Being in a Positive Mood Increases Choice Deferral

When Being in a Positive Mood Increases Choice Deferral Abstract Consumers’ choices are often accompanied by unrelated incidental moods. The positive mood caused by receiving a compliment, for example, may persist when one is choosing what service to book or which product to buy. How might being in a positive mood affect consumers’ subsequent, unrelated choices? The present research demonstrates that being in a positive mood can make consumers more likely to defer choice. Four studies show that when choosing requires trade-offs between important choice attributes, being in a positive (vs. neutral) mood makes choosing more difficult and therefore increases the likelihood of deferring choice altogether. The findings further understanding of how incidental factors shape choice processes and outcomes and the role of emotions in decision making. choice, choice deferral, choice difficulty, positive mood, trade-offs Consumers’ mood can change in response to various events. Getting stuck in traffic or reading a sad news article, for instance, puts people in a negative mood, whereas receiving a compliment or watching a funny video puts people in a positive mood. These incidental mood states can persist during subsequent, unrelated judgments and decisions (Pham 1998; Schwarz and Clore 1983, 1996; Wyer, Clore, and Isbell 1999). The positive mood caused by watching a funny video, for example, may persist when one is choosing what to make for dinner or which flight to book for an upcoming trip. How might being in a positive mood affect consumers’ subsequent, unrelated choices? In the current research, we propose that being in a positive mood can make consumers more likely to defer choice. Choice options are often described by multiple attributes, and preferences for these attributes tend to conflict. For instance, one available option may offer the best level of one attribute, whereas another may offer the best level of a different attribute (e.g., one flight has the lowest price and another has the fewest number of stops). In order to choose, consumers must make trade-offs between these conflicting attributes (Tversky and Shafir 1992), and such trade-offs can make choice feel more difficult (Chatterjee and Heath 1996; Dhar and Simonson 2003) and choice deferral more likely (Dhar 1996; Dhar and Nowlis 1999; Tversky and Shafir 1992). We argue that being in a positive mood will increase the extent to which consumers focus on important (vs. unimportant) choice attributes. Consequently, when multiple attributes are important and different options are best on each (requiring consumers to make trade-offs between important attributes in order to choose), being in a positive mood should make choosing feel more difficult and increase consumers’ propensity to defer choice altogether. This work makes two main contributions. First, our findings advance understanding of how mood influences decision making. That the emotions caused by choice can impact choice processes and outcomes is well established (Garbarino and Edell 1997; Luce 1998; Luce, Payne and Bettman 1999), yet how incidental mood states—and a positive mood in particular—might shape such outcomes has received less attention. Informing this question, our work demonstrates that being in a positive mood can make subsequent, unrelated choices feel more difficult and therefore increase choice deferral. Further, contributing to the prior work on whether positive mood helps or hinders decision making (Isen and Means 1983; Meloy 2000; Pyone and Isen 2011), our findings suggest that being in a positive mood can help consumers stay focused on the important choice attributes, but proves disadvantageous—increasing choice difficulty and the rate of choice deferral—when those important attributes conflict. Second, our findings contribute to emerging research on the role of incidental factors in choice difficulty and deferral (Novemsky et al. 2007; Xu, Jiang, and Dhar 2013). Going beyond prior work’s emphasis on choice set composition as a driver of choice deferral (Dhar 1996; Tversky and Shafir 1992), we identify a novel incidental factor—a positive mood unrelated to the choice itself—that can increase choice difficulty and deferral. THEORETICAL BACKGROUND The option to defer choice is prevalent in many, if not most, consumer decisions. Consumers defer choice when they postpone or delay selecting an option from among available alternatives (Anderson 2003; Dhar 1996, 1997; Greenleaf and Lehmann 1995). When considering flight options for an upcoming trip, for instance, one could defer choice by deciding to spend more time considering the available options, deciding to seek additional information about the available options, or deciding to explore additional options beyond the current consideration set. One common reason consumers defer choice is because choosing requires trade-offs (Dhar 1997; Dhar and Simonson 2003; Redelmeier and Shafir 1995; Tversky and Shafir 1992). For most consumer choices (e.g., booking a flight, buying a new car, choosing a vacation destination), the choice options are described by multiple attributes (e.g., for a flight, price, number of stops, and on-board internet), with each option having a different level of those attributes. For example, when booking a flight, one flight option might have a lower fare but no on-board internet, whereas another might have a higher fare but offer internet. In order to make a choice, then, consumers must make trade-offs (i.e., give up something of value to obtain something else of value; Bettman et al. 1993; Einhorn and Hogarth 1981) among the various attributes. Choosing the former flight, for instance, would mean sacrificing the ability to stay connected in flight, whereas choosing the latter would require paying more money. We propose that when choosing requires trade-offs between important choice attributes, being in a positive mood will increase consumers’ propensity to defer choice. Our reasoning is based on the notion that a positive mood enables people to be more discerning in their judgments and behavior, with task importance being a key moderator of their responses (Aspinwall 1998, Isen and Reeve 2005). Several distinct lines of research support this view. Examining intrinsic motivation, Isen and Reeve (2005) demonstrate that being in a positive mood increases people’s propensity to prioritize important work tasks (i.e., ones related to higher-order goals) over other enjoyable activities. In the domain of risk-related behaviors, several articles show that being in a positive mood makes people more risk-averse on tasks of high (vs. low) importance (i.e., high vs. low stakes; Isen 1993; Isen and Geva 1987). More germane to the present context, being in a positive mood enhances consumers’ ability to discern important choice attributes from less important ones (Isen 2001). Pyone and Isen (2011), for instance, show that a positive mood affects intertemporal choice (a trade-off between the size of the reward and wait time) by increasing choosers’ focus on the more important aspect of this trade-off (i.e., the size of the reward). Thought protocols revealed that when the choice was important (i.e., the reward was sufficiently valuable), positive-mood participants generated more positive thoughts about reward size than controls, but this effect disappeared when the choice was less important (i.e., the reward was less valuable). Further, Isen and Means (1983) demonstrate that consumers in a positive mood are more likely than controls to eliminate choice alternatives that do not meet set criteria on important dimensions or to ignore unimportant dimensions altogether, an effect that was replicated in a highly consequential setting (i.e., medical decision making; Isen, Rosenzweig, and Young 1991). Building on these findings, we suggest that being in a positive mood will increase the extent to which consumers focus on the important aspects of a decision—namely, the attributes of the choice that are most important. We further argue, however, that this increased focus on important choice attributes can be detrimental—increasing choice difficulty and ultimately choice deferral—when the important attributes conflict (i.e., when different options are best on each, requiring consumers to make trade-offs between them in order to choose). By making salient what one must give up, trade-offs highlight the forgone, underscoring potential losses on each attribute (Brenner, Rottenstreich, and Sood 1999; Carmon and Ariely 2000; Carmon, Wertenbroch, and Zeelenberg 2003). When those potential losses arise from more (vs. less) important attributes, choosing should feel more difficult. For example, if picking a flight required consumers to sacrifice price for fewer stops (both important attributes), this choice would feel more difficult than if choosing instead required consumers to sacrifice price (an important attribute) for on-board entertainment (a relatively unimportant attribute). Because experiencing greater choice difficulty makes choice deferral more likely (Dhar and Nowlis 1999; Tversky and Shafir 1992), when the important attributes of a choice conflict, focusing on them to a greater extent should increase consumers’ propensity to defer choice.1 Consequently, by increasing consumers’ focus on important choice attributes, when multiple attributes are important and different options are best on each (requiring consumers to make trade-offs between important attributes in order to choose), we propose that being in a positive mood will increase choice difficulty, thereby increasing consumers’ propensity to defer choice altogether. H1: When choosing requires trade-offs between important choice attributes, being in a positive mood increases the rate of choice deferral. H2: This increased propensity to defer choice occurs because a positive mood makes choosing between alternatives best on different important attributes more difficult. Further, we argue that the nature of requisite trade-offs (i.e., between important vs. unimportant choice attributes) will moderate these effects. If the need to make trade-offs between important choice attributes plays the critical role we suggest, then when such trade-offs are not required (e.g., choosing only requires trade-offs between unimportant attributes), positive mood’s effects should be attenuated. Consistent with this reasoning, a recent article (Pocheptsova, Peterson, and Etkin 2015) found that in the pursuit of multiple goals, a positive mood influenced subsequent judgments only when those goals were in a high degree of conflict—that is, when consumers’ multiple goals were important and could not easily be prioritized (e.g., goals related to one’s family and work). Although the mechanism (perceived differences between goals) and outcome (evaluation of means to reach goals) in that article differ from the ones examined here, the findings nevertheless suggest that the nature of requisite trade-offs (in the current research, whether those trade-offs are between important choice attributes) may likewise moderate the effects. Consequently, we predict that, whereas a positive mood should increase choice difficulty and deferral when the important choice attributes conflict, when only the unimportant attributes conflict, these effects should be reduced. H3: When choosing requires trade-offs only between unimportant choice attributes, the effect of positive mood on choice difficulty and deferral is attenuated. In summary, we predict that, by increasing the extent to which consumers focus on important (vs. unimportant) choice attributes, when multiple attributes are important and those attributes conflict (requiring consumers to make trade-offs between them in order to choose), being in a positive mood will make choosing feel more difficult, and this increased choice difficulty will increase the likelihood of deferring choice. OVERVIEW OF STUDIES Four studies examine how being in a positive mood affects consumers’ propensity to defer choice. Study 1 provides initial evidence that when multiple choice attributes are important and those attributes conflict (requiring consumers to make trade-offs between them in order to choose), being in a positive mood makes consumers more likely to defer choice. Studies 2–4 replicate this effect in multiple consequential choice settings using different measures of choice deferral and test the proposed underlying role of choice difficulty: when multiple choice attributes are important and choosing requires trade-offs between them, being in a positive mood makes choice feel more difficult, and this increased choice difficulty increases consumers’ propensity to defer choice. In addition to testing for mediation, studies 2 and 3 also explore the proposed underlying process through moderation. Study 2 manipulates the extent to which choosers focus on the important (vs. unimportant) choice attributes: directing (all) participants’ attention to the important choice attributes eliminates positive mood’s effects. Study 3 manipulates whether choosing requires trade-offs between important versus unimportant choice attributes: when the important attributes do not conflict (e.g., a single option is best on all of them), being in a positive mood no longer increases choice difficulty and deferral (hypothesis 3). Lastly, study 4 rules out two important alternative explanations. We shuggest that a positive mood increases choice difficulty and deferral by increasing the extent to which choosers focus on important choice attributes. In contrast to this systematic information processing argument, being in a positive mood could prompt more heuristic information processing (Bless et al. 1990; Schwarz and Bless 1991; Schwarz and Clore 1996), which is also linked to choice deferral (Pocheptsova et al. 2009). Or instead, if making trade-offs between important attributes is emotionally aversive (Einhorn and Hogarth 1981), choosers in positive mood may defer choice in order to avoid resolving those trade-offs (i.e., a mood maintenance-based account; Meloy 2000; Wegener and Petty 1994; Wegener, Petty, and Smith 1995). To show that the systematic processing of important choice attribute information underlies positive mood’s effect, study 4 manipulates choosers’ cognitive capacity: consistent with the proposed theoretical account and inconsistent with alternative explanations, for those in a positive (vs. neutral) mood, imposing cognitive load reduces choice difficulty and therefore decreases the likelihood of choice deferral. Together these studies provide robust and consistent evidence for our predictions: when choosing requires trade-offs between important choice attributes, being in a positive mood increases choice difficulty and choice deferral. STUDY 1: POSITIVE MOOD INCREASES CHOICE DEFERRAL Study 1 tested our first hypothesis. After inducing a positive or neutral mood, we asked participants to choose between two alternatives (flight options) best on different important attributes: one flight option had the most desirable level of one important attribute (price) and the other had the most desirable level of another important attribute (number of stops), requiring participants to make trade-offs between these attributes in order to choose. Notably, in addition to the two alternatives, participants also had the option to defer choice, and we measured the proportion of participants who selected the deferral option. We predicted that in this choice context, where choosing requires trade-offs between important choice attributes, being in a positive mood would increase the propensity to defer choice (hypothesis 1). Design and Method Participants (N = 132; average age 35 years, 49% female) recruited through Amazon’s Mechanical Turk were randomly assigned to a mood condition: positive versus neutral mood. For this and subsequent online studies, a target rule of 50–70 participants per cell determined the sample size. Participants completed two ostensibly unrelated tasks. In the first task, we manipulated incidental mood with a word-association task from prior research (Pyone and Isen 2011). Participants in the positive-mood condition saw 10 positive words (e.g., laughter, fun) and wrote down the first word that came to mind in response to each. Participants in the neutral-mood condition saw 10 neutral words (e.g., hat, chair) and wrote down the first word that came to mind in response to each. Pretest results (N = 51) supported this manipulation. Pretest participants assigned to view the positive words reported more positive feelings on a 10-item five-point PANAS scale (α = .96; Watson, Clark, and Tellegen 1988) than those assigned to view the neutral words (Mpositive = 3.88 vs. Mneutral = 3.43; F(1, 49) = 5.70, p = .021). In the second task, we asked participants to choose a flight for an imagined upcoming trip. Participants viewed two flight options that “fit the time frame of your trip,” each of which was described by four attributes (adapted from Sela and Berger 2012; see web appendix A). We designed the choice set such that each flight option was best on one important attribute but inferior on another. To determine attribute importance, we conducted a pretest. A separate sample of mTurk participants (N = 55, average age 32.9 years, 30.9% women) viewed the same four flight attributes (cost, number of stops, on-board internet, and entertainment) and rated the importance of each (“When choosing a flight, how important are the following attributes to you?”) on a seven-point scale (1 = Very unimportant, 7 = Very important). Examining the means revealed two important attributes and two unimportant attributes: participants rated price (M = 6.69; t(54) = 36.96, p < .001) and number of stops (M = 5.16; t(54) = 5.12, p < .001) as important (above the scale midpoint), and rated on-board internet (M = 3.09; t(54) = –3.21, p = .002) and entertainment (M = 3.00; t(54) = –3.63, p = .001) as unimportant (below the scale midpoint). In the main choice task, we thus considered price and number of stops (on-board internet and entertainment) to be important (unimportant) attributes. We assigned flight option 1 the fewest number of stops and flight option 2 the lowest price, so that choosing required participants to make trade-offs between important choice attributes (see web appendix A). Importantly, in addition to the two flight alternatives, we also gave participants the option to defer choice (in favor of “searching for more options”). We recorded which option (flight option 1, flight option 2, or deferral) they chose.2 Results and Discussion Manipulation Check Among participants who chose one of the two flight options (84.1% of the total sample), mood had no effect on preference between them (χ2(1) = 1.98, p = .159). Overall, 32.4% chose option 1 and 67.6% chose option 2, suggesting that neither option dominated (although option 2, which offered the lowest price, was preferred, consistent with the pretested attribute-importance ratings) and choosing between them required trade-offs between important attributes. Choice Deferral As predicted (hypothesis 1), a positive mood significantly increased the rate of choice deferral (χ2(1) = 6.46, p = .011). Compared to a neutral mood (P = 7.7%), being in a positive mood made participants about three times more likely to defer choice (P = 23.9%). Study 1 provides initial support for our predictions. On a choice task that required trade-offs between important choice attributes, being in a positive (vs. neutral) mood increased the likelihood of deferring choice (hypothesis 1). Despite being incidental to the choice itself, a positive mood can thus carry over to affect consumers’ subsequent, unrelated decisions. STUDY 2: CHOICE DIFFICULTY AND THE ROLE OF ATTRIBUTE FOCUS Study 2 explored the proposed underlying process in two ways. First, we measured choice difficulty and tested for mediation. After inducing a positive or neutral mood, we gave participants a choice between alternatives that were best on different important attributes, as in study 1, and in addition to their propensity to defer choice, we measured how difficult choosing felt. We predicted that being in a positive mood would make choosing feel more difficult, and that this increased difficulty would make participants more likely to defer choice (hypothesis 2). Second, we manipulated the extent to which participants focused on the important choice attributes. If being in a positive mood increases choice difficulty and deferral by increasing the extent to which choosers focus on important choice attributes, as we suggest, then explicitly directing consumers in a neutral mood to focus on the important attributes should attenuate the effects. To test this, for half of participants, we supplemented the choice stimuli with a visual cue that drew attention to the important choice attributes. In the absence of this cue (i.e., the control condition), we expected that being in a positive mood would make choosing feel more difficult and increase choice deferral (hypothesis 1, hypothesis 2). In the presence of this cue, however, we expected that neutral-mood participants would have as much difficulty choosing, and thus be as likely to defer choice, as positive-mood participants in the control condition. In addition, study 2 extended study 1 by examining a consequential choice setting and measuring choice deferral in a different way. We invited lab participants to complete a take-home survey for extra pay, varying the levels of the important survey attributes such that one option was best on one important attribute and the other was best on another. Rather than defer choice in favor of seeking more options, as in study 1, we limited how long participants had to consider the two survey options and, when that time expired, asked them if they wanted more time to think about their choice. In addition to demonstrating the robustness of our choice deferral effect, operationalizing choice deferral in this way casts initial doubt on a potential alternative explanation due to heuristic processing (see study Discussion for rationale). Design and Method Participants (N = 209; 60% female, age not collected) recruited from a university’s behavioral lab were randomly assigned to one condition of a 2 (mood: positive vs. neutral) × 2 (attribute focus: control, enhanced focus) between-subjects design. For this and subsequent lab studies, lab capacity and participant availability determined the sample size. Participants completed two ostensibly unrelated tasks. In the first task, we manipulated incidental mood using the word-association task from study 1. In the second task, we asked participants to choose a take-home survey to complete for extra pay. Participants read, “As part of today’s lab session, you have the opportunity to complete a take-home survey for extra pay. We are collecting data for several surveys, so you can choose which you would like to complete. Students who complete the survey will be entered in a lottery for a cash prize.” Participants viewed two survey options, each of which was described by six attributes (see web appendix A). We designed the choice set such that each survey option was the best on one important attribute but inferior on another. To determine attribute importance, we conducted a pretest. A separate sample of participants (N = 52, 63.5% women) from the same lab pool viewed the same six survey attributes (lottery prize, odds of winning, duration, content, topic, and format) and rated the importance of each (“When choosing a paid take-home survey, how important are the following attributes to you?”) on a seven-point scale (1 = Very unimportant, 7 = Very important). Examining the means revealed two important and four unimportant attributes: participants rated the lottery prize (M = 5.44; t(51) = 6.65, p < .001) and odds of winning (M = 6.15; t(51) = 12.19, p < .001) as important (above the scale midpoint), and rated duration (M = 4.44; t(51) = 1.87, p = .068), content (M = 3.90; t(51) = –.35, p = .729), topic (M = 3.75, t(51) = –.99, p = .328), and format (M = 3.69; t(51) = –1.21, p = .231) as unimportant (equal to or below the scale midpoint). In the main choice task, we thus considered lottery prize and odds of winning (duration, content, topic, and format) to be important (unimportant) attributes. We assigned survey option 1 the highest prize and survey option 2 the best odds, so that choosing required participants to make trade-offs between important choice attributes (see web appendix A). To manipulate participants’ focus on the important choice attributes, we varied the visual depiction of the choice attribute information (see web appendix A). In the enhanced-focus condition, we used bright colors to highlight the important attributes. In the control condition, we provided no such visual cue. A pretest (mTurk panelists, N = 206) supported this manipulation, demonstrating that visually highlighting the important choice attributes increased the extent to which choosers focused on those attributes (and thus trade-offs between them; see web appendix B). Importantly, prior to choosing between the survey options, participants were given the option to defer choice (in favor of spending more time deciding which option they preferred). We told participants that they had 30 seconds to consider the two survey options and then make their choice. At the end of those 30 seconds, we asked them, “Would you like more time to think about your choice?” with response options “Yes, I’d like more time” (i.e., the deferral option) or “No, I will choose now.” We recorded which option participants chose, and those who indicated wanting to choose right away then made their selection (those who indicated wanting more time were told they could make their choice as they left the lab).3,4 Next, on a separate page, we measured choice difficulty using three items (Iyengar and Lepper 2000; see also Xu et al. 2013): “How difficult was it for you to choose the survey that you wanted?” “How frustrated did you feel when making the choice?” and “How hesitant did you feel when making the choice?” (1 = Not at all, 7 = Very much). We combined these items to form a choice-difficulty index (α = .89). Finally, to better understand positive mood’s effects, we collected ancillary measures. We have argued that being in a positive mood makes choosing between alternatives best on different important attributes more difficult because it increases the extent to which choosers focus on the important attributes (which are in conflict). If our reasoning is correct, then positive mood should influence the extent to which participants relied on the important (vs. unimportant) attributes in making their choice: compared to a neutral mood, being in a positive mood should lead the important (unimportant) attributes to play more (less) of a role. To explore this reasoning, we showed participants the six survey attributes and asked them, “How important was each of the following attributes to you in choosing a survey?” (six items, all 1 = Very unimportant, 7 = Very important). In addition, we measured anticipated regret. Anticipated regret can increase choice deferral (Anderson 2003). Because making trade-offs is aversive (Einhorn and Hogarth 1981), and individuals in a positive mood are motivated to maintain that positive state (Clark and Isen 1982; Isen and Levin 1972), one could wonder whether anticipated regret contributes to the present effects. To test this, we asked participants, “How much do you anticipate regretting your choice?” (1 = Not at all, 7 = Very much). Results Manipulation Check A logistic regression of preference on mood, attribute-focus condition, and their interaction revealed no significant effects (ps > .550). Overall, 36.8% of these participants chose option 1 and 63.2% chose option 2, suggesting neither option dominated (although option 2 was preferred) and choosing between them required trade-offs between important attributes. Choice Deferral A logistic regression of choice deferral on mood condition, attribute-focus condition, and their interaction revealed main effects of mood (β = .79, Wald = 3.51, p = .061) and attribute-focus condition (β = .80, Wald = 3.65, p = .056), qualified by the expected interaction (β = –1.02, Wald = 3.07, p = .080; figure 1a). FIGURE 1A View largeDownload slide FOCUSING ON THE IMPORANT ATTRIBUTES MODERATES THE CHOICE DEFERRAL EFFECT FIGURE 1A View largeDownload slide FOCUSING ON THE IMPORANT ATTRIBUTES MODERATES THE CHOICE DEFERRAL EFFECT As predicted and consistent with study 1, in the control condition, a positive mood increased preference for the deferral option (hypothesis 1). When choosing between alternatives best on different important attributes (requiring participants to make trade-offs between those attributes in order to choose), being in a positive mood made participants more likely to defer choice (Ppositive = 43.6% vs. Pneutral = 26.0%; χ2(1) = 3.57, p = .059). In the enhanced-focus condition, however, this effect was attenuated. When the important survey attributes were visually highlighted, the difference between mood conditions in the propensity to defer choice vanished (Ppositive = 38.3% vs. Pneutral = 43.9%; χ2 < 1). Supporting our reasoning, the neutral-mood condition drove this attenuation. Among neutral-mood participants, directing attention to the important choice attributes made deferral more likely (Menhanced-focus = 43.9% vs. Mcontrol = 26.0%; χ2(1) = 3.71, p = .054). However, because positive-mood participants should spontaneously focus on such attributes to a greater extent, we found no effect among them (Menhanced-focus = 38.3% vs. Mcontrol = 43.6%; χ2 < 1). Choice Difficulty A 2 (mood) × 2 (attribute-focus) ANOVA revealed a main effect of mood (Mpositive = 4.29 vs. Mneutral = 3.80; F(1, 205) = 9.90, p = .002), qualified by the expected interaction (F(1, 205) = 4.49, p = .035; figure 1b). There was no main effect of attribute-focus condition (F(1, 205) = 1.98, p = .161). FIGURE 1B View largeDownload slide FOCUSING ON THE IMPORANT ATTRIBUTES MODERATES THE CHOICE DIFFICULTY EFFECT FIGURE 1B View largeDownload slide FOCUSING ON THE IMPORANT ATTRIBUTES MODERATES THE CHOICE DIFFICULTY EFFECT As predicted, in the control condition, a positive mood increased choice difficulty (hypothesis 2). When choosing required participants to make trade-offs between important choice attributes, being in a positive mood made the choice feel more difficult (Mpositive = 4.34 vs. Mneutral = 3.51; F(1, 205) = 13.97, p < .001). In the enhanced-focus condition, however, this effect was attenuated. When the important survey attributes were visually highlighted, there was no difference in choice difficulty across mood conditions (Mpositive = 4.23 vs. Mneutral = 4.06; F < 1). As expected and consistent with the choice deferral results, the neutral-mood condition drove this attenuation. Among neutral-mood participants, directing attention to the important choice attributes made choosing feel more difficult (Menhanced-focus = 4.06 vs. Mcontrol = 3.51; F(1, 205) = 6.37, p = .012), but because positive-mood participants should spontaneously focus on such attributes to a greater extent, we found no effect among them (Menhanced-focus = 4.23 vs. Mcontrol = 4.34; F < 1). Moderated Mediation A bootstrapping moderated mediation analysis (model 7; Hayes 2013) with choice difficulty as the mediator revealed a significant index of moderated mediation (Index = –.68, 95% CI [–1.46 to –.03]). Supporting our reasoning, in the control condition, the indirect effect was positive and significant (ab = .84, 95% CI [.38 to 1.41]): being in a positive mood increased the rate of choice deferral by making choosing feel more difficult. In the enhanced-focus condition, however, the indirect effect was not significant (ab = .16, 95% CI [–.35 to .66]): when the important survey attributes were visually highlighted, being in a positive mood no longer influenced choice difficulty, and thus no longer affected choice deferral. Further supporting our theory, in the neutral-mood condition, the indirect effect of choice difficulty was positive and significant (ab = .55, 95% CI [.11 to 1.10]): directing attention to the important choice attributes made neutral-mood participants more likely to defer choice because it made choosing feel more difficult. In the positive-mood condition, however, the indirect effect was not significant (ab = –.11, 95% CI [–.58 to .35]): among positive-mood participants, making the important attributes more salient did not affect choice difficulty, and thus did not affect choice deferral. Ancillary Analyses First, we analyzed the attribute importance ratings. To explore whether positive mood’s effects are due to an increased focus on the important (vs. unimportant) choice attributes, as we argued, we created an “important attributes” score (the average of the importance ratings for the survey attributes pretested to be important: lottery prize and odds of winning) and an “unimportant attributes” score (the average of the importance ratings for the survey attributes pretested to be unimportant: duration, content, topic, and format) and conducted a 2 (mood) × 2 (attribute-focus) × 2 (attribute-type) repeated-measures ANOVA (see web appendix C for raw scores). This analysis revealed a main effect of the within-subjects attribute-type factor (F(1, 205) = 336.72, p < .001), qualified by a three-way interaction (F(1, 205) = 7.30, p = .007). There were no other main effects (Fs < 1) or interactions (ps > .20). Examining the results within each attribute-focus condition supported our reasoning. In the control condition (i.e., in the absence of a cue to focus on the important attributes), being in a positive mood influenced the attribute importance ratings. In addition to a main effect of the within-subjects attribute-type factor (F(1, 205) = 182.50, p < .001), a mood × attribute-type interaction emerged (F(1, 205) = 7.75, p = .006). While there was no difference across mood conditions in the perceived importance of the “important” survey attributes (Mpositive = 5.89 vs. Mneutral = 5.56; F(1, 205) = 1.89, p = .171), mood did influence the perceived importance of the “unimportant” attributes (F(1, 205) = 6.57, p = .011): compared to neutral-mood participants (M = 3.74), those in a positive mood rated these unimportant attributes as less important (M = 3.12). Consistent with the prior work that finds greater marginal effects on the perceived importance of unimportant (vs. important) choice attributes (Nowlis and Simonson 1996; Sela and Berger 2012), these results suggest that being in a positive mood increases choosers’ relative focus on important (vs. unimportant) choice attributes. In the enhanced-focus condition (i.e., when we visually highlighted the important choice attributes), however, while there was still a main effect of the within-subjects attribute-type factor (F(1, 205) = 154.76, p < .001), there was no longer a mood × attribute-type interaction (F(1, 205) = 1.06, p = .304). Consistent with our theory, when we encouraged neutral-mood participants to focus on the important choice attributes (as those in a positive mood do naturally), mood did not influence the perceived importance of the “important” (Mpositive = 5.54 vs. Mneutral = 5.65; F < 1) or the “unimportant” (Mpositive = 3.62 vs. Mneutral = 3.37; F(1, 205) = 1.03, p = .311) attributes. Second, we examined anticipated regret. A 2 (mood) × 2 (attribute-focus) ANOVA only revealed a main effect of attribute-focus condition (Menhanced-focus = 2.54, Mcontrol = 2.04; F(1, 205) = 5.17, p = .024). We found no main effect of mood (F(1, 205) = 1.18, p = .278) or mood × attribute-focus interaction (F(1, 205) = 1.11, p = .294). Thus, although anticipated regret can contribute to choice deferral (Anderson 2003), in the present context, it does not seem to play a critical role. Discussion Study 2 supports our predictions in a consequential choice setting and demonstrates the proposed underlying process in two ways. First, we measured choice difficulty and showed that this drove the choice deferral effect. As predicted (hypothesis 2), being in a positive mood made choosing feel more difficult, and these increased feelings of choice difficulty explained why a positive mood increased choosers’ propensity to defer choice. Second, we manipulated the extent to which participants focused on the important choice attributes and found that this moderated positive mood’s effects. In the control condition, being in a positive mood increased choice difficulty and the rate of choice deferral (hypothesis 1, hypothesis 2), but when we visually highlighted the important choice attributes, these effects disappeared. Supporting our argument that a positive mood increases consumers’ focus on important choice attributes, explicitly directing neutral-mood participants to focus on the important attributes made them have as much difficulty choosing, and thus be as likely to defer choice, as positive-mood participants in the control condition. Ancillary analyses of the perceived attribute important measures further underscored the role of attribute focus in determining positive mood’s effects. In the control condition, compared to neutral-mood participants, those in a positive mood relied less on the unimportant attributes (and thus relatively more on the important attributes) to make their choice. However, when we encouraged all participants, regardless of their mood state, to focus on the important choice attributes, all participants equally prioritized the important over the unimportant attributes. These results support our reasoning that being in a positive mood makes choosing more difficult and choice deferral more likely by increasing the extent to which consumers focus on important (vs. unimportant) choice attributes. Finally, the results of study 2 cast initial doubt on a potential alternative explanation based on heuristic processing. Although a large body of work documents that a positive mood leads people to process information more systematically (Aspinwall 1998; Erez and Isen 2002; Isen, Daubman, and Nowicki 1987), other findings suggest a positive mood can instead prompt heuristic information processing (Bless et al. 1990; Schwarz and Bless 1991; Schwarz and Clore 1996). Because heuristic processing can also increase choice deferral (at least in some situations; Pocheptsova et al. 2009), one could wonder whether a positive mood makes choice deferral more likely by increasing heuristic, rather than systematic (as we argue; see also Dhar 1997; Dhar and Nowlis 1999), processing of choice attribute information. Casting initial doubt on this possibility, study 2 participants deferred choice in favor of spending more time considering the available options—a delay strategy that is inconsistent with a heuristic-processing account (which would suggest spending less time deciding). Studies 3 and 4 further rule out this account by again measuring the decision to defer choice in favor of spending more time deciding (study 3) and by manipulating cognitive capacity (study 4). STUDY 3: THE MODERATING ROLE OF TRADE-OFF IMPORTANCE Study 3 further examined the underlying mechanism by manipulating whether the important choice attributes were in conflict. If making trade-offs between important choice attributes plays the critical role we suggest, then when such trade-offs are not required (e.g., choosing requires trade-offs only between unimportant attributes), positive mood’s effects should be attenuated (hypothesis 3). To test this, we had half of participants make a choice that required trade-offs between important choice attributes, as in studies 1 and 2, and for the remaining half, we altered the choice attribute values such that choosing required trade-offs only between unimportant attributes. Consistent with the previous results, we expected that when the requisite trade-offs were important, being in a positive mood would make choosing feel more difficult and therefore increase choice deferral (hypothesis 1, hypothesis 2). When the requisite trade-offs were unimportant, however, we expected no such differences across mood conditions (hypothesis 3). In addition, study 3 explored the generalizability of our effects by testing them in a different consequential choice setting. Design and Method Participants (N = 243; average age 21 years, 38% female) recruited from a university’s behavioral lab were randomly assigned to one condition of a 2 (mood: positive vs. neutral) × 2 (trade-off importance: important vs. unimportant) between-subjects design. Participants completed two ostensibly unrelated tasks. In the first task, we manipulated incidental mood using the word-association task from the prior studies. In the second task, we asked participants to choose a portable wireless speaker. Participants read, “As part of today’s study, you will be entered in a lottery for the speaker of your choice (valued at $199). At the conclusion of the study, we will notify the winners to come pick up their speaker.” All participants viewed two available options, each of which was described by seven attributes (see web appendix A). As in the prior studies, in the important trade-offs condition, we designed the choice set such that each speaker option was best on one important attribute but inferior on another. In the unimportant trade-offs condition, we designed the choice set such that one option dominated the other on all important attributes. To determine attribute importance, we ran a pretest. A separate sample of participants (N = 47, average age 23 years, 49% women) from a university’s behavioral lab viewed the same seven speaker attributes (battery life, Bluetooth, power supply, weight, color, range, micro-USB ports) and rated the importance of each (“When choosing a portable wireless speaker, how important are the following attributes to you?”) on a seven-point scale (1 = Very unimportant, 7 = Very important). Examining the means identified four important and three unimportant attributes: participants rated battery life (M = 5.85; t(46) = 9.08, p < .001), range (M = 5.77; t(46) = 6.98, p < .001), power supply (M = 4.91; t(46) = 3.53, p = .001), and Bluetooth (M = 5.06; t(46) = 4.39, p < .001) as important (above the scale midpoint), and rated weight (M = 3.53; t(46) = –2.01, p = .051), color (M = 3.85; t < 1), and micro-USB ports (M = 3.79; t < 1) as unimportant (equal to or below the scale midpoint). In the main experiment, we thus considered battery life, range, power supply, and Bluetooth (weight, color, and micro-USB ports) to be important (unimportant) attributes. To manipulate trade-off importance, we varied the attribute levels of the important speaker attributes. In the important trade-offs condition, we assigned speaker option 1 (Bose) the furthest range and speaker option 2 (Jawbone) the longest battery life (and the same power supply and Bluetooth levels to both), so that choosing required trade-offs between important choice attributes. In the unimportant trade-offs condition, we assigned speaker option 2 (Jawbone) the best level of both range and battery life (holding the other important attributes constant), so that choosing required trade-offs only between the unimportant attributes (weight, color, and micro-USB ports; see web appendix A). Importantly, prior to choosing between the speaker options, participants were given the option to defer choice (in favor of spending more time deciding which option they preferred). As in study 2, we asked them, “Would you like more time to think about your choice?” with response options “Yes, I’d like more time” (i.e., the deferral option) or “No, I will choose now.” We recorded which option participants chose, and those who indicated wanting to choose right away then made their selection (those who indicated wanting more time were told they could make their choice at the end of the study).5,6 Finally, on a separate page, we measured choice difficulty using the same three items from study 2 (1 = Not at all, 7 = Very much): “How difficult was it for you to choose the speaker that you wanted?” “How frustrated did you feel when making the choice?” and “How hesitant did you feel when making the choice?” combined to form a choice-difficulty index (α = .83). Results Manipulation Check A logistic regression of preference on mood, trade-off importance condition, and their interaction revealed only the expected main effect of trade-off importance (Wald = 16.94, p < .001). There was no main effect of mood or interaction (mood: Wald < 1; interaction: Wald = 1.58, p = .210). Supporting our manipulation, although neither option dominated in the important trade-offs condition (51.6% chose option 1 and 48.4% chose option 2), option 2 (which was best on the important attributes) dominated option 1 in the unimportant trade-offs condition (20.2% chose option 1 and 79.8% chose option 2; χ2(1) = 23.19, p < .001). Choice Deferral A logistic regression of choice deferral on mood condition, trade-off importance condition, and their interaction revealed only the expected interaction (β = 1.51, Wald = 3.23, p = .072; figure 2a). There were no main effects of mood or trade-off importance condition (Wald’s < 1). As predicted and consistent with studies 1 and 2, in the important trade-offs condition, a positive mood increased preference for the deferral option (hypothesis 1). When choosing required trade-offs between important choice attributes, being in a positive mood made participants more likely to defer choice (Ppositive = 25.4% vs. Pneutral = 8.9%; χ2(1) = 5.44, p = .020). In the unimportant trade-offs condition, however, this effect was eliminated (hypothesis 3). When choosing required trade-offs only between unimportant attributes, being in a positive mood no longer increased the rate of choice deferral (Ppositive = 7.6% vs. Pneutral = 9.7%; χ2 < 1). FIGURE 2A View largeDownload slide TRADE-OFF IMPORTANCE MODERATES THE CHOICE DEFERRAL EFFECT FIGURE 2A View largeDownload slide TRADE-OFF IMPORTANCE MODERATES THE CHOICE DEFERRAL EFFECT Choice Difficulty A 2 (mood) × 2 (trade-off importance) ANOVA on choice difficulty revealed a main effect of mood (Mpositive = 3.86 vs. Mneutral = 3.44; F(1, 239) = 7.87, p = .005) and a main effect of trade-off importance condition (F(1, 239) = 7.29, p = .007). As could be expected, participants had more difficulty choosing when doing so required trade-offs between important (vs. unimportant) attributes (Mimportant = 3.88 vs. Munimportant = 3.45). Importantly, these main effects were qualified by the predicted interaction (F(1, 239) = 5.85, p = .016; figure 2b). A expected and consistent with study 2, when choosing required trade-offs between important choice attributes, being in a positive mood made the speaker choice feel more difficult (Mpositive = 4.28 vs. Mneutral = 3.46; F(1, 239) = 12.95, p < .001). When choosing required trade-offs only between unimportant attributes, however, this effect was attenuated (Mpositive = 3.47 vs. Mneutral = 3.41; F < 1). FIGURE 2B View largeDownload slide TRADE-OFF IMPORTANCE MODERATES THE CHOICE DIFFICULTY EFFECT FIGURE 2B View largeDownload slide TRADE-OFF IMPORTANCE MODERATES THE CHOICE DIFFICULTY EFFECT Moderated Mediation A bootstrapping moderated mediation analysis (model 7; Hayes 2013) with choice difficulty as the mediator revealed a significant index of moderated mediation (Index = .87, 95% CI [.16 to 1.81]). As predicted and consistent with study 2, in the important trade-offs condition, the indirect effect was positive and significant (ab = .94, 95% CI [.38 to 1.75]): being in a positive mood increased the likelihood of deferring choice by making choosing feel more difficult. In the unimportant trade-offs condition, however, the indirect effect was not significant (ab = .07, 95% CI [–.45 to .67]): when choosing required trade-offs only between unimportant attributes, being in a positive mood no longer influenced choice difficulty, and thus no longer affected choice deferral. Discussion Study 3 supports our predictions in a different consequential choice setting and further demonstrates the proposed underlying process in two ways. First, as in study 2, we measured choice difficulty and demonstrated that this drove the effect of positive mood on choice deferral (hypothesis 2). Second, we manipulated whether choosing required trade-offs between important choice attributes and showed this moderated positive mood’s effects (hypothesis 3). Consistent with the prior results, when choosing required trade-offs between important attributes, being in a positive mood increased choice difficulty, and consequently, choice deferral (hypothesis 1, hypothesis 2). When choosing required trade-offs only between unimportant attributes, however, being in a positive mood no longer increased choice difficulty, and consequently, did not affect the propensity to defer choice (hypothesis 3). In line with prior research showing that the nature of requisite trade-offs moderates positive mood’s effects (e.g., Pocheptsova et al. 2015), being in a positive mood increased choice deferral only when choosing required trade-offs between important choice attributes. Note, this moderation occurred even though participants in both trade-off importance conditions had to make trade-offs in order to choose (although these trade-offs were presumably easier in the unimportant trade-offs condition). Thus, the need to make trade-offs alone is not what leads a positive mood to increase choice difficulty and choice deferral. Consistent with prior work showing that a positive mood is discerning and helps people disregard unimportant information (Isen and Means 1983; Pyone and Isen 2011), being in a positive mood increases choice difficulty (and therefore choice deferral) only when choosing requires sacrificing one important attribute for another. The results of study 3 also further argue against a potential alternative explanation based on heuristic processing. As in study 2, study 3 participants deferred choice in favor of spending more time considering the available options—a delay strategy consistent with the proposed systematic-processing account and inconsistent with an alternative one based on heuristic processing. Study 4, reported next, further rules out this potential alternative explanation. STUDY 4: RULING OUT ALTERNATIVE EXPLANATIONS Our final study ruled out two potential alternative explanations for why a positive mood increases choice deferral by manipulating choosers’ cognitive capacity. Reducing cognitive capacity (via cognitive load) impedes systematic information processing (Gilbert, Giesler, and Morris 1995; Shiv and Huber 2000). Thus, if positive mood’s effects are due to the more systematic processing of choice attribute information, as we argue, then cognitive load should primarily influence individuals in a positive mood. Specifically, because they can no longer elaborate on trade-offs between important choice attributes, positive-mood participants under cognitive load should find choosing less difficult and be less likely to defer choice. If positive mood’s effects are instead due to more heuristic information processing, however, then imposing cognitive load should have no such effects. Similarly, if positive mood’s effects are due to a desire to maintain one’s positive mood by avoiding resolving difficult trade-offs (i.e., a mood-maintenance-based account; Meloy 2000; Wegener and Petty 1994; Wegener et al. 1995), then cognitive load should have no impact on positive-mood participants’ propensity to defer choice. Finding that cognitive load reduces choice difficulty and deferral for positive-mood participants would thus support our theory and cast strong doubt on both of these alternative accounts. In addition, study 4 underscored the robustness and external validity of our effects by testing them in a different consequential choice setting and measuring choice deferral in a different way (i.e., deferring in favor of learning more information about the available options). Design and Method Participants (N = 320; average age 21 years, 64% female) recruited from a university’s behavioral lab were randomly assigned to one condition of a 2 (mood: positive vs. neutral) × 2 (cognitive load: control vs. load) between-subjects design. Eight individuals reported audio problems and were excluded, leaving a sample of 312. Participants completed two ostensibly unrelated tasks. In the first task, we manipulated incidental mood in a different way. Participants viewed one of two video clips pretested to elicit the target mood state (positive or neutral, depending on condition). Supporting the manipulation, pretest participants (N = 94) assigned to view the positive-mood clip reported more positive feelings on a 10-item seven-point PANAS scale (α = .97) than those assigned to view the neutral-mood clip (Mpositive = 5.75 vs. Mnetural = 4.34; F(1, 92) = 41.33, p < .001). In the second task, we asked participants to choose a single-serving protein powder packet (see web appendix A). We told all participants that as part of today’s study, they would receive their preferred alternative at the end of the session. Prior to showing participants the available alternatives, we manipulated cognitive load. In the cognitive-load condition, we asked participants to remember an eight-digit number for the remainder of the study. Remembering a multidigit number has reliably been shown to reduce individuals’ cognitive capacity and impede their ability to systematically process information (Gilbert et al. 1995; Shiv and Huber 2000). We told participants the purpose of remembering the number was to help us learn about the decision-making process, and that we would ask them to recall it toward the end of the survey. In the control (no-load) condition, participants proceeded directly to the choice task. Participants viewed three protein powder options, each of which was described by five attributes (see web appendix A), which we chose based on typical nutritional information provided on the packaging of protein supplements (calories, size, protein, sugar, and flavor). Unlike in the previous studies, here we explicitly told participants which attributes were important (protein and sugar content). Participants read, “Industry experts suggest that the protein and sugar content are the most important attributes when evaluating different protein powders.” Importantly, prior to choosing between the protein powder options, participants were given the option to defer choice (in favor of learning more about the available alternatives). Rather than provide complete information about both important attributes, as in the prior studies, we revealed the protein content of each powder option but hid the sugar content on two of them (see web appendix A). Thus, based on the information provided, participants could see that option 3 (Designer Whey) had the least amount of protein and option 2 (Biochem) had the most, but they were unable to discern how the alternatives differed on sugar content. This meant that participants could choose based on one important attribute (protein), but lacked information to choose based on both. We told participants that before making their choice, they had the option to acquire more information about the options, and specifically, that “the cells marked // (i.e., the missing sugar values) in the table will populate.” Participants further read that this additional information could take a few minutes to load and they would have to wait. Then we asked them, “Would you like more information before making your choice?” with response options “Yes, I’d like more information” (i.e., the deferral option) or “No, I will choose now.” We recorded which option participants chose, and those who indicated wanting to choose right away then made their selection (those who indicated wanting more information were told they could see the complete table and make their choice at the end of the study).7,8 Finally, on a separate page, we removed the cognitive load from participants in that condition by asking them to type the number we had asked them to remember. Then, we measured choice difficulty using the same three items from the prior studies (1 = Not at all, 7 = Very much): “How difficult was it for you to choose the protein powder that you wanted?” “How frustrated did you feel when making the choice?” and “How hesitant did you feel when making the choice?” which we combined to form a choice-difficulty index (α = .77). Results Manipulation Check A series of logistic regressions (each comparing preference between two of the three alternatives) on mood, cognitive-load condition, and their interaction revealed no significant effects (ps > .20). Overall, 57% of participants chose option 1 (Quest), 21% chose option 2 (Biochem), and 22% chose option 3 (Designer), suggesting that no option dominated (although option 1, which had the most protein, was preferred), and choosing between them required trade-offs between important choice attributes. Choice Deferral A logistic regression of choice deferral on mood condition, cognitive-load condition, and their interaction revealed a significant main effect of mood (β = .78, Wald = 5.67, p = .017), qualified by the predicted interaction (β = –.94, Wald = 4.18, p = .041; figure 3a). There was no main effect of cognitive load (Wald < 1). As predicted and consistent with the previous studies, in the absence of cognitive load, being in a positive mood made participants more likely to defer choice (Ppositive = 63.3% vs. Pneutral = 44.2%; χ2(1) = 5.75, p = .017). Under cognitive load, however, this effect was eliminated. When we imposed a cognitive load prior to considering the choice options, the propensity to defer choice no longer differed across mood conditions (Ppositive = 44.1% vs. Pneutral = 48.1%; χ2 < 1). FIGURE 3A View largeDownload slide COGNITIVE LOAD MODERATES THE CHOICE DEFERRAL EFFECT FIGURE 3A View largeDownload slide COGNITIVE LOAD MODERATES THE CHOICE DEFERRAL EFFECT Supporting our reasoning and ruling out potential alternative accounts, the positive-mood condition drove this attenuation. Putting positive-mood participants under cognitive load decreased their rate of choice deferral to the level of neutral-mood participants (Pload = 44.1% vs. Pcontrol = 63.3%; χ2(1) = 5.75, p = .017), but did not alter neutral-mood participants’ propensity to defer choice (Pload = 48.1% vs. Pcontrol = 44.2%; χ2 < 1). This suggests that positive mood’s effects are due to the more systematic processing of choice attribute information. Choice Difficulty A 2 (mood) × 2 (cognitive load) ANOVA on choice difficulty revealed only the predicted interaction (F(1, 308) = 4.04, p = .045; figure 3b). No main effects of mood (F(1, 308) = 1.74, p = .188) or cognitive-load condition (F < 1) emerged. As predicted and consistent with studies 2 and 3, in the absence of cognitive load, being in a positive mood made choosing feel more difficult (Mpositive = 4.01 vs. Mneutral = 3.53; F(1, 308) = 5.55, p = .019). Under cognitive load, however, this effect was eliminated. When we imposed a cognitive load prior to considering the choice options, choice difficulty no longer differed across mood conditions (Mpositive = 3.60 vs. Mneutral = 3.70; F < 1). FIGURE 3B View largeDownload slide COGNITIVE LOAD MODERATES THE CHOICE DIFFICULTY EFFECT FIGURE 3B View largeDownload slide COGNITIVE LOAD MODERATES THE CHOICE DIFFICULTY EFFECT Consistent with the choice deferral results and supporting our theory, the positive-mood condition drove this attenuation. Putting positive-mood participants under cognitive load made choosing feel less difficult (Mload = 3.60 vs. Mcontrol = 4.01; F(1, 308) = 3.96, p = .047), but did not affect choice difficulty among neutral-mood participants (Mload = 3.70 vs. Mcontrol = 3.53; F < 1). This further suggests that positive mood’s effects cannot be explained by alternative accounts due to heuristic processing or mood maintenance. Moderated Mediation As in studies 2 and 3, we ran a bootstrapping moderated mediation analysis (model 7; Hayes 2013) with choice difficulty as the mediator to test the proposed underlying process. This analysis revealed a significant index of moderated mediation (Index = –.27, 95% CI [–.63 to –.02]). Supporting our theory, in the control condition, the indirect effect was positive and significant (ab = .22, 95% CI [.05 to .49]): being in a positive mood increased the rate of choice deferral by making choosing feel more difficult. In the cognitive-load condition, however, the indirect effect was not significant (ab = –.05, 95% CI [–.26 to .14]): when participants had fewer cognitive resources to process the choice options, being in a positive mood no longer made choosing feel more difficult, and thus no longer increased consumers’ propensity to defer choice. Further supporting our reasoning, in the positive-mood condition, the indirect effect of choice difficulty was negative and significant (ab = –.19, 95% CI [–.43 to –.01]): adding a cognitive load made positive-mood participants less likely to defer choice because it made choosing less difficult. In the neutral-mood condition, however, the indirect effect was not significant (ab = .05, 95% CI [–.46 to .61]): among neutral-mood participants, adding a cognitive load did not influence choice difficulty, and thus did not affect choice deferral. Discussion Study 4 underscores our theorizing in a different consequential choice domain and rules out a key alternative explanation. We manipulated participants’ cognitive resources prior to considering the choice options and showed that this moderated positive mood’s effects. Consistent with the prior results, in the control condition, being in a positive mood increased choice difficulty and the rate of choice deferral (hypothesis 1, hypothesis 2). Under cognitive load, however, these effects disappeared. Supporting our argument that a positive mood increases the extent to which consumers focus on important choice attributes (and thus engage in trade-offs between them), which requires cognitive resources (Montgomery 1983; Pocheptsova et al. 2009), reducing positive-mood participants’ cognitive capacity attenuated the effects on choice difficulty and choice deferral. The nature of the moderation further rules out potential alternative explanations. Imposing cognitive load caused positive-mood participants to find choosing less difficult and therefore be less likely to defer choice, similar to neutral-mood participants in the control condition. That cognitive load influenced choice difficulty and deferral among positive-mood, rather than neutral-mood, participants is inconsistent with arguments based on heuristic processing or mood maintenance, underscoring that positive mood’s effects are due to the proposed systematic processing of choice attribute information. That being in a positive mood made participants more likely to defer choice in favor of acquiring more information about the important attributes is also inconsistent with both alternative accounts. GENERAL DISCUSSION Incidental moods accompany most, if not all, of consumers’ decisions. While a few prior articles have looked at the effect of a negative mood on decision making (Hammond and Doyle 1991; Lewinsohn and Mano 1993), how a positive mood might impact subsequent, unrelated choices remains largely unexplored (for exceptions, see Isen and Means 1983; Meloy 2000; Meloy, Russo, and Miller 2006). Our work contributes to this line of research by examining how being in a positive mood shapes consumers’ propensity to defer choice. Results of four studies demonstrate that when multiple choice attributes are important and those attributes conflict (requiring trade-offs between them in order to choose), being in a positive mood makes consumers more likely to defer choice (hypothesis 1). We replicated this effect across several choice contexts, both hypothetical (study 1) and consequential (studies 2–4), and with multiple measures of choice deferral: participants in a positive mood were more likely to defer choice in favor of seeking additional alternatives (study 1), having more time to decide (studies 2–3), and learning more information about the available choice options (study 4). Further, the choice deferral effect was robust to a range of important attribute conflicts, including approach/approach (e.g., in study 3, battery life vs. range), approach/avoidance (e.g., in study 4, protein vs. sugar), and avoidance/avoidance (e.g., in study 2, price vs. the number of stops). The studies also provide insight into the underlying process. When multiple choice attributes are important and those important attributes conflict, being in a positive mood makes choosing feel more difficult, which increases consumers’ propensity to defer choice (hypothesis 2, studies 2–4). Based on the prior work that shows a positive mood enables consumers to discern important from unimportant information (Isen and Means 1983; Pyone and Isen 2011), we argue that this occurs because being in a positive mood increases the extent to which consumers focus on important choice attributes. In support, an ancillary analysis in study 2 showed that compared to neutral-mood participants, those in a positive mood relied less on the unimportant attributes (and thus relatively more on the important attributes) to make their choice. Consequently, when choosing required trade-offs only between unimportant attributes (hypothesis 3, study 3), being in a positive mood no longer increased choice difficulty, thereby attenuating its effect on choice deferral. The findings also cast doubt on potential alternative explanations. While one could wonder whether a positive mood increases choice difficulty and deferral by prompting heuristic processing, study 4 rules out this possibility. Because reducing cognitive capacity impedes systematic information processing (Gilbert et al. 1995; Shiv and Huber 2000), that imposing a cognitive load reduced choice difficulty and deferral among positive-mood participants, but had no such effects among neutral-mood participants, demonstrates that positive mood’s effects are due to systematic (vs. heuristic) information processing. That positive-mood participants deferred choice in favor of spending more time deciding (studies 2–3) and learning additional information about the available options (study 4) casts further doubt on this alternative account. In addition, while one could wonder whether being in a positive mood increases choice deferral due to a mood-maintenance-based alternative explanation (because making trade-offs is aversive and people are motivated to maintain positive mood states), our findings cast doubt on this possibility in two ways. First, the moderation by cognitive load in study 4 suggests that it is the extent to which consumers process trade-offs between important attributes—not their motivation to protect a positive mood state—that underlies the present effects. Second, that we found no effect of mood on anticipated regret (study 2) suggests that the motivation to maintain one’s positive mood is not the primary underlying driver. Notably, this null effect also casts doubt on the possibility that positive mood’s effects are due to the increased desirability of important choice attributes (which could increase the anticipated regret associated with forgoing one attribute in favor of another). Theoretical Contributions This research advances understanding of how mood influences decision making. That the emotions experienced during a choice can impact the process and outcomes of that choice is well established (Garbarino and Edell 1997; Luce 1998; Luce, Bettman, and Payne 1997; Luce et al. 1999). Yet while such “integral” mood states have received much attention in the choice literature, less work has explored whether incidental mood states—and a positive mood in particular—might also influence choice processes and outcomes. Our research suggests a novel way that being in a positive mood influences subsequent, unrelated decisions: by increasing the extent to which choosers focus on important (vs. unimportant) choice attributes. When consumers choose between alternatives that are best on different important attributes (requiring trade-offs between them in order to choose), being in a positive mood makes choosing feel more difficult and increases consumers’ propensity to defer choice. This work also informs how mood interacts with choice attribute conflict to impact choice deferral. Separate streams of research have identified two reasons why the need to make trade-offs between important choice attributes can increase consumers’ propensity to defer choice. On one hand, making trade-offs between important attributes increases preference uncertainty, which can make selecting one from among many options more difficult (Dhar 1996, 1997; Dhar and Nowlis 1999; Dhar and Sherman 1996; Dhar and Simonson 2003; Redelmeier and Shafir 1995; Tversky and Shafir 1992). On the other hand, making trade-offs between important attributes is emotionally aversive, which can lead consumers to avoid such trade-offs (Luce 1998; Luce et al. 1997; Luce et al. 1999). Consistent with the former perspective, the current findings show that a positive mood makes consumers more likely to defer choice by increasing the extent to which they focus on important choice attributes (and thus engage in trade-offs between them). The current research also relates to whether being in a positive mood results in better or worse decisions. Prior research provides mixed results on this point. Whereas some work finds that a positive mood improves decision making by helping consumers process information more efficiently (ignoring unimportant information and focusing on what is important; Isen and Means 1983), other work shows that a positive mood results in suboptimal decisions by leading consumers to distort information about the choice options (Meloy 2000; Meloy et al. 2006). Our findings fall in between: being in a positive mood can help consumers stay focused on the important choice attributes, which should generally result in better choice outcomes (Sela, Berger, and Nardini 2017), but proves disadvantageous—increasing choice difficulty and the rate of choice deferral—when those important attributes conflict. Finally, this research contributes to emerging work on the role of incidental factors in choice difficulty and deferral. Prior work on the antecedents of choice difficulty and deferral has primarily focused on structural aspects of the choice set. For example, choosing from a larger number of options (Brenner et al.; Carmon et al. 2003; Iyengar and Lepper 2000; Sela, Berger, and Liu 2009; Shugan 1980), from more varied assortments (Chernev 2006; Huffman and Kahn 1998; Townsend and Kahn 2014), and from choice sets that lack acceptable alternatives (Ratchford 1982; Stigler 1961; Weitzman 1979) can make choosing more difficult and choice deferral more likely. More recently, marketing scholars have begun to explore how incidental factors unrelated to the focal choice might also play a role. For example, processing disfluency and metacognitive difficulty (Novemsky et al. 2007; Schrift et al. 2011; Sela and Berger 2012) and mental abstraction (Kim, Khan, and Dhar 2013; Xu et al. 2013) can impact choice difficulty and deferral. Adding to these findings, we demonstrate that an incidental positive mood can influence consumers’ feelings of choice difficulty and subsequent propensity to defer choice altogether. Implications and Future Research Directions This research has several practical implications. From the consumer’s perspective, the findings further understanding of why choosing might feel difficult. Consumers often find choice be difficult, but may be unaware of the reason(s) why. Unlike more readily observable aspects of choice, such as the appeal of individual options or the size of the choice set, incidental influences on choice difficulty may be harder to diagnose, and thus harder to correct. By identifying a novel factor that impacts choice difficulty (and the likelihood of choice deferral)—being in a positive mood—this research better enables consumers to understand their feelings of choice difficulty. From the marketer’s perspective, our findings have implications for the use of positive mood appeals as a persuasion tactic. Marketers often craft content intended to influence consumers’ mood states (Aaker, Stayman, and Hagerty 1986; Burke and Edell 1989; Goldberg and Gorn 1987) and assume that evoking a positive mood will have desirable effects. While putting consumers in a good mood may indeed result in more favorable evaluations of individual products (Barone, Miniard, and Romeo 2000; Dommermuth and Milliard 1967; Gorn, Goldberg, and Basu 1993), when multiple choice attributes are important and those attributes conflict, it may also inadvertently make choosing between options feel more difficult and choice deferral more likely. Positive mood-boosting tactics may therefore be more effectively employed in situations where consumers are evaluating a single product in isolation, as opposed to comparing multiple products and making trade-offs between them. Similarly, a positive mood may be advantageous when marketers want to encourage consumers to select multiple options from a choice set. While in our studies we did not allow participants to acquire several options at the same time, it is possible that instead of deferring choice, consumers would resolve conflict between important choice attributes by purchasing multiple options from the choice set. Such an effect would provide another theoretical explanation for Kahn and Isen (1993)’s finding that being in a positive mood increases variety seeking among attractive options and would be consistent with the prior work that demonstrates increased variety seeking when more options in the choice set have favorable evaluations (Faraji-Rad, Moeini-Jazani, and Warlop 2013; Goukens et al. 2007). The findings also suggest interesting directions for future research. First, might the desirability versus feasibility of choice option attributes influence positive mood’s effects? Prior work has distinguished between low-level (feasibility) attributes and high-level (desirability) attributes and proposed that construing choice at a higher level should lead consumers to prefer desirability attributes over feasibility ones (Liberman and Trope 1998). In study 1, one could argue that one important choice attribute is high on feasibility (price) and the other on desirability (number of stops), yet we do not find that a positive mood increases preference for the shorter flight option. Dhar and Simonson (1999) suggest one way to reconcile our findings. They show that a trade-off between money and another important choice attribute can be perceived as a trade-off between a resource and a goal or between two goals (with the former leading to prioritization and the latter leading to more balancing). It is possible that in study 1, participants saw money as a goal rather than a resource, leading them to engage in trade-offs between price and number of stops rather than prioritize the number of stops over price. Future work could explore when a positive mood increases the extent to which consumers focus on certain types of attributes (e.g., desirability vs. feasibility), rather than distinguish between important and unimportant attributes more generally. Second, under what situations might being in a positive mood lead consumers to prioritize among important choice attributes (and choose based on just one)? In this work we argued that a positive mood increases consumers’ relative focus on important (vs. unimportant) attributes. In support, in study 2, positive-mood participants perceived the unimportant attributes to be less important than neutral-mood participants; along similar lines, Isen and Meads (1993) found that positive-mood participants more effectively distinguished between multiple more and less important attributes (rather than focus on just one). Thus, empirically, being in a positive mood does not seem to spontaneously lead to such absolute attribute prioritization. Pocheptsova et al. (2015), however, suggests that when important goals serve a single overarching purpose (such as goals to exercise and eat healthy), being in a positive mood may possibly encourage such prioritization. Future research could examine whether highlighting the interconnectedness of important choice attributes would likewise affect their prioritization. Conclusion Consumers’ mood changes in response to various events and such incidental mood states can persist during subsequent, unrelated decisions. Four studies explored how being in a positive mood impacts choice deferral. Results showed that when multiple attributes are important and choosing requires trade-offs between them, being in a positive mood makes choice feel more difficult, which increases consumers’ propensity to defer choice. In addition to structural aspects of choice sets and other contextual factors the incidental moods that accompany consumers’ choices can therefore impact how likely they are to defer choice altogether. DATA COLLECTION INFORMATION The second author supervised data collection by the lab manager at the University of South Carolina behavioral lab (studies 2, 3, and 4). The first author collected the data for study 1 on Amazon’s Mechanical Turk. Study 1 ran in fall 2016. Study 2 ran in spring 2017. Study 3 ran in fall 2015. Study 4 ran in spring 2016. The first author analyzed the data, with input and advice from the second author. Supplementary materials are included in the web appendix accompanying the online version of this article. Footnotes 1 Although disregarding unimportant attributes can in some cases make choosing easier (e.g., by reducing the number of attributes under consideration; Luce et al. 1999; Schrift, Netzer, and Kivetz 2011; Sela and Berger 2012; Tversky and Shafir 1992), when multiple attributes are important and different options are best on each, focusing on the important attributes should exacerbate choice difficulty. 2 Time spent choosing did not differ between mood conditions (Mpositive = 27.27 seconds, Mneutral = 34.33 seconds; F(1, 130) = 1.02, p = .315). Note, here and in subsequent studies, we did not have an a priori prediction for how a positive mood might impact decision time: whereas our prediction that being in a positive mood increases choice difficulty might suggest positive- (vs. neutral-) mood participants spend more time deciding, the prior work that finds being in a positive mood leads people to process choice information more efficiently (Isen and Means 1983) would suggest that positive- (vs. neutral-) mood participants instead spend less time deciding. Consequently, we did not have an expectation for how (or whether) decision time differs across mood conditions. 3 Time spent choosing was capped in this study (at 30 seconds). 4 We drew lottery winners, consistent with the stated odds, from among participants who returned the take-home survey and arranged for those individuals to receive their prize. 5 A 2 (mood) × 2 (trade-off importance) ANOVA on time spent choosing revealed only a main effect of trade-off importance condition (F(1, 239) = 19.87, p < .001). As might be expected, participants spent more time deciding when the trade-offs were between important (vs. unimportant) attributes (Mhigh-importance = 34.09 seconds vs. Mlow-importance = 27.09 seconds). There was no effect of mood (Mpositive = 29.74 seconds, Mneutral = 31.10 seconds, F < 1) or interaction (F < 1). 6 We drew lottery winners (one from each experimental session, approximately a one in 25 chance of winning) and arranged for those individuals to receive their speaker of choice. 7 A 2 (mood) × 2 (cognitive load) ANOVA on time spent choosing revealed a significant main effect of cognitive load (F(1, 308) = 55.23, p < .001). Consistent with the notion that cognitive load reduces information processing capacity, participants spent less time choosing when under cognitive load (Mload = 34.45 seconds vs. Mcontrol = 51.08 seconds). We found no main effect of mood (F < 1), but did find an interaction (F(1, 308) = 4.21, p = .041), such that imposing a cognitive load had a bigger effect on decision time in the positive- (Mload = 32.38 seconds vs. Mcontrol = 53.59 seconds; F(1, 308) = 44.98, p < .001) versus neutral-mood condition (Mload = 36.47 seconds vs. Mcontrol = 48.50 seconds; F(1, 308) = 14.47, p < .001). 8 All participants received their choice of protein powder as they exited the lab. REFERENCES Aaker David A. , Stayman Douglas M. , Hagerty Michael R. ( 1986 ), “Warmth in Advertising: Measurement, Impact, and Sequence Effects,” Journal of Consumer Research , 12 4 , 365 – 81 . Google Scholar CrossRef Search ADS Anderson Christopher J. ( 2003 ), “The Psychology of Doing Nothing: Forms of Decision Avoidance Result from Reason and Emotion,” Psychological Bulletin , 129 ( January ), 139 – 67 . Google Scholar CrossRef Search ADS Aspinwall Lisa G. ( 1998 ), “Rethinking the Role of Positive Affect in Self-Regulation,” Motivation and Emotion , 22 ( March ), 35 – 68 . Barone Michael J. , Miniard Paul W. , Romeo Jean B. ( 2000 ), “The Influence of Positive Mood on Brand Extension Evaluations,” Journal of Consumer Research , 26 4 , 386 – 400 . Google Scholar CrossRef Search ADS Bettman James R. , Johnson Eric J. , Luce Mary Frances , Payne John W. ( 1993 ), “Correlation, Conflict, and Choice,” Journal of Experimental Psychology: Learning, Memory, and Cognition , 19 ( July ), 931 – 51 . Google Scholar CrossRef Search ADS Bless Herbert , Bohner Gerd , Schwarz Norbert , Strack Fritz ( 1990 ), “Mood and Persuasion: A Cognitive Response Model,” Personality and Social Psychology Bulletin , 16 ( June ), 331 – 45 . Google Scholar CrossRef Search ADS Brenner Lyle , Rottenstreich Yuval , Sood Sanjay ( 1999 ), “Comparison, Grouping, and Preference,” Psychological Science , 10 3 , 225 – 9 . Google Scholar CrossRef Search ADS Burke Marian C. , Edell Julie ( 1989 ), “The Impact of Feelings on Ad-Based Affect and Cognition,” Journal of Marketing Research , 26 ( February ), 69 – 83 . Google Scholar CrossRef Search ADS Carmon Ziv , Ariely Dan ( 2000 ), “Focusing on the Forgone: How Value Can Appear So Different to Buyers and Sellers,” Journal of Consumer Research , 27 3 , 360 – 70 . Google Scholar CrossRef Search ADS Carmon Ziv , Wertenbroch Klaus , Zeelenberg Marcel ( 2003 ), “Option Attachment: When Deliberating Makes Choosing Feel Like Losing,” Journal of Consumer Research , 30 1 , 15 – 29 . Google Scholar CrossRef Search ADS Chatterjee Subimal , Heath Timothy B. ( 1996 ), “Conflict and Loss Aversion in Multiattribute Choice: The Effects of Trade-Off Size and Reference Dependence on Decision Difficulty,” Organizational Behavior and Human Decision Processes 67 2 , 144 – 55 . Google Scholar CrossRef Search ADS Chernev Alexander ( 2006 ), “Decision Focus and Consumer Choice among Assortments,” Journal of Consumer Research 33 1 , 50 – 9 . Google Scholar CrossRef Search ADS Dhar Ravi ( 1996 ), “The Effect of Decision Strategy on Deciding to Defer Choice,” Journal of Behavioral Decision Making , 9 ( December ), 265 – 81 . Google Scholar CrossRef Search ADS Dhar Ravi ( 1997 ), “Consumer Preference for a No-Choice Option,” Journal of Consumer Research , 24 2 , 215 – 31 . 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 Dhar Ravi , Sherman Steven J. ( 1996 ), “The Effect of Common and Unique Features in Consumer Choice,” Journal of Consumer Research , 23 3 , 193 – 203 . Google Scholar CrossRef Search ADS Dhar Ravi , Simonson Itamar ( 2003 ), “The Effect of Forced Choice on Choice,” Journal of Marketing Research , 40 2 , 146 – 60 . Google Scholar CrossRef Search ADS Dommermuth William P. , Millard William J. Jr. ( 1967 ), “Consumption Coincidence in Product Evaluation,” Journal of Marketing Research , 4 4 , 388 – 90 . Google Scholar CrossRef Search ADS Einhorn Hillel J. , Hogarth Robin M. ( 1981 ), “Behavioral Decision Theory: Processes of Judgment and Choice,” Journal of Accounting Research , 19 ( Spring ), 1 – 31 . Google Scholar CrossRef Search ADS Erez Amir , Isen Alice M. ( 2002 ), “The Influence of Positive Affect on the Components of Expectancy Motivation,” Journal of Applied Psychology , 87 ( December ), 1055 – 67 . Google Scholar CrossRef Search ADS Faraji-Rad Ali , Moeini-Jazani Mehrad , Warlop Luk ( 2013 ), “Women Seek More Variety in Rewards when Closer to Ovulation,” Journal of Consumer Psychology , 23 4 , 503 – 8 . Google Scholar CrossRef Search ADS Garbarino Ellen C. , Edell Julie A. ( 1997 ), “Cognitive Effort, Affect, and Choice,” Journal of Consumer Research , 24 2 , 147 – 58 . Google Scholar CrossRef Search ADS Gilbert Daniel T. , Giesler R. Brian , Morris Kathryn A. ( 1995 ), “When Comparisons Arise,” Journal of Personality and Social Psychology , 69 2 , 227 – 36 . Google Scholar CrossRef Search ADS Goldberg Marvin E. , Gorn Gerald J. ( 1987 ), “Happy and Sad TV Programs: How They Affect Reactions to Commercials,” Journal of Consumer Research , 2 3 , 237 – 56 . Gorn Gerald J. , Goldberg Marvin E. , Basu Kunal ( 1993 ), “Mood, Awareness, and Product Evaluation,” Journal of Consumer Psychology , 2 3 , 237 – 56 . Google Scholar CrossRef Search ADS Goukens Caroline , Dewitte Siegried , Pandelaere Mario , Warlop Luk ( 2007 ), “Wanting a Bit(e) of Everything: Extending the Valuation Effect to Variety Seeking,” Journal of Consumer Research , 34 3 , 386 – 94 . Google Scholar CrossRef Search ADS Greenleaf Eric A. , Lehmann Donald R. ( 1995 ), “Reasons for Substantial Delay in Consumer Decision Making,” Journal of Consumer Research , 22 2 , 186 – 99 . Google Scholar CrossRef Search ADS Hammond Kenneth R. , Doyle James K. ( 1991 ), Effects of Stress on Judgment and Decision Making: Part II , Research Triangle Park, NC : US Army Research Office . Hayes Andrew F. ( 2013 ), Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach , New York : Guilford . Huffman Cynthia , Kahn Barbara E. ( 1998 ), “Variety for Sale: Mass Customization or Mass Confusion?” Journal of Retailing , 74 ( Winter ), 45 – 53 . Isen Alice M. ( 1993 ), “Positive Affect and Decision Making,” in Handbook of Emotions , ed. Lewis Michael , Haviland Jeannette M. , New York : Guilford , 261 – 77 . Isen Alice M. ( 2001 ), “An Influence of Positive Affect on Decision Making in Complex Situations: Theoretical Issues with Practical Implications,” Journal of Consumer Psychology , 11 2 , 75 – 85 . Google Scholar CrossRef Search ADS Isen Alice M. , Daubman Kimberly A. , Nowicki Gary P. ( 1987 ), “Positive Affect Facilitates Creative Problem Solving,” Journal of Personality and Social Psychology , 52 ( June ), 1122 – 31 . Google Scholar CrossRef Search ADS Isen Alice M. , Geva Nehemia ( 1987 ), “The Influence of Positive Affect on Acceptable Level of Risk: The Person with a Large Canoe Has a Large Worry,” Organizational Behavior and Human Decision Processes , 39 2 , 145 – 54 . Google Scholar CrossRef Search ADS Isen Alice M. , Levin Paula F. ( 1972 ), “Effect of Feeling Good on Helping: Cookies and Kindness,” Journal of Personality and Social Psychology , 21 3 , 384 – 8 . Google Scholar CrossRef Search ADS Isen Alice M. , Means Barbara ( 1983 ), “The Influence of Positive Affect on Decision-Making Strategy,” Social Cognition , 2 1 , 18 – 31 . Google Scholar CrossRef Search ADS Isen Alice M. , Reeve Johnmarshall ( 2005 ), “The Influence of Positive Affect on Intrinsic and Extrinsic Motivation: Facilitating Enjoyment of Play, Responsible Work Behavior, and Self-Control,” Motivation and Emotion , 29 ( December ), 295 – 323 . Google Scholar CrossRef Search ADS Isen Alice M. , Rosenzweig Andrew S. , Young Mark J. ( 1991 ), “The Influence of Positive Affect on Clinical Problem Solving,” Medical Decision Making , 11 3 , 221 – 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 ( December ), 995 – 1006 . Google Scholar CrossRef Search ADS Kahn Barbara E. , Isen Alice M. ( 1993 ), “The Influence of Positive Affect on Variety Seeking among Safe, Enjoyable Products,” Journal of Consumer Research , 20 2 , 257 – 70 . Google Scholar CrossRef Search ADS Kim-Cho Eunice , Khan Uzma , Dhar Ravi ( 2013 ), “Comparing Apples to Apples or Apples to Oranges: The Role of Mental Representation in Choice Difficulty,” Journal of Marketing Research , 50 4 , 505 – 16 . Google Scholar CrossRef Search ADS Lewinsohn Shai , Mano Haim ( 1993 ), “Multi-Attribute Choice and Affect: The Influence of Naturally Occurring and Manipulated Moods on Choice Processes,” Journal of Behavioral Decision Making , 6 1 , 33 – 51 . Google Scholar CrossRef Search ADS Liberman Nira , Trope Yaacov ( 1998 ), “The Role of Desirability and Feasibility Considerations in Near and Distant Future Decisions: A Test of Temporal Construal Theory,” Journal of Personality and Social Psychology , 75 1 , 5 – 18 . Google Scholar CrossRef Search ADS 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 Luce Mary Frances , Bettman James R. , Payne John W. ( 1997 ), “Choice Processing in Emotionally Difficult Decisions,” Journal of Experimental Psychology: Learning, Memory, and Cognition , 23 ( March ), 384 – 405 . Google Scholar CrossRef Search ADS Luce Mary Frances , Payne John W. , Bettman James R. ( 1999 ), “Emotional Trade-Off Difficulty and Choice,” Journal of Marketing Research , 36 ( May ), 143 – 59 . Google Scholar CrossRef Search ADS Meloy Margaret G. ( 2000 ), “Mood-Driven Distortion of Product Information,” Journal of Consumer Research , 27 3 , 345 – 59 . Google Scholar CrossRef Search ADS Meloy Margaret G. , Russo J. Edward , Miller Elizabeth Gelfand ( 2006 ), “Monetary Incentives and Mood,” Journal of Marketing Research , 43 2 , 267 – 75 . Google Scholar CrossRef Search ADS Montgomery Henry ( 1983 ), “Decision Rules and the Search for a Dominance Structure: Towards a Process Model of Decision Making,” Advances in Psychology , 14 , 343 – 69 . Google Scholar CrossRef Search ADS Novemsky Nathan , Dhar Ravi , Schwarz Norbert , Simonson Itamar ( 2007 ), “Preference Fluency in Choice,” Journal of Marketing Research , 44 ( August ), 347 – 56 . Google Scholar CrossRef Search ADS Nowlis Stephen M. , Simonson Itamar ( 1996 ), “The Effect of New Product Features on Brand Choice,” Journal of Marketing Research , 33 ( February ), 36 – 46 . Google Scholar CrossRef Search ADS Pham Michel Tuan ( 1998 ), “Representativeness, Relevance, and the Use of Feelings in Decision Making,” Journal of Consumer Research , 25 ( September ), 144 – 59 . http://dx.doi.org/10.1111/1468-0106.00037 Google Scholar CrossRef Search ADS Pocheptsova Anastasiya , Amir On , Dhar Ravi , Baumeister Roy F. ( 2009 ), “Deciding Without Resources: Resource Depletion and Choice in Context,” Journal of Marketing Research , 46 ( June ), 344 – 55 . Google Scholar CrossRef Search ADS Pocheptsova Anastasiya , Peterson Francine , Etkin Jordan ( 2015 ), “Using Products to Pursuit Multiple Goals: The Effect of Positive Mood on Perceived Means Instrumentality,” Journal of Consumer Psychology , 25 2 , 296 – 303 . Google Scholar CrossRef Search ADS Pyone Jin Seok , Isen Alice M. ( 2011 ), “Positive Affect, Intertemporal Choice, and Levels of Thinking: Increasing Consumers’ Willingness to Wait,” Journal of Marketing Research , 48 ( June ), 532 – 43 . Google Scholar CrossRef Search ADS Ratchford Brian T. ( 1982 ), “Cost-Benefit Models for Explaining Consumer Choice and Information Seeking Behavior,” Management Science , 28 ( February ), 197 – 212 . Google Scholar CrossRef Search ADS Redelmeier Donald A. , Shafir Eldar ( 1995 ), “Medical Decision Making in Situations that Offer Multiple Alternatives,” JAMA , 273 ( January ), 302 – 5 . Google Scholar CrossRef Search ADS Schrift Rom Y. , Netzer Oded , Kivetz Ran ( 2011 ), “Complicating Choice,” Journal of Marketing Research , 48 2 , 308 – 26 . Google Scholar CrossRef Search ADS Schwarz Norbert , Bless Herbert ( 1991 ), “Happy and Mindless, but Sad and Smart? The Impact of Affective States on Analytic Reasoning,” in Emotion and Social Judgments , ed. Forgas Joseph Paul , Elmsford, New York : Pergamon Press , 55 – 71 . Schwarz Norbert , Clore Gerald L. ( 1983 ), “Mood, Misattribution, and Judgments of Well-Being: Informative and Directive Functions of Affective States,” Journal of Personality and Social Psychology , 45 ( September ), 513 – 23 . Google Scholar CrossRef Search ADS Schwarz Norbert , Clore Gerald L. ( 1996 ), “Feelings and Phenomenal Experiences,” in Social Psychology: Handbook of Basic Principles , ed. Kruglanski Arie W. , Higgins E. Tory , New York : Guilford , 385 – 407 . Sela Aner , Berger Jonah ( 2012 ), “Decision Quicksand: How Trivial Choices Suck Us In,” Journal of Consumer Research , 39 2 , 360 – 70 . Google Scholar CrossRef Search ADS Sela Aner , Berger Jonah , Liu Wendy ( 2009 ), “Variety, Vice, and Virtue: How Assortment Size Influences Option Choice,” Journal of Consumer Research , 35 6 , 941 – 51 . Google Scholar CrossRef Search ADS Sela Aner , Berger Jonah , Nardini Gia ( 2017 ), “How Tradeoffs Shrink Attribute Hierarchy,” working paper, Department of Marketing, University of Florida, Gainesville, FL 32611. Shiv Baba , Huber Joel ( 2000 ), “The Impact of Anticipating Satisfaction on Consumer Choice,” Journal of Consumer Research , 27 2 , 202 – 16 . Google Scholar CrossRef Search ADS Shugan Steven M. ( 1980 ), “The Cost of Thinking,” Journal of Consumer Research , 7 2 , 99 – 111 . Google Scholar CrossRef Search ADS Stigler George J. ( 1961 ), “The Economics of Information,” Journal of Political Economy , 69 ( June ), 213 – 25 . Google Scholar CrossRef Search ADS Townsend Claudia , Kahn Barbara E. ( 2014 ), “The Visual Preference Heuristic: The Influence of Visual versus Verbal Depiction on Assortment Processing, Perceived Variety, and Choice Overload,” Journal of Consumer Research , 40 5 , 993 – 1015 . Google Scholar CrossRef Search ADS Tversky Amos , Shafir Eldar ( 1992 ), “Choice under Conflict: The Dynamics of Deferred Decision,” Psychological Science , 3 ( November ), 358 – 61 . Google Scholar CrossRef Search ADS Watson David , Clark Lee A. , Tellegen Auke ( 1988 ), “Development and Validation of Brief Measures of Positive and Negative Affect: The PANAS Scales,” Journal of Personality and Social Psychology , 54 6 , 1063 – 70 . Google Scholar CrossRef Search ADS Wegener Duane T. , Petty Richard E. ( 1994 ), “Mood Management across Affective States: The Hedonic Contingency Hypothesis,” Journal of Personality and Social Psychology , 66 6 , 1034 – 48 . Google Scholar CrossRef Search ADS Wegener Duane T. , Petty Richard E. , Smith Stephen M. ( 1995 ), “Positive Mood Can Increase or Decrease Message Scrutiny: The Hedonic Contingency View of Mood and Message Processing,” Journal of Personality and Social Psychology , 69 1 , 5 – 15 . Google Scholar CrossRef Search ADS Weitzman Martin L. ( 1979 ), “Optimal Search for the Best Alternative,” Econometrica: Journal of the Econometric Society , 47 ( May ), 641 – 54 . Google Scholar CrossRef Search ADS Wyer Robert S. , Clore Gerald L. , Isbell Linda M. ( 1999 ), “Affect and Information Processing,” Advances in Experimental Social Psychology , 31 , 1 – 77 . Google Scholar CrossRef Search ADS Xu Jing , Jiang Zixi , Dhar Ravi ( 2013 ), “Mental Representation and Perceived Similarity: How Abstract Mindset Aids Choice from Large Assortments,” Journal of Marketing Research , 50 ( August ), 548 – 59 . 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 This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Consumer Research Oxford University Press

When Being in a Positive Mood Increases Choice Deferral

Loading next page...
 
/lp/ou_press/when-being-in-a-positive-mood-increases-choice-deferral-371ria52NJ
Publisher
Oxford University 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/ucx121
Publisher site
See Article on Publisher Site

Abstract

Abstract Consumers’ choices are often accompanied by unrelated incidental moods. The positive mood caused by receiving a compliment, for example, may persist when one is choosing what service to book or which product to buy. How might being in a positive mood affect consumers’ subsequent, unrelated choices? The present research demonstrates that being in a positive mood can make consumers more likely to defer choice. Four studies show that when choosing requires trade-offs between important choice attributes, being in a positive (vs. neutral) mood makes choosing more difficult and therefore increases the likelihood of deferring choice altogether. The findings further understanding of how incidental factors shape choice processes and outcomes and the role of emotions in decision making. choice, choice deferral, choice difficulty, positive mood, trade-offs Consumers’ mood can change in response to various events. Getting stuck in traffic or reading a sad news article, for instance, puts people in a negative mood, whereas receiving a compliment or watching a funny video puts people in a positive mood. These incidental mood states can persist during subsequent, unrelated judgments and decisions (Pham 1998; Schwarz and Clore 1983, 1996; Wyer, Clore, and Isbell 1999). The positive mood caused by watching a funny video, for example, may persist when one is choosing what to make for dinner or which flight to book for an upcoming trip. How might being in a positive mood affect consumers’ subsequent, unrelated choices? In the current research, we propose that being in a positive mood can make consumers more likely to defer choice. Choice options are often described by multiple attributes, and preferences for these attributes tend to conflict. For instance, one available option may offer the best level of one attribute, whereas another may offer the best level of a different attribute (e.g., one flight has the lowest price and another has the fewest number of stops). In order to choose, consumers must make trade-offs between these conflicting attributes (Tversky and Shafir 1992), and such trade-offs can make choice feel more difficult (Chatterjee and Heath 1996; Dhar and Simonson 2003) and choice deferral more likely (Dhar 1996; Dhar and Nowlis 1999; Tversky and Shafir 1992). We argue that being in a positive mood will increase the extent to which consumers focus on important (vs. unimportant) choice attributes. Consequently, when multiple attributes are important and different options are best on each (requiring consumers to make trade-offs between important attributes in order to choose), being in a positive mood should make choosing feel more difficult and increase consumers’ propensity to defer choice altogether. This work makes two main contributions. First, our findings advance understanding of how mood influences decision making. That the emotions caused by choice can impact choice processes and outcomes is well established (Garbarino and Edell 1997; Luce 1998; Luce, Payne and Bettman 1999), yet how incidental mood states—and a positive mood in particular—might shape such outcomes has received less attention. Informing this question, our work demonstrates that being in a positive mood can make subsequent, unrelated choices feel more difficult and therefore increase choice deferral. Further, contributing to the prior work on whether positive mood helps or hinders decision making (Isen and Means 1983; Meloy 2000; Pyone and Isen 2011), our findings suggest that being in a positive mood can help consumers stay focused on the important choice attributes, but proves disadvantageous—increasing choice difficulty and the rate of choice deferral—when those important attributes conflict. Second, our findings contribute to emerging research on the role of incidental factors in choice difficulty and deferral (Novemsky et al. 2007; Xu, Jiang, and Dhar 2013). Going beyond prior work’s emphasis on choice set composition as a driver of choice deferral (Dhar 1996; Tversky and Shafir 1992), we identify a novel incidental factor—a positive mood unrelated to the choice itself—that can increase choice difficulty and deferral. THEORETICAL BACKGROUND The option to defer choice is prevalent in many, if not most, consumer decisions. Consumers defer choice when they postpone or delay selecting an option from among available alternatives (Anderson 2003; Dhar 1996, 1997; Greenleaf and Lehmann 1995). When considering flight options for an upcoming trip, for instance, one could defer choice by deciding to spend more time considering the available options, deciding to seek additional information about the available options, or deciding to explore additional options beyond the current consideration set. One common reason consumers defer choice is because choosing requires trade-offs (Dhar 1997; Dhar and Simonson 2003; Redelmeier and Shafir 1995; Tversky and Shafir 1992). For most consumer choices (e.g., booking a flight, buying a new car, choosing a vacation destination), the choice options are described by multiple attributes (e.g., for a flight, price, number of stops, and on-board internet), with each option having a different level of those attributes. For example, when booking a flight, one flight option might have a lower fare but no on-board internet, whereas another might have a higher fare but offer internet. In order to make a choice, then, consumers must make trade-offs (i.e., give up something of value to obtain something else of value; Bettman et al. 1993; Einhorn and Hogarth 1981) among the various attributes. Choosing the former flight, for instance, would mean sacrificing the ability to stay connected in flight, whereas choosing the latter would require paying more money. We propose that when choosing requires trade-offs between important choice attributes, being in a positive mood will increase consumers’ propensity to defer choice. Our reasoning is based on the notion that a positive mood enables people to be more discerning in their judgments and behavior, with task importance being a key moderator of their responses (Aspinwall 1998, Isen and Reeve 2005). Several distinct lines of research support this view. Examining intrinsic motivation, Isen and Reeve (2005) demonstrate that being in a positive mood increases people’s propensity to prioritize important work tasks (i.e., ones related to higher-order goals) over other enjoyable activities. In the domain of risk-related behaviors, several articles show that being in a positive mood makes people more risk-averse on tasks of high (vs. low) importance (i.e., high vs. low stakes; Isen 1993; Isen and Geva 1987). More germane to the present context, being in a positive mood enhances consumers’ ability to discern important choice attributes from less important ones (Isen 2001). Pyone and Isen (2011), for instance, show that a positive mood affects intertemporal choice (a trade-off between the size of the reward and wait time) by increasing choosers’ focus on the more important aspect of this trade-off (i.e., the size of the reward). Thought protocols revealed that when the choice was important (i.e., the reward was sufficiently valuable), positive-mood participants generated more positive thoughts about reward size than controls, but this effect disappeared when the choice was less important (i.e., the reward was less valuable). Further, Isen and Means (1983) demonstrate that consumers in a positive mood are more likely than controls to eliminate choice alternatives that do not meet set criteria on important dimensions or to ignore unimportant dimensions altogether, an effect that was replicated in a highly consequential setting (i.e., medical decision making; Isen, Rosenzweig, and Young 1991). Building on these findings, we suggest that being in a positive mood will increase the extent to which consumers focus on the important aspects of a decision—namely, the attributes of the choice that are most important. We further argue, however, that this increased focus on important choice attributes can be detrimental—increasing choice difficulty and ultimately choice deferral—when the important attributes conflict (i.e., when different options are best on each, requiring consumers to make trade-offs between them in order to choose). By making salient what one must give up, trade-offs highlight the forgone, underscoring potential losses on each attribute (Brenner, Rottenstreich, and Sood 1999; Carmon and Ariely 2000; Carmon, Wertenbroch, and Zeelenberg 2003). When those potential losses arise from more (vs. less) important attributes, choosing should feel more difficult. For example, if picking a flight required consumers to sacrifice price for fewer stops (both important attributes), this choice would feel more difficult than if choosing instead required consumers to sacrifice price (an important attribute) for on-board entertainment (a relatively unimportant attribute). Because experiencing greater choice difficulty makes choice deferral more likely (Dhar and Nowlis 1999; Tversky and Shafir 1992), when the important attributes of a choice conflict, focusing on them to a greater extent should increase consumers’ propensity to defer choice.1 Consequently, by increasing consumers’ focus on important choice attributes, when multiple attributes are important and different options are best on each (requiring consumers to make trade-offs between important attributes in order to choose), we propose that being in a positive mood will increase choice difficulty, thereby increasing consumers’ propensity to defer choice altogether. H1: When choosing requires trade-offs between important choice attributes, being in a positive mood increases the rate of choice deferral. H2: This increased propensity to defer choice occurs because a positive mood makes choosing between alternatives best on different important attributes more difficult. Further, we argue that the nature of requisite trade-offs (i.e., between important vs. unimportant choice attributes) will moderate these effects. If the need to make trade-offs between important choice attributes plays the critical role we suggest, then when such trade-offs are not required (e.g., choosing only requires trade-offs between unimportant attributes), positive mood’s effects should be attenuated. Consistent with this reasoning, a recent article (Pocheptsova, Peterson, and Etkin 2015) found that in the pursuit of multiple goals, a positive mood influenced subsequent judgments only when those goals were in a high degree of conflict—that is, when consumers’ multiple goals were important and could not easily be prioritized (e.g., goals related to one’s family and work). Although the mechanism (perceived differences between goals) and outcome (evaluation of means to reach goals) in that article differ from the ones examined here, the findings nevertheless suggest that the nature of requisite trade-offs (in the current research, whether those trade-offs are between important choice attributes) may likewise moderate the effects. Consequently, we predict that, whereas a positive mood should increase choice difficulty and deferral when the important choice attributes conflict, when only the unimportant attributes conflict, these effects should be reduced. H3: When choosing requires trade-offs only between unimportant choice attributes, the effect of positive mood on choice difficulty and deferral is attenuated. In summary, we predict that, by increasing the extent to which consumers focus on important (vs. unimportant) choice attributes, when multiple attributes are important and those attributes conflict (requiring consumers to make trade-offs between them in order to choose), being in a positive mood will make choosing feel more difficult, and this increased choice difficulty will increase the likelihood of deferring choice. OVERVIEW OF STUDIES Four studies examine how being in a positive mood affects consumers’ propensity to defer choice. Study 1 provides initial evidence that when multiple choice attributes are important and those attributes conflict (requiring consumers to make trade-offs between them in order to choose), being in a positive mood makes consumers more likely to defer choice. Studies 2–4 replicate this effect in multiple consequential choice settings using different measures of choice deferral and test the proposed underlying role of choice difficulty: when multiple choice attributes are important and choosing requires trade-offs between them, being in a positive mood makes choice feel more difficult, and this increased choice difficulty increases consumers’ propensity to defer choice. In addition to testing for mediation, studies 2 and 3 also explore the proposed underlying process through moderation. Study 2 manipulates the extent to which choosers focus on the important (vs. unimportant) choice attributes: directing (all) participants’ attention to the important choice attributes eliminates positive mood’s effects. Study 3 manipulates whether choosing requires trade-offs between important versus unimportant choice attributes: when the important attributes do not conflict (e.g., a single option is best on all of them), being in a positive mood no longer increases choice difficulty and deferral (hypothesis 3). Lastly, study 4 rules out two important alternative explanations. We shuggest that a positive mood increases choice difficulty and deferral by increasing the extent to which choosers focus on important choice attributes. In contrast to this systematic information processing argument, being in a positive mood could prompt more heuristic information processing (Bless et al. 1990; Schwarz and Bless 1991; Schwarz and Clore 1996), which is also linked to choice deferral (Pocheptsova et al. 2009). Or instead, if making trade-offs between important attributes is emotionally aversive (Einhorn and Hogarth 1981), choosers in positive mood may defer choice in order to avoid resolving those trade-offs (i.e., a mood maintenance-based account; Meloy 2000; Wegener and Petty 1994; Wegener, Petty, and Smith 1995). To show that the systematic processing of important choice attribute information underlies positive mood’s effect, study 4 manipulates choosers’ cognitive capacity: consistent with the proposed theoretical account and inconsistent with alternative explanations, for those in a positive (vs. neutral) mood, imposing cognitive load reduces choice difficulty and therefore decreases the likelihood of choice deferral. Together these studies provide robust and consistent evidence for our predictions: when choosing requires trade-offs between important choice attributes, being in a positive mood increases choice difficulty and choice deferral. STUDY 1: POSITIVE MOOD INCREASES CHOICE DEFERRAL Study 1 tested our first hypothesis. After inducing a positive or neutral mood, we asked participants to choose between two alternatives (flight options) best on different important attributes: one flight option had the most desirable level of one important attribute (price) and the other had the most desirable level of another important attribute (number of stops), requiring participants to make trade-offs between these attributes in order to choose. Notably, in addition to the two alternatives, participants also had the option to defer choice, and we measured the proportion of participants who selected the deferral option. We predicted that in this choice context, where choosing requires trade-offs between important choice attributes, being in a positive mood would increase the propensity to defer choice (hypothesis 1). Design and Method Participants (N = 132; average age 35 years, 49% female) recruited through Amazon’s Mechanical Turk were randomly assigned to a mood condition: positive versus neutral mood. For this and subsequent online studies, a target rule of 50–70 participants per cell determined the sample size. Participants completed two ostensibly unrelated tasks. In the first task, we manipulated incidental mood with a word-association task from prior research (Pyone and Isen 2011). Participants in the positive-mood condition saw 10 positive words (e.g., laughter, fun) and wrote down the first word that came to mind in response to each. Participants in the neutral-mood condition saw 10 neutral words (e.g., hat, chair) and wrote down the first word that came to mind in response to each. Pretest results (N = 51) supported this manipulation. Pretest participants assigned to view the positive words reported more positive feelings on a 10-item five-point PANAS scale (α = .96; Watson, Clark, and Tellegen 1988) than those assigned to view the neutral words (Mpositive = 3.88 vs. Mneutral = 3.43; F(1, 49) = 5.70, p = .021). In the second task, we asked participants to choose a flight for an imagined upcoming trip. Participants viewed two flight options that “fit the time frame of your trip,” each of which was described by four attributes (adapted from Sela and Berger 2012; see web appendix A). We designed the choice set such that each flight option was best on one important attribute but inferior on another. To determine attribute importance, we conducted a pretest. A separate sample of mTurk participants (N = 55, average age 32.9 years, 30.9% women) viewed the same four flight attributes (cost, number of stops, on-board internet, and entertainment) and rated the importance of each (“When choosing a flight, how important are the following attributes to you?”) on a seven-point scale (1 = Very unimportant, 7 = Very important). Examining the means revealed two important attributes and two unimportant attributes: participants rated price (M = 6.69; t(54) = 36.96, p < .001) and number of stops (M = 5.16; t(54) = 5.12, p < .001) as important (above the scale midpoint), and rated on-board internet (M = 3.09; t(54) = –3.21, p = .002) and entertainment (M = 3.00; t(54) = –3.63, p = .001) as unimportant (below the scale midpoint). In the main choice task, we thus considered price and number of stops (on-board internet and entertainment) to be important (unimportant) attributes. We assigned flight option 1 the fewest number of stops and flight option 2 the lowest price, so that choosing required participants to make trade-offs between important choice attributes (see web appendix A). Importantly, in addition to the two flight alternatives, we also gave participants the option to defer choice (in favor of “searching for more options”). We recorded which option (flight option 1, flight option 2, or deferral) they chose.2 Results and Discussion Manipulation Check Among participants who chose one of the two flight options (84.1% of the total sample), mood had no effect on preference between them (χ2(1) = 1.98, p = .159). Overall, 32.4% chose option 1 and 67.6% chose option 2, suggesting that neither option dominated (although option 2, which offered the lowest price, was preferred, consistent with the pretested attribute-importance ratings) and choosing between them required trade-offs between important attributes. Choice Deferral As predicted (hypothesis 1), a positive mood significantly increased the rate of choice deferral (χ2(1) = 6.46, p = .011). Compared to a neutral mood (P = 7.7%), being in a positive mood made participants about three times more likely to defer choice (P = 23.9%). Study 1 provides initial support for our predictions. On a choice task that required trade-offs between important choice attributes, being in a positive (vs. neutral) mood increased the likelihood of deferring choice (hypothesis 1). Despite being incidental to the choice itself, a positive mood can thus carry over to affect consumers’ subsequent, unrelated decisions. STUDY 2: CHOICE DIFFICULTY AND THE ROLE OF ATTRIBUTE FOCUS Study 2 explored the proposed underlying process in two ways. First, we measured choice difficulty and tested for mediation. After inducing a positive or neutral mood, we gave participants a choice between alternatives that were best on different important attributes, as in study 1, and in addition to their propensity to defer choice, we measured how difficult choosing felt. We predicted that being in a positive mood would make choosing feel more difficult, and that this increased difficulty would make participants more likely to defer choice (hypothesis 2). Second, we manipulated the extent to which participants focused on the important choice attributes. If being in a positive mood increases choice difficulty and deferral by increasing the extent to which choosers focus on important choice attributes, as we suggest, then explicitly directing consumers in a neutral mood to focus on the important attributes should attenuate the effects. To test this, for half of participants, we supplemented the choice stimuli with a visual cue that drew attention to the important choice attributes. In the absence of this cue (i.e., the control condition), we expected that being in a positive mood would make choosing feel more difficult and increase choice deferral (hypothesis 1, hypothesis 2). In the presence of this cue, however, we expected that neutral-mood participants would have as much difficulty choosing, and thus be as likely to defer choice, as positive-mood participants in the control condition. In addition, study 2 extended study 1 by examining a consequential choice setting and measuring choice deferral in a different way. We invited lab participants to complete a take-home survey for extra pay, varying the levels of the important survey attributes such that one option was best on one important attribute and the other was best on another. Rather than defer choice in favor of seeking more options, as in study 1, we limited how long participants had to consider the two survey options and, when that time expired, asked them if they wanted more time to think about their choice. In addition to demonstrating the robustness of our choice deferral effect, operationalizing choice deferral in this way casts initial doubt on a potential alternative explanation due to heuristic processing (see study Discussion for rationale). Design and Method Participants (N = 209; 60% female, age not collected) recruited from a university’s behavioral lab were randomly assigned to one condition of a 2 (mood: positive vs. neutral) × 2 (attribute focus: control, enhanced focus) between-subjects design. For this and subsequent lab studies, lab capacity and participant availability determined the sample size. Participants completed two ostensibly unrelated tasks. In the first task, we manipulated incidental mood using the word-association task from study 1. In the second task, we asked participants to choose a take-home survey to complete for extra pay. Participants read, “As part of today’s lab session, you have the opportunity to complete a take-home survey for extra pay. We are collecting data for several surveys, so you can choose which you would like to complete. Students who complete the survey will be entered in a lottery for a cash prize.” Participants viewed two survey options, each of which was described by six attributes (see web appendix A). We designed the choice set such that each survey option was the best on one important attribute but inferior on another. To determine attribute importance, we conducted a pretest. A separate sample of participants (N = 52, 63.5% women) from the same lab pool viewed the same six survey attributes (lottery prize, odds of winning, duration, content, topic, and format) and rated the importance of each (“When choosing a paid take-home survey, how important are the following attributes to you?”) on a seven-point scale (1 = Very unimportant, 7 = Very important). Examining the means revealed two important and four unimportant attributes: participants rated the lottery prize (M = 5.44; t(51) = 6.65, p < .001) and odds of winning (M = 6.15; t(51) = 12.19, p < .001) as important (above the scale midpoint), and rated duration (M = 4.44; t(51) = 1.87, p = .068), content (M = 3.90; t(51) = –.35, p = .729), topic (M = 3.75, t(51) = –.99, p = .328), and format (M = 3.69; t(51) = –1.21, p = .231) as unimportant (equal to or below the scale midpoint). In the main choice task, we thus considered lottery prize and odds of winning (duration, content, topic, and format) to be important (unimportant) attributes. We assigned survey option 1 the highest prize and survey option 2 the best odds, so that choosing required participants to make trade-offs between important choice attributes (see web appendix A). To manipulate participants’ focus on the important choice attributes, we varied the visual depiction of the choice attribute information (see web appendix A). In the enhanced-focus condition, we used bright colors to highlight the important attributes. In the control condition, we provided no such visual cue. A pretest (mTurk panelists, N = 206) supported this manipulation, demonstrating that visually highlighting the important choice attributes increased the extent to which choosers focused on those attributes (and thus trade-offs between them; see web appendix B). Importantly, prior to choosing between the survey options, participants were given the option to defer choice (in favor of spending more time deciding which option they preferred). We told participants that they had 30 seconds to consider the two survey options and then make their choice. At the end of those 30 seconds, we asked them, “Would you like more time to think about your choice?” with response options “Yes, I’d like more time” (i.e., the deferral option) or “No, I will choose now.” We recorded which option participants chose, and those who indicated wanting to choose right away then made their selection (those who indicated wanting more time were told they could make their choice as they left the lab).3,4 Next, on a separate page, we measured choice difficulty using three items (Iyengar and Lepper 2000; see also Xu et al. 2013): “How difficult was it for you to choose the survey that you wanted?” “How frustrated did you feel when making the choice?” and “How hesitant did you feel when making the choice?” (1 = Not at all, 7 = Very much). We combined these items to form a choice-difficulty index (α = .89). Finally, to better understand positive mood’s effects, we collected ancillary measures. We have argued that being in a positive mood makes choosing between alternatives best on different important attributes more difficult because it increases the extent to which choosers focus on the important attributes (which are in conflict). If our reasoning is correct, then positive mood should influence the extent to which participants relied on the important (vs. unimportant) attributes in making their choice: compared to a neutral mood, being in a positive mood should lead the important (unimportant) attributes to play more (less) of a role. To explore this reasoning, we showed participants the six survey attributes and asked them, “How important was each of the following attributes to you in choosing a survey?” (six items, all 1 = Very unimportant, 7 = Very important). In addition, we measured anticipated regret. Anticipated regret can increase choice deferral (Anderson 2003). Because making trade-offs is aversive (Einhorn and Hogarth 1981), and individuals in a positive mood are motivated to maintain that positive state (Clark and Isen 1982; Isen and Levin 1972), one could wonder whether anticipated regret contributes to the present effects. To test this, we asked participants, “How much do you anticipate regretting your choice?” (1 = Not at all, 7 = Very much). Results Manipulation Check A logistic regression of preference on mood, attribute-focus condition, and their interaction revealed no significant effects (ps > .550). Overall, 36.8% of these participants chose option 1 and 63.2% chose option 2, suggesting neither option dominated (although option 2 was preferred) and choosing between them required trade-offs between important attributes. Choice Deferral A logistic regression of choice deferral on mood condition, attribute-focus condition, and their interaction revealed main effects of mood (β = .79, Wald = 3.51, p = .061) and attribute-focus condition (β = .80, Wald = 3.65, p = .056), qualified by the expected interaction (β = –1.02, Wald = 3.07, p = .080; figure 1a). FIGURE 1A View largeDownload slide FOCUSING ON THE IMPORANT ATTRIBUTES MODERATES THE CHOICE DEFERRAL EFFECT FIGURE 1A View largeDownload slide FOCUSING ON THE IMPORANT ATTRIBUTES MODERATES THE CHOICE DEFERRAL EFFECT As predicted and consistent with study 1, in the control condition, a positive mood increased preference for the deferral option (hypothesis 1). When choosing between alternatives best on different important attributes (requiring participants to make trade-offs between those attributes in order to choose), being in a positive mood made participants more likely to defer choice (Ppositive = 43.6% vs. Pneutral = 26.0%; χ2(1) = 3.57, p = .059). In the enhanced-focus condition, however, this effect was attenuated. When the important survey attributes were visually highlighted, the difference between mood conditions in the propensity to defer choice vanished (Ppositive = 38.3% vs. Pneutral = 43.9%; χ2 < 1). Supporting our reasoning, the neutral-mood condition drove this attenuation. Among neutral-mood participants, directing attention to the important choice attributes made deferral more likely (Menhanced-focus = 43.9% vs. Mcontrol = 26.0%; χ2(1) = 3.71, p = .054). However, because positive-mood participants should spontaneously focus on such attributes to a greater extent, we found no effect among them (Menhanced-focus = 38.3% vs. Mcontrol = 43.6%; χ2 < 1). Choice Difficulty A 2 (mood) × 2 (attribute-focus) ANOVA revealed a main effect of mood (Mpositive = 4.29 vs. Mneutral = 3.80; F(1, 205) = 9.90, p = .002), qualified by the expected interaction (F(1, 205) = 4.49, p = .035; figure 1b). There was no main effect of attribute-focus condition (F(1, 205) = 1.98, p = .161). FIGURE 1B View largeDownload slide FOCUSING ON THE IMPORANT ATTRIBUTES MODERATES THE CHOICE DIFFICULTY EFFECT FIGURE 1B View largeDownload slide FOCUSING ON THE IMPORANT ATTRIBUTES MODERATES THE CHOICE DIFFICULTY EFFECT As predicted, in the control condition, a positive mood increased choice difficulty (hypothesis 2). When choosing required participants to make trade-offs between important choice attributes, being in a positive mood made the choice feel more difficult (Mpositive = 4.34 vs. Mneutral = 3.51; F(1, 205) = 13.97, p < .001). In the enhanced-focus condition, however, this effect was attenuated. When the important survey attributes were visually highlighted, there was no difference in choice difficulty across mood conditions (Mpositive = 4.23 vs. Mneutral = 4.06; F < 1). As expected and consistent with the choice deferral results, the neutral-mood condition drove this attenuation. Among neutral-mood participants, directing attention to the important choice attributes made choosing feel more difficult (Menhanced-focus = 4.06 vs. Mcontrol = 3.51; F(1, 205) = 6.37, p = .012), but because positive-mood participants should spontaneously focus on such attributes to a greater extent, we found no effect among them (Menhanced-focus = 4.23 vs. Mcontrol = 4.34; F < 1). Moderated Mediation A bootstrapping moderated mediation analysis (model 7; Hayes 2013) with choice difficulty as the mediator revealed a significant index of moderated mediation (Index = –.68, 95% CI [–1.46 to –.03]). Supporting our reasoning, in the control condition, the indirect effect was positive and significant (ab = .84, 95% CI [.38 to 1.41]): being in a positive mood increased the rate of choice deferral by making choosing feel more difficult. In the enhanced-focus condition, however, the indirect effect was not significant (ab = .16, 95% CI [–.35 to .66]): when the important survey attributes were visually highlighted, being in a positive mood no longer influenced choice difficulty, and thus no longer affected choice deferral. Further supporting our theory, in the neutral-mood condition, the indirect effect of choice difficulty was positive and significant (ab = .55, 95% CI [.11 to 1.10]): directing attention to the important choice attributes made neutral-mood participants more likely to defer choice because it made choosing feel more difficult. In the positive-mood condition, however, the indirect effect was not significant (ab = –.11, 95% CI [–.58 to .35]): among positive-mood participants, making the important attributes more salient did not affect choice difficulty, and thus did not affect choice deferral. Ancillary Analyses First, we analyzed the attribute importance ratings. To explore whether positive mood’s effects are due to an increased focus on the important (vs. unimportant) choice attributes, as we argued, we created an “important attributes” score (the average of the importance ratings for the survey attributes pretested to be important: lottery prize and odds of winning) and an “unimportant attributes” score (the average of the importance ratings for the survey attributes pretested to be unimportant: duration, content, topic, and format) and conducted a 2 (mood) × 2 (attribute-focus) × 2 (attribute-type) repeated-measures ANOVA (see web appendix C for raw scores). This analysis revealed a main effect of the within-subjects attribute-type factor (F(1, 205) = 336.72, p < .001), qualified by a three-way interaction (F(1, 205) = 7.30, p = .007). There were no other main effects (Fs < 1) or interactions (ps > .20). Examining the results within each attribute-focus condition supported our reasoning. In the control condition (i.e., in the absence of a cue to focus on the important attributes), being in a positive mood influenced the attribute importance ratings. In addition to a main effect of the within-subjects attribute-type factor (F(1, 205) = 182.50, p < .001), a mood × attribute-type interaction emerged (F(1, 205) = 7.75, p = .006). While there was no difference across mood conditions in the perceived importance of the “important” survey attributes (Mpositive = 5.89 vs. Mneutral = 5.56; F(1, 205) = 1.89, p = .171), mood did influence the perceived importance of the “unimportant” attributes (F(1, 205) = 6.57, p = .011): compared to neutral-mood participants (M = 3.74), those in a positive mood rated these unimportant attributes as less important (M = 3.12). Consistent with the prior work that finds greater marginal effects on the perceived importance of unimportant (vs. important) choice attributes (Nowlis and Simonson 1996; Sela and Berger 2012), these results suggest that being in a positive mood increases choosers’ relative focus on important (vs. unimportant) choice attributes. In the enhanced-focus condition (i.e., when we visually highlighted the important choice attributes), however, while there was still a main effect of the within-subjects attribute-type factor (F(1, 205) = 154.76, p < .001), there was no longer a mood × attribute-type interaction (F(1, 205) = 1.06, p = .304). Consistent with our theory, when we encouraged neutral-mood participants to focus on the important choice attributes (as those in a positive mood do naturally), mood did not influence the perceived importance of the “important” (Mpositive = 5.54 vs. Mneutral = 5.65; F < 1) or the “unimportant” (Mpositive = 3.62 vs. Mneutral = 3.37; F(1, 205) = 1.03, p = .311) attributes. Second, we examined anticipated regret. A 2 (mood) × 2 (attribute-focus) ANOVA only revealed a main effect of attribute-focus condition (Menhanced-focus = 2.54, Mcontrol = 2.04; F(1, 205) = 5.17, p = .024). We found no main effect of mood (F(1, 205) = 1.18, p = .278) or mood × attribute-focus interaction (F(1, 205) = 1.11, p = .294). Thus, although anticipated regret can contribute to choice deferral (Anderson 2003), in the present context, it does not seem to play a critical role. Discussion Study 2 supports our predictions in a consequential choice setting and demonstrates the proposed underlying process in two ways. First, we measured choice difficulty and showed that this drove the choice deferral effect. As predicted (hypothesis 2), being in a positive mood made choosing feel more difficult, and these increased feelings of choice difficulty explained why a positive mood increased choosers’ propensity to defer choice. Second, we manipulated the extent to which participants focused on the important choice attributes and found that this moderated positive mood’s effects. In the control condition, being in a positive mood increased choice difficulty and the rate of choice deferral (hypothesis 1, hypothesis 2), but when we visually highlighted the important choice attributes, these effects disappeared. Supporting our argument that a positive mood increases consumers’ focus on important choice attributes, explicitly directing neutral-mood participants to focus on the important attributes made them have as much difficulty choosing, and thus be as likely to defer choice, as positive-mood participants in the control condition. Ancillary analyses of the perceived attribute important measures further underscored the role of attribute focus in determining positive mood’s effects. In the control condition, compared to neutral-mood participants, those in a positive mood relied less on the unimportant attributes (and thus relatively more on the important attributes) to make their choice. However, when we encouraged all participants, regardless of their mood state, to focus on the important choice attributes, all participants equally prioritized the important over the unimportant attributes. These results support our reasoning that being in a positive mood makes choosing more difficult and choice deferral more likely by increasing the extent to which consumers focus on important (vs. unimportant) choice attributes. Finally, the results of study 2 cast initial doubt on a potential alternative explanation based on heuristic processing. Although a large body of work documents that a positive mood leads people to process information more systematically (Aspinwall 1998; Erez and Isen 2002; Isen, Daubman, and Nowicki 1987), other findings suggest a positive mood can instead prompt heuristic information processing (Bless et al. 1990; Schwarz and Bless 1991; Schwarz and Clore 1996). Because heuristic processing can also increase choice deferral (at least in some situations; Pocheptsova et al. 2009), one could wonder whether a positive mood makes choice deferral more likely by increasing heuristic, rather than systematic (as we argue; see also Dhar 1997; Dhar and Nowlis 1999), processing of choice attribute information. Casting initial doubt on this possibility, study 2 participants deferred choice in favor of spending more time considering the available options—a delay strategy that is inconsistent with a heuristic-processing account (which would suggest spending less time deciding). Studies 3 and 4 further rule out this account by again measuring the decision to defer choice in favor of spending more time deciding (study 3) and by manipulating cognitive capacity (study 4). STUDY 3: THE MODERATING ROLE OF TRADE-OFF IMPORTANCE Study 3 further examined the underlying mechanism by manipulating whether the important choice attributes were in conflict. If making trade-offs between important choice attributes plays the critical role we suggest, then when such trade-offs are not required (e.g., choosing requires trade-offs only between unimportant attributes), positive mood’s effects should be attenuated (hypothesis 3). To test this, we had half of participants make a choice that required trade-offs between important choice attributes, as in studies 1 and 2, and for the remaining half, we altered the choice attribute values such that choosing required trade-offs only between unimportant attributes. Consistent with the previous results, we expected that when the requisite trade-offs were important, being in a positive mood would make choosing feel more difficult and therefore increase choice deferral (hypothesis 1, hypothesis 2). When the requisite trade-offs were unimportant, however, we expected no such differences across mood conditions (hypothesis 3). In addition, study 3 explored the generalizability of our effects by testing them in a different consequential choice setting. Design and Method Participants (N = 243; average age 21 years, 38% female) recruited from a university’s behavioral lab were randomly assigned to one condition of a 2 (mood: positive vs. neutral) × 2 (trade-off importance: important vs. unimportant) between-subjects design. Participants completed two ostensibly unrelated tasks. In the first task, we manipulated incidental mood using the word-association task from the prior studies. In the second task, we asked participants to choose a portable wireless speaker. Participants read, “As part of today’s study, you will be entered in a lottery for the speaker of your choice (valued at $199). At the conclusion of the study, we will notify the winners to come pick up their speaker.” All participants viewed two available options, each of which was described by seven attributes (see web appendix A). As in the prior studies, in the important trade-offs condition, we designed the choice set such that each speaker option was best on one important attribute but inferior on another. In the unimportant trade-offs condition, we designed the choice set such that one option dominated the other on all important attributes. To determine attribute importance, we ran a pretest. A separate sample of participants (N = 47, average age 23 years, 49% women) from a university’s behavioral lab viewed the same seven speaker attributes (battery life, Bluetooth, power supply, weight, color, range, micro-USB ports) and rated the importance of each (“When choosing a portable wireless speaker, how important are the following attributes to you?”) on a seven-point scale (1 = Very unimportant, 7 = Very important). Examining the means identified four important and three unimportant attributes: participants rated battery life (M = 5.85; t(46) = 9.08, p < .001), range (M = 5.77; t(46) = 6.98, p < .001), power supply (M = 4.91; t(46) = 3.53, p = .001), and Bluetooth (M = 5.06; t(46) = 4.39, p < .001) as important (above the scale midpoint), and rated weight (M = 3.53; t(46) = –2.01, p = .051), color (M = 3.85; t < 1), and micro-USB ports (M = 3.79; t < 1) as unimportant (equal to or below the scale midpoint). In the main experiment, we thus considered battery life, range, power supply, and Bluetooth (weight, color, and micro-USB ports) to be important (unimportant) attributes. To manipulate trade-off importance, we varied the attribute levels of the important speaker attributes. In the important trade-offs condition, we assigned speaker option 1 (Bose) the furthest range and speaker option 2 (Jawbone) the longest battery life (and the same power supply and Bluetooth levels to both), so that choosing required trade-offs between important choice attributes. In the unimportant trade-offs condition, we assigned speaker option 2 (Jawbone) the best level of both range and battery life (holding the other important attributes constant), so that choosing required trade-offs only between the unimportant attributes (weight, color, and micro-USB ports; see web appendix A). Importantly, prior to choosing between the speaker options, participants were given the option to defer choice (in favor of spending more time deciding which option they preferred). As in study 2, we asked them, “Would you like more time to think about your choice?” with response options “Yes, I’d like more time” (i.e., the deferral option) or “No, I will choose now.” We recorded which option participants chose, and those who indicated wanting to choose right away then made their selection (those who indicated wanting more time were told they could make their choice at the end of the study).5,6 Finally, on a separate page, we measured choice difficulty using the same three items from study 2 (1 = Not at all, 7 = Very much): “How difficult was it for you to choose the speaker that you wanted?” “How frustrated did you feel when making the choice?” and “How hesitant did you feel when making the choice?” combined to form a choice-difficulty index (α = .83). Results Manipulation Check A logistic regression of preference on mood, trade-off importance condition, and their interaction revealed only the expected main effect of trade-off importance (Wald = 16.94, p < .001). There was no main effect of mood or interaction (mood: Wald < 1; interaction: Wald = 1.58, p = .210). Supporting our manipulation, although neither option dominated in the important trade-offs condition (51.6% chose option 1 and 48.4% chose option 2), option 2 (which was best on the important attributes) dominated option 1 in the unimportant trade-offs condition (20.2% chose option 1 and 79.8% chose option 2; χ2(1) = 23.19, p < .001). Choice Deferral A logistic regression of choice deferral on mood condition, trade-off importance condition, and their interaction revealed only the expected interaction (β = 1.51, Wald = 3.23, p = .072; figure 2a). There were no main effects of mood or trade-off importance condition (Wald’s < 1). As predicted and consistent with studies 1 and 2, in the important trade-offs condition, a positive mood increased preference for the deferral option (hypothesis 1). When choosing required trade-offs between important choice attributes, being in a positive mood made participants more likely to defer choice (Ppositive = 25.4% vs. Pneutral = 8.9%; χ2(1) = 5.44, p = .020). In the unimportant trade-offs condition, however, this effect was eliminated (hypothesis 3). When choosing required trade-offs only between unimportant attributes, being in a positive mood no longer increased the rate of choice deferral (Ppositive = 7.6% vs. Pneutral = 9.7%; χ2 < 1). FIGURE 2A View largeDownload slide TRADE-OFF IMPORTANCE MODERATES THE CHOICE DEFERRAL EFFECT FIGURE 2A View largeDownload slide TRADE-OFF IMPORTANCE MODERATES THE CHOICE DEFERRAL EFFECT Choice Difficulty A 2 (mood) × 2 (trade-off importance) ANOVA on choice difficulty revealed a main effect of mood (Mpositive = 3.86 vs. Mneutral = 3.44; F(1, 239) = 7.87, p = .005) and a main effect of trade-off importance condition (F(1, 239) = 7.29, p = .007). As could be expected, participants had more difficulty choosing when doing so required trade-offs between important (vs. unimportant) attributes (Mimportant = 3.88 vs. Munimportant = 3.45). Importantly, these main effects were qualified by the predicted interaction (F(1, 239) = 5.85, p = .016; figure 2b). A expected and consistent with study 2, when choosing required trade-offs between important choice attributes, being in a positive mood made the speaker choice feel more difficult (Mpositive = 4.28 vs. Mneutral = 3.46; F(1, 239) = 12.95, p < .001). When choosing required trade-offs only between unimportant attributes, however, this effect was attenuated (Mpositive = 3.47 vs. Mneutral = 3.41; F < 1). FIGURE 2B View largeDownload slide TRADE-OFF IMPORTANCE MODERATES THE CHOICE DIFFICULTY EFFECT FIGURE 2B View largeDownload slide TRADE-OFF IMPORTANCE MODERATES THE CHOICE DIFFICULTY EFFECT Moderated Mediation A bootstrapping moderated mediation analysis (model 7; Hayes 2013) with choice difficulty as the mediator revealed a significant index of moderated mediation (Index = .87, 95% CI [.16 to 1.81]). As predicted and consistent with study 2, in the important trade-offs condition, the indirect effect was positive and significant (ab = .94, 95% CI [.38 to 1.75]): being in a positive mood increased the likelihood of deferring choice by making choosing feel more difficult. In the unimportant trade-offs condition, however, the indirect effect was not significant (ab = .07, 95% CI [–.45 to .67]): when choosing required trade-offs only between unimportant attributes, being in a positive mood no longer influenced choice difficulty, and thus no longer affected choice deferral. Discussion Study 3 supports our predictions in a different consequential choice setting and further demonstrates the proposed underlying process in two ways. First, as in study 2, we measured choice difficulty and demonstrated that this drove the effect of positive mood on choice deferral (hypothesis 2). Second, we manipulated whether choosing required trade-offs between important choice attributes and showed this moderated positive mood’s effects (hypothesis 3). Consistent with the prior results, when choosing required trade-offs between important attributes, being in a positive mood increased choice difficulty, and consequently, choice deferral (hypothesis 1, hypothesis 2). When choosing required trade-offs only between unimportant attributes, however, being in a positive mood no longer increased choice difficulty, and consequently, did not affect the propensity to defer choice (hypothesis 3). In line with prior research showing that the nature of requisite trade-offs moderates positive mood’s effects (e.g., Pocheptsova et al. 2015), being in a positive mood increased choice deferral only when choosing required trade-offs between important choice attributes. Note, this moderation occurred even though participants in both trade-off importance conditions had to make trade-offs in order to choose (although these trade-offs were presumably easier in the unimportant trade-offs condition). Thus, the need to make trade-offs alone is not what leads a positive mood to increase choice difficulty and choice deferral. Consistent with prior work showing that a positive mood is discerning and helps people disregard unimportant information (Isen and Means 1983; Pyone and Isen 2011), being in a positive mood increases choice difficulty (and therefore choice deferral) only when choosing requires sacrificing one important attribute for another. The results of study 3 also further argue against a potential alternative explanation based on heuristic processing. As in study 2, study 3 participants deferred choice in favor of spending more time considering the available options—a delay strategy consistent with the proposed systematic-processing account and inconsistent with an alternative one based on heuristic processing. Study 4, reported next, further rules out this potential alternative explanation. STUDY 4: RULING OUT ALTERNATIVE EXPLANATIONS Our final study ruled out two potential alternative explanations for why a positive mood increases choice deferral by manipulating choosers’ cognitive capacity. Reducing cognitive capacity (via cognitive load) impedes systematic information processing (Gilbert, Giesler, and Morris 1995; Shiv and Huber 2000). Thus, if positive mood’s effects are due to the more systematic processing of choice attribute information, as we argue, then cognitive load should primarily influence individuals in a positive mood. Specifically, because they can no longer elaborate on trade-offs between important choice attributes, positive-mood participants under cognitive load should find choosing less difficult and be less likely to defer choice. If positive mood’s effects are instead due to more heuristic information processing, however, then imposing cognitive load should have no such effects. Similarly, if positive mood’s effects are due to a desire to maintain one’s positive mood by avoiding resolving difficult trade-offs (i.e., a mood-maintenance-based account; Meloy 2000; Wegener and Petty 1994; Wegener et al. 1995), then cognitive load should have no impact on positive-mood participants’ propensity to defer choice. Finding that cognitive load reduces choice difficulty and deferral for positive-mood participants would thus support our theory and cast strong doubt on both of these alternative accounts. In addition, study 4 underscored the robustness and external validity of our effects by testing them in a different consequential choice setting and measuring choice deferral in a different way (i.e., deferring in favor of learning more information about the available options). Design and Method Participants (N = 320; average age 21 years, 64% female) recruited from a university’s behavioral lab were randomly assigned to one condition of a 2 (mood: positive vs. neutral) × 2 (cognitive load: control vs. load) between-subjects design. Eight individuals reported audio problems and were excluded, leaving a sample of 312. Participants completed two ostensibly unrelated tasks. In the first task, we manipulated incidental mood in a different way. Participants viewed one of two video clips pretested to elicit the target mood state (positive or neutral, depending on condition). Supporting the manipulation, pretest participants (N = 94) assigned to view the positive-mood clip reported more positive feelings on a 10-item seven-point PANAS scale (α = .97) than those assigned to view the neutral-mood clip (Mpositive = 5.75 vs. Mnetural = 4.34; F(1, 92) = 41.33, p < .001). In the second task, we asked participants to choose a single-serving protein powder packet (see web appendix A). We told all participants that as part of today’s study, they would receive their preferred alternative at the end of the session. Prior to showing participants the available alternatives, we manipulated cognitive load. In the cognitive-load condition, we asked participants to remember an eight-digit number for the remainder of the study. Remembering a multidigit number has reliably been shown to reduce individuals’ cognitive capacity and impede their ability to systematically process information (Gilbert et al. 1995; Shiv and Huber 2000). We told participants the purpose of remembering the number was to help us learn about the decision-making process, and that we would ask them to recall it toward the end of the survey. In the control (no-load) condition, participants proceeded directly to the choice task. Participants viewed three protein powder options, each of which was described by five attributes (see web appendix A), which we chose based on typical nutritional information provided on the packaging of protein supplements (calories, size, protein, sugar, and flavor). Unlike in the previous studies, here we explicitly told participants which attributes were important (protein and sugar content). Participants read, “Industry experts suggest that the protein and sugar content are the most important attributes when evaluating different protein powders.” Importantly, prior to choosing between the protein powder options, participants were given the option to defer choice (in favor of learning more about the available alternatives). Rather than provide complete information about both important attributes, as in the prior studies, we revealed the protein content of each powder option but hid the sugar content on two of them (see web appendix A). Thus, based on the information provided, participants could see that option 3 (Designer Whey) had the least amount of protein and option 2 (Biochem) had the most, but they were unable to discern how the alternatives differed on sugar content. This meant that participants could choose based on one important attribute (protein), but lacked information to choose based on both. We told participants that before making their choice, they had the option to acquire more information about the options, and specifically, that “the cells marked // (i.e., the missing sugar values) in the table will populate.” Participants further read that this additional information could take a few minutes to load and they would have to wait. Then we asked them, “Would you like more information before making your choice?” with response options “Yes, I’d like more information” (i.e., the deferral option) or “No, I will choose now.” We recorded which option participants chose, and those who indicated wanting to choose right away then made their selection (those who indicated wanting more information were told they could see the complete table and make their choice at the end of the study).7,8 Finally, on a separate page, we removed the cognitive load from participants in that condition by asking them to type the number we had asked them to remember. Then, we measured choice difficulty using the same three items from the prior studies (1 = Not at all, 7 = Very much): “How difficult was it for you to choose the protein powder that you wanted?” “How frustrated did you feel when making the choice?” and “How hesitant did you feel when making the choice?” which we combined to form a choice-difficulty index (α = .77). Results Manipulation Check A series of logistic regressions (each comparing preference between two of the three alternatives) on mood, cognitive-load condition, and their interaction revealed no significant effects (ps > .20). Overall, 57% of participants chose option 1 (Quest), 21% chose option 2 (Biochem), and 22% chose option 3 (Designer), suggesting that no option dominated (although option 1, which had the most protein, was preferred), and choosing between them required trade-offs between important choice attributes. Choice Deferral A logistic regression of choice deferral on mood condition, cognitive-load condition, and their interaction revealed a significant main effect of mood (β = .78, Wald = 5.67, p = .017), qualified by the predicted interaction (β = –.94, Wald = 4.18, p = .041; figure 3a). There was no main effect of cognitive load (Wald < 1). As predicted and consistent with the previous studies, in the absence of cognitive load, being in a positive mood made participants more likely to defer choice (Ppositive = 63.3% vs. Pneutral = 44.2%; χ2(1) = 5.75, p = .017). Under cognitive load, however, this effect was eliminated. When we imposed a cognitive load prior to considering the choice options, the propensity to defer choice no longer differed across mood conditions (Ppositive = 44.1% vs. Pneutral = 48.1%; χ2 < 1). FIGURE 3A View largeDownload slide COGNITIVE LOAD MODERATES THE CHOICE DEFERRAL EFFECT FIGURE 3A View largeDownload slide COGNITIVE LOAD MODERATES THE CHOICE DEFERRAL EFFECT Supporting our reasoning and ruling out potential alternative accounts, the positive-mood condition drove this attenuation. Putting positive-mood participants under cognitive load decreased their rate of choice deferral to the level of neutral-mood participants (Pload = 44.1% vs. Pcontrol = 63.3%; χ2(1) = 5.75, p = .017), but did not alter neutral-mood participants’ propensity to defer choice (Pload = 48.1% vs. Pcontrol = 44.2%; χ2 < 1). This suggests that positive mood’s effects are due to the more systematic processing of choice attribute information. Choice Difficulty A 2 (mood) × 2 (cognitive load) ANOVA on choice difficulty revealed only the predicted interaction (F(1, 308) = 4.04, p = .045; figure 3b). No main effects of mood (F(1, 308) = 1.74, p = .188) or cognitive-load condition (F < 1) emerged. As predicted and consistent with studies 2 and 3, in the absence of cognitive load, being in a positive mood made choosing feel more difficult (Mpositive = 4.01 vs. Mneutral = 3.53; F(1, 308) = 5.55, p = .019). Under cognitive load, however, this effect was eliminated. When we imposed a cognitive load prior to considering the choice options, choice difficulty no longer differed across mood conditions (Mpositive = 3.60 vs. Mneutral = 3.70; F < 1). FIGURE 3B View largeDownload slide COGNITIVE LOAD MODERATES THE CHOICE DIFFICULTY EFFECT FIGURE 3B View largeDownload slide COGNITIVE LOAD MODERATES THE CHOICE DIFFICULTY EFFECT Consistent with the choice deferral results and supporting our theory, the positive-mood condition drove this attenuation. Putting positive-mood participants under cognitive load made choosing feel less difficult (Mload = 3.60 vs. Mcontrol = 4.01; F(1, 308) = 3.96, p = .047), but did not affect choice difficulty among neutral-mood participants (Mload = 3.70 vs. Mcontrol = 3.53; F < 1). This further suggests that positive mood’s effects cannot be explained by alternative accounts due to heuristic processing or mood maintenance. Moderated Mediation As in studies 2 and 3, we ran a bootstrapping moderated mediation analysis (model 7; Hayes 2013) with choice difficulty as the mediator to test the proposed underlying process. This analysis revealed a significant index of moderated mediation (Index = –.27, 95% CI [–.63 to –.02]). Supporting our theory, in the control condition, the indirect effect was positive and significant (ab = .22, 95% CI [.05 to .49]): being in a positive mood increased the rate of choice deferral by making choosing feel more difficult. In the cognitive-load condition, however, the indirect effect was not significant (ab = –.05, 95% CI [–.26 to .14]): when participants had fewer cognitive resources to process the choice options, being in a positive mood no longer made choosing feel more difficult, and thus no longer increased consumers’ propensity to defer choice. Further supporting our reasoning, in the positive-mood condition, the indirect effect of choice difficulty was negative and significant (ab = –.19, 95% CI [–.43 to –.01]): adding a cognitive load made positive-mood participants less likely to defer choice because it made choosing less difficult. In the neutral-mood condition, however, the indirect effect was not significant (ab = .05, 95% CI [–.46 to .61]): among neutral-mood participants, adding a cognitive load did not influence choice difficulty, and thus did not affect choice deferral. Discussion Study 4 underscores our theorizing in a different consequential choice domain and rules out a key alternative explanation. We manipulated participants’ cognitive resources prior to considering the choice options and showed that this moderated positive mood’s effects. Consistent with the prior results, in the control condition, being in a positive mood increased choice difficulty and the rate of choice deferral (hypothesis 1, hypothesis 2). Under cognitive load, however, these effects disappeared. Supporting our argument that a positive mood increases the extent to which consumers focus on important choice attributes (and thus engage in trade-offs between them), which requires cognitive resources (Montgomery 1983; Pocheptsova et al. 2009), reducing positive-mood participants’ cognitive capacity attenuated the effects on choice difficulty and choice deferral. The nature of the moderation further rules out potential alternative explanations. Imposing cognitive load caused positive-mood participants to find choosing less difficult and therefore be less likely to defer choice, similar to neutral-mood participants in the control condition. That cognitive load influenced choice difficulty and deferral among positive-mood, rather than neutral-mood, participants is inconsistent with arguments based on heuristic processing or mood maintenance, underscoring that positive mood’s effects are due to the proposed systematic processing of choice attribute information. That being in a positive mood made participants more likely to defer choice in favor of acquiring more information about the important attributes is also inconsistent with both alternative accounts. GENERAL DISCUSSION Incidental moods accompany most, if not all, of consumers’ decisions. While a few prior articles have looked at the effect of a negative mood on decision making (Hammond and Doyle 1991; Lewinsohn and Mano 1993), how a positive mood might impact subsequent, unrelated choices remains largely unexplored (for exceptions, see Isen and Means 1983; Meloy 2000; Meloy, Russo, and Miller 2006). Our work contributes to this line of research by examining how being in a positive mood shapes consumers’ propensity to defer choice. Results of four studies demonstrate that when multiple choice attributes are important and those attributes conflict (requiring trade-offs between them in order to choose), being in a positive mood makes consumers more likely to defer choice (hypothesis 1). We replicated this effect across several choice contexts, both hypothetical (study 1) and consequential (studies 2–4), and with multiple measures of choice deferral: participants in a positive mood were more likely to defer choice in favor of seeking additional alternatives (study 1), having more time to decide (studies 2–3), and learning more information about the available choice options (study 4). Further, the choice deferral effect was robust to a range of important attribute conflicts, including approach/approach (e.g., in study 3, battery life vs. range), approach/avoidance (e.g., in study 4, protein vs. sugar), and avoidance/avoidance (e.g., in study 2, price vs. the number of stops). The studies also provide insight into the underlying process. When multiple choice attributes are important and those important attributes conflict, being in a positive mood makes choosing feel more difficult, which increases consumers’ propensity to defer choice (hypothesis 2, studies 2–4). Based on the prior work that shows a positive mood enables consumers to discern important from unimportant information (Isen and Means 1983; Pyone and Isen 2011), we argue that this occurs because being in a positive mood increases the extent to which consumers focus on important choice attributes. In support, an ancillary analysis in study 2 showed that compared to neutral-mood participants, those in a positive mood relied less on the unimportant attributes (and thus relatively more on the important attributes) to make their choice. Consequently, when choosing required trade-offs only between unimportant attributes (hypothesis 3, study 3), being in a positive mood no longer increased choice difficulty, thereby attenuating its effect on choice deferral. The findings also cast doubt on potential alternative explanations. While one could wonder whether a positive mood increases choice difficulty and deferral by prompting heuristic processing, study 4 rules out this possibility. Because reducing cognitive capacity impedes systematic information processing (Gilbert et al. 1995; Shiv and Huber 2000), that imposing a cognitive load reduced choice difficulty and deferral among positive-mood participants, but had no such effects among neutral-mood participants, demonstrates that positive mood’s effects are due to systematic (vs. heuristic) information processing. That positive-mood participants deferred choice in favor of spending more time deciding (studies 2–3) and learning additional information about the available options (study 4) casts further doubt on this alternative account. In addition, while one could wonder whether being in a positive mood increases choice deferral due to a mood-maintenance-based alternative explanation (because making trade-offs is aversive and people are motivated to maintain positive mood states), our findings cast doubt on this possibility in two ways. First, the moderation by cognitive load in study 4 suggests that it is the extent to which consumers process trade-offs between important attributes—not their motivation to protect a positive mood state—that underlies the present effects. Second, that we found no effect of mood on anticipated regret (study 2) suggests that the motivation to maintain one’s positive mood is not the primary underlying driver. Notably, this null effect also casts doubt on the possibility that positive mood’s effects are due to the increased desirability of important choice attributes (which could increase the anticipated regret associated with forgoing one attribute in favor of another). Theoretical Contributions This research advances understanding of how mood influences decision making. That the emotions experienced during a choice can impact the process and outcomes of that choice is well established (Garbarino and Edell 1997; Luce 1998; Luce, Bettman, and Payne 1997; Luce et al. 1999). Yet while such “integral” mood states have received much attention in the choice literature, less work has explored whether incidental mood states—and a positive mood in particular—might also influence choice processes and outcomes. Our research suggests a novel way that being in a positive mood influences subsequent, unrelated decisions: by increasing the extent to which choosers focus on important (vs. unimportant) choice attributes. When consumers choose between alternatives that are best on different important attributes (requiring trade-offs between them in order to choose), being in a positive mood makes choosing feel more difficult and increases consumers’ propensity to defer choice. This work also informs how mood interacts with choice attribute conflict to impact choice deferral. Separate streams of research have identified two reasons why the need to make trade-offs between important choice attributes can increase consumers’ propensity to defer choice. On one hand, making trade-offs between important attributes increases preference uncertainty, which can make selecting one from among many options more difficult (Dhar 1996, 1997; Dhar and Nowlis 1999; Dhar and Sherman 1996; Dhar and Simonson 2003; Redelmeier and Shafir 1995; Tversky and Shafir 1992). On the other hand, making trade-offs between important attributes is emotionally aversive, which can lead consumers to avoid such trade-offs (Luce 1998; Luce et al. 1997; Luce et al. 1999). Consistent with the former perspective, the current findings show that a positive mood makes consumers more likely to defer choice by increasing the extent to which they focus on important choice attributes (and thus engage in trade-offs between them). The current research also relates to whether being in a positive mood results in better or worse decisions. Prior research provides mixed results on this point. Whereas some work finds that a positive mood improves decision making by helping consumers process information more efficiently (ignoring unimportant information and focusing on what is important; Isen and Means 1983), other work shows that a positive mood results in suboptimal decisions by leading consumers to distort information about the choice options (Meloy 2000; Meloy et al. 2006). Our findings fall in between: being in a positive mood can help consumers stay focused on the important choice attributes, which should generally result in better choice outcomes (Sela, Berger, and Nardini 2017), but proves disadvantageous—increasing choice difficulty and the rate of choice deferral—when those important attributes conflict. Finally, this research contributes to emerging work on the role of incidental factors in choice difficulty and deferral. Prior work on the antecedents of choice difficulty and deferral has primarily focused on structural aspects of the choice set. For example, choosing from a larger number of options (Brenner et al.; Carmon et al. 2003; Iyengar and Lepper 2000; Sela, Berger, and Liu 2009; Shugan 1980), from more varied assortments (Chernev 2006; Huffman and Kahn 1998; Townsend and Kahn 2014), and from choice sets that lack acceptable alternatives (Ratchford 1982; Stigler 1961; Weitzman 1979) can make choosing more difficult and choice deferral more likely. More recently, marketing scholars have begun to explore how incidental factors unrelated to the focal choice might also play a role. For example, processing disfluency and metacognitive difficulty (Novemsky et al. 2007; Schrift et al. 2011; Sela and Berger 2012) and mental abstraction (Kim, Khan, and Dhar 2013; Xu et al. 2013) can impact choice difficulty and deferral. Adding to these findings, we demonstrate that an incidental positive mood can influence consumers’ feelings of choice difficulty and subsequent propensity to defer choice altogether. Implications and Future Research Directions This research has several practical implications. From the consumer’s perspective, the findings further understanding of why choosing might feel difficult. Consumers often find choice be difficult, but may be unaware of the reason(s) why. Unlike more readily observable aspects of choice, such as the appeal of individual options or the size of the choice set, incidental influences on choice difficulty may be harder to diagnose, and thus harder to correct. By identifying a novel factor that impacts choice difficulty (and the likelihood of choice deferral)—being in a positive mood—this research better enables consumers to understand their feelings of choice difficulty. From the marketer’s perspective, our findings have implications for the use of positive mood appeals as a persuasion tactic. Marketers often craft content intended to influence consumers’ mood states (Aaker, Stayman, and Hagerty 1986; Burke and Edell 1989; Goldberg and Gorn 1987) and assume that evoking a positive mood will have desirable effects. While putting consumers in a good mood may indeed result in more favorable evaluations of individual products (Barone, Miniard, and Romeo 2000; Dommermuth and Milliard 1967; Gorn, Goldberg, and Basu 1993), when multiple choice attributes are important and those attributes conflict, it may also inadvertently make choosing between options feel more difficult and choice deferral more likely. Positive mood-boosting tactics may therefore be more effectively employed in situations where consumers are evaluating a single product in isolation, as opposed to comparing multiple products and making trade-offs between them. Similarly, a positive mood may be advantageous when marketers want to encourage consumers to select multiple options from a choice set. While in our studies we did not allow participants to acquire several options at the same time, it is possible that instead of deferring choice, consumers would resolve conflict between important choice attributes by purchasing multiple options from the choice set. Such an effect would provide another theoretical explanation for Kahn and Isen (1993)’s finding that being in a positive mood increases variety seeking among attractive options and would be consistent with the prior work that demonstrates increased variety seeking when more options in the choice set have favorable evaluations (Faraji-Rad, Moeini-Jazani, and Warlop 2013; Goukens et al. 2007). The findings also suggest interesting directions for future research. First, might the desirability versus feasibility of choice option attributes influence positive mood’s effects? Prior work has distinguished between low-level (feasibility) attributes and high-level (desirability) attributes and proposed that construing choice at a higher level should lead consumers to prefer desirability attributes over feasibility ones (Liberman and Trope 1998). In study 1, one could argue that one important choice attribute is high on feasibility (price) and the other on desirability (number of stops), yet we do not find that a positive mood increases preference for the shorter flight option. Dhar and Simonson (1999) suggest one way to reconcile our findings. They show that a trade-off between money and another important choice attribute can be perceived as a trade-off between a resource and a goal or between two goals (with the former leading to prioritization and the latter leading to more balancing). It is possible that in study 1, participants saw money as a goal rather than a resource, leading them to engage in trade-offs between price and number of stops rather than prioritize the number of stops over price. Future work could explore when a positive mood increases the extent to which consumers focus on certain types of attributes (e.g., desirability vs. feasibility), rather than distinguish between important and unimportant attributes more generally. Second, under what situations might being in a positive mood lead consumers to prioritize among important choice attributes (and choose based on just one)? In this work we argued that a positive mood increases consumers’ relative focus on important (vs. unimportant) attributes. In support, in study 2, positive-mood participants perceived the unimportant attributes to be less important than neutral-mood participants; along similar lines, Isen and Meads (1993) found that positive-mood participants more effectively distinguished between multiple more and less important attributes (rather than focus on just one). Thus, empirically, being in a positive mood does not seem to spontaneously lead to such absolute attribute prioritization. Pocheptsova et al. (2015), however, suggests that when important goals serve a single overarching purpose (such as goals to exercise and eat healthy), being in a positive mood may possibly encourage such prioritization. Future research could examine whether highlighting the interconnectedness of important choice attributes would likewise affect their prioritization. Conclusion Consumers’ mood changes in response to various events and such incidental mood states can persist during subsequent, unrelated decisions. Four studies explored how being in a positive mood impacts choice deferral. Results showed that when multiple attributes are important and choosing requires trade-offs between them, being in a positive mood makes choice feel more difficult, which increases consumers’ propensity to defer choice. In addition to structural aspects of choice sets and other contextual factors the incidental moods that accompany consumers’ choices can therefore impact how likely they are to defer choice altogether. DATA COLLECTION INFORMATION The second author supervised data collection by the lab manager at the University of South Carolina behavioral lab (studies 2, 3, and 4). The first author collected the data for study 1 on Amazon’s Mechanical Turk. Study 1 ran in fall 2016. Study 2 ran in spring 2017. Study 3 ran in fall 2015. Study 4 ran in spring 2016. The first author analyzed the data, with input and advice from the second author. Supplementary materials are included in the web appendix accompanying the online version of this article. Footnotes 1 Although disregarding unimportant attributes can in some cases make choosing easier (e.g., by reducing the number of attributes under consideration; Luce et al. 1999; Schrift, Netzer, and Kivetz 2011; Sela and Berger 2012; Tversky and Shafir 1992), when multiple attributes are important and different options are best on each, focusing on the important attributes should exacerbate choice difficulty. 2 Time spent choosing did not differ between mood conditions (Mpositive = 27.27 seconds, Mneutral = 34.33 seconds; F(1, 130) = 1.02, p = .315). Note, here and in subsequent studies, we did not have an a priori prediction for how a positive mood might impact decision time: whereas our prediction that being in a positive mood increases choice difficulty might suggest positive- (vs. neutral-) mood participants spend more time deciding, the prior work that finds being in a positive mood leads people to process choice information more efficiently (Isen and Means 1983) would suggest that positive- (vs. neutral-) mood participants instead spend less time deciding. Consequently, we did not have an expectation for how (or whether) decision time differs across mood conditions. 3 Time spent choosing was capped in this study (at 30 seconds). 4 We drew lottery winners, consistent with the stated odds, from among participants who returned the take-home survey and arranged for those individuals to receive their prize. 5 A 2 (mood) × 2 (trade-off importance) ANOVA on time spent choosing revealed only a main effect of trade-off importance condition (F(1, 239) = 19.87, p < .001). As might be expected, participants spent more time deciding when the trade-offs were between important (vs. unimportant) attributes (Mhigh-importance = 34.09 seconds vs. Mlow-importance = 27.09 seconds). There was no effect of mood (Mpositive = 29.74 seconds, Mneutral = 31.10 seconds, F < 1) or interaction (F < 1). 6 We drew lottery winners (one from each experimental session, approximately a one in 25 chance of winning) and arranged for those individuals to receive their speaker of choice. 7 A 2 (mood) × 2 (cognitive load) ANOVA on time spent choosing revealed a significant main effect of cognitive load (F(1, 308) = 55.23, p < .001). Consistent with the notion that cognitive load reduces information processing capacity, participants spent less time choosing when under cognitive load (Mload = 34.45 seconds vs. Mcontrol = 51.08 seconds). We found no main effect of mood (F < 1), but did find an interaction (F(1, 308) = 4.21, p = .041), such that imposing a cognitive load had a bigger effect on decision time in the positive- (Mload = 32.38 seconds vs. Mcontrol = 53.59 seconds; F(1, 308) = 44.98, p < .001) versus neutral-mood condition (Mload = 36.47 seconds vs. Mcontrol = 48.50 seconds; F(1, 308) = 14.47, p < .001). 8 All participants received their choice of protein powder as they exited the lab. REFERENCES Aaker David A. , Stayman Douglas M. , Hagerty Michael R. ( 1986 ), “Warmth in Advertising: Measurement, Impact, and Sequence Effects,” Journal of Consumer Research , 12 4 , 365 – 81 . Google Scholar CrossRef Search ADS Anderson Christopher J. ( 2003 ), “The Psychology of Doing Nothing: Forms of Decision Avoidance Result from Reason and Emotion,” Psychological Bulletin , 129 ( January ), 139 – 67 . Google Scholar CrossRef Search ADS Aspinwall Lisa G. ( 1998 ), “Rethinking the Role of Positive Affect in Self-Regulation,” Motivation and Emotion , 22 ( March ), 35 – 68 . Barone Michael J. , Miniard Paul W. , Romeo Jean B. ( 2000 ), “The Influence of Positive Mood on Brand Extension Evaluations,” Journal of Consumer Research , 26 4 , 386 – 400 . Google Scholar CrossRef Search ADS Bettman James R. , Johnson Eric J. , Luce Mary Frances , Payne John W. ( 1993 ), “Correlation, Conflict, and Choice,” Journal of Experimental Psychology: Learning, Memory, and Cognition , 19 ( July ), 931 – 51 . Google Scholar CrossRef Search ADS Bless Herbert , Bohner Gerd , Schwarz Norbert , Strack Fritz ( 1990 ), “Mood and Persuasion: A Cognitive Response Model,” Personality and Social Psychology Bulletin , 16 ( June ), 331 – 45 . Google Scholar CrossRef Search ADS Brenner Lyle , Rottenstreich Yuval , Sood Sanjay ( 1999 ), “Comparison, Grouping, and Preference,” Psychological Science , 10 3 , 225 – 9 . Google Scholar CrossRef Search ADS Burke Marian C. , Edell Julie ( 1989 ), “The Impact of Feelings on Ad-Based Affect and Cognition,” Journal of Marketing Research , 26 ( February ), 69 – 83 . Google Scholar CrossRef Search ADS Carmon Ziv , Ariely Dan ( 2000 ), “Focusing on the Forgone: How Value Can Appear So Different to Buyers and Sellers,” Journal of Consumer Research , 27 3 , 360 – 70 . Google Scholar CrossRef Search ADS Carmon Ziv , Wertenbroch Klaus , Zeelenberg Marcel ( 2003 ), “Option Attachment: When Deliberating Makes Choosing Feel Like Losing,” Journal of Consumer Research , 30 1 , 15 – 29 . Google Scholar CrossRef Search ADS Chatterjee Subimal , Heath Timothy B. ( 1996 ), “Conflict and Loss Aversion in Multiattribute Choice: The Effects of Trade-Off Size and Reference Dependence on Decision Difficulty,” Organizational Behavior and Human Decision Processes 67 2 , 144 – 55 . Google Scholar CrossRef Search ADS Chernev Alexander ( 2006 ), “Decision Focus and Consumer Choice among Assortments,” Journal of Consumer Research 33 1 , 50 – 9 . Google Scholar CrossRef Search ADS Dhar Ravi ( 1996 ), “The Effect of Decision Strategy on Deciding to Defer Choice,” Journal of Behavioral Decision Making , 9 ( December ), 265 – 81 . Google Scholar CrossRef Search ADS Dhar Ravi ( 1997 ), “Consumer Preference for a No-Choice Option,” Journal of Consumer Research , 24 2 , 215 – 31 . 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 Dhar Ravi , Sherman Steven J. ( 1996 ), “The Effect of Common and Unique Features in Consumer Choice,” Journal of Consumer Research , 23 3 , 193 – 203 . Google Scholar CrossRef Search ADS Dhar Ravi , Simonson Itamar ( 2003 ), “The Effect of Forced Choice on Choice,” Journal of Marketing Research , 40 2 , 146 – 60 . Google Scholar CrossRef Search ADS Dommermuth William P. , Millard William J. Jr. ( 1967 ), “Consumption Coincidence in Product Evaluation,” Journal of Marketing Research , 4 4 , 388 – 90 . Google Scholar CrossRef Search ADS Einhorn Hillel J. , Hogarth Robin M. ( 1981 ), “Behavioral Decision Theory: Processes of Judgment and Choice,” Journal of Accounting Research , 19 ( Spring ), 1 – 31 . Google Scholar CrossRef Search ADS Erez Amir , Isen Alice M. ( 2002 ), “The Influence of Positive Affect on the Components of Expectancy Motivation,” Journal of Applied Psychology , 87 ( December ), 1055 – 67 . Google Scholar CrossRef Search ADS Faraji-Rad Ali , Moeini-Jazani Mehrad , Warlop Luk ( 2013 ), “Women Seek More Variety in Rewards when Closer to Ovulation,” Journal of Consumer Psychology , 23 4 , 503 – 8 . Google Scholar CrossRef Search ADS Garbarino Ellen C. , Edell Julie A. ( 1997 ), “Cognitive Effort, Affect, and Choice,” Journal of Consumer Research , 24 2 , 147 – 58 . Google Scholar CrossRef Search ADS Gilbert Daniel T. , Giesler R. Brian , Morris Kathryn A. ( 1995 ), “When Comparisons Arise,” Journal of Personality and Social Psychology , 69 2 , 227 – 36 . Google Scholar CrossRef Search ADS Goldberg Marvin E. , Gorn Gerald J. ( 1987 ), “Happy and Sad TV Programs: How They Affect Reactions to Commercials,” Journal of Consumer Research , 2 3 , 237 – 56 . Gorn Gerald J. , Goldberg Marvin E. , Basu Kunal ( 1993 ), “Mood, Awareness, and Product Evaluation,” Journal of Consumer Psychology , 2 3 , 237 – 56 . Google Scholar CrossRef Search ADS Goukens Caroline , Dewitte Siegried , Pandelaere Mario , Warlop Luk ( 2007 ), “Wanting a Bit(e) of Everything: Extending the Valuation Effect to Variety Seeking,” Journal of Consumer Research , 34 3 , 386 – 94 . Google Scholar CrossRef Search ADS Greenleaf Eric A. , Lehmann Donald R. ( 1995 ), “Reasons for Substantial Delay in Consumer Decision Making,” Journal of Consumer Research , 22 2 , 186 – 99 . Google Scholar CrossRef Search ADS Hammond Kenneth R. , Doyle James K. ( 1991 ), Effects of Stress on Judgment and Decision Making: Part II , Research Triangle Park, NC : US Army Research Office . Hayes Andrew F. ( 2013 ), Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach , New York : Guilford . Huffman Cynthia , Kahn Barbara E. ( 1998 ), “Variety for Sale: Mass Customization or Mass Confusion?” Journal of Retailing , 74 ( Winter ), 45 – 53 . Isen Alice M. ( 1993 ), “Positive Affect and Decision Making,” in Handbook of Emotions , ed. Lewis Michael , Haviland Jeannette M. , New York : Guilford , 261 – 77 . Isen Alice M. ( 2001 ), “An Influence of Positive Affect on Decision Making in Complex Situations: Theoretical Issues with Practical Implications,” Journal of Consumer Psychology , 11 2 , 75 – 85 . Google Scholar CrossRef Search ADS Isen Alice M. , Daubman Kimberly A. , Nowicki Gary P. ( 1987 ), “Positive Affect Facilitates Creative Problem Solving,” Journal of Personality and Social Psychology , 52 ( June ), 1122 – 31 . Google Scholar CrossRef Search ADS Isen Alice M. , Geva Nehemia ( 1987 ), “The Influence of Positive Affect on Acceptable Level of Risk: The Person with a Large Canoe Has a Large Worry,” Organizational Behavior and Human Decision Processes , 39 2 , 145 – 54 . Google Scholar CrossRef Search ADS Isen Alice M. , Levin Paula F. ( 1972 ), “Effect of Feeling Good on Helping: Cookies and Kindness,” Journal of Personality and Social Psychology , 21 3 , 384 – 8 . Google Scholar CrossRef Search ADS Isen Alice M. , Means Barbara ( 1983 ), “The Influence of Positive Affect on Decision-Making Strategy,” Social Cognition , 2 1 , 18 – 31 . Google Scholar CrossRef Search ADS Isen Alice M. , Reeve Johnmarshall ( 2005 ), “The Influence of Positive Affect on Intrinsic and Extrinsic Motivation: Facilitating Enjoyment of Play, Responsible Work Behavior, and Self-Control,” Motivation and Emotion , 29 ( December ), 295 – 323 . Google Scholar CrossRef Search ADS Isen Alice M. , Rosenzweig Andrew S. , Young Mark J. ( 1991 ), “The Influence of Positive Affect on Clinical Problem Solving,” Medical Decision Making , 11 3 , 221 – 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 ( December ), 995 – 1006 . Google Scholar CrossRef Search ADS Kahn Barbara E. , Isen Alice M. ( 1993 ), “The Influence of Positive Affect on Variety Seeking among Safe, Enjoyable Products,” Journal of Consumer Research , 20 2 , 257 – 70 . Google Scholar CrossRef Search ADS Kim-Cho Eunice , Khan Uzma , Dhar Ravi ( 2013 ), “Comparing Apples to Apples or Apples to Oranges: The Role of Mental Representation in Choice Difficulty,” Journal of Marketing Research , 50 4 , 505 – 16 . Google Scholar CrossRef Search ADS Lewinsohn Shai , Mano Haim ( 1993 ), “Multi-Attribute Choice and Affect: The Influence of Naturally Occurring and Manipulated Moods on Choice Processes,” Journal of Behavioral Decision Making , 6 1 , 33 – 51 . Google Scholar CrossRef Search ADS Liberman Nira , Trope Yaacov ( 1998 ), “The Role of Desirability and Feasibility Considerations in Near and Distant Future Decisions: A Test of Temporal Construal Theory,” Journal of Personality and Social Psychology , 75 1 , 5 – 18 . Google Scholar CrossRef Search ADS 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 Luce Mary Frances , Bettman James R. , Payne John W. ( 1997 ), “Choice Processing in Emotionally Difficult Decisions,” Journal of Experimental Psychology: Learning, Memory, and Cognition , 23 ( March ), 384 – 405 . Google Scholar CrossRef Search ADS Luce Mary Frances , Payne John W. , Bettman James R. ( 1999 ), “Emotional Trade-Off Difficulty and Choice,” Journal of Marketing Research , 36 ( May ), 143 – 59 . Google Scholar CrossRef Search ADS Meloy Margaret G. ( 2000 ), “Mood-Driven Distortion of Product Information,” Journal of Consumer Research , 27 3 , 345 – 59 . Google Scholar CrossRef Search ADS Meloy Margaret G. , Russo J. Edward , Miller Elizabeth Gelfand ( 2006 ), “Monetary Incentives and Mood,” Journal of Marketing Research , 43 2 , 267 – 75 . Google Scholar CrossRef Search ADS Montgomery Henry ( 1983 ), “Decision Rules and the Search for a Dominance Structure: Towards a Process Model of Decision Making,” Advances in Psychology , 14 , 343 – 69 . Google Scholar CrossRef Search ADS Novemsky Nathan , Dhar Ravi , Schwarz Norbert , Simonson Itamar ( 2007 ), “Preference Fluency in Choice,” Journal of Marketing Research , 44 ( August ), 347 – 56 . Google Scholar CrossRef Search ADS Nowlis Stephen M. , Simonson Itamar ( 1996 ), “The Effect of New Product Features on Brand Choice,” Journal of Marketing Research , 33 ( February ), 36 – 46 . Google Scholar CrossRef Search ADS Pham Michel Tuan ( 1998 ), “Representativeness, Relevance, and the Use of Feelings in Decision Making,” Journal of Consumer Research , 25 ( September ), 144 – 59 . http://dx.doi.org/10.1111/1468-0106.00037 Google Scholar CrossRef Search ADS Pocheptsova Anastasiya , Amir On , Dhar Ravi , Baumeister Roy F. ( 2009 ), “Deciding Without Resources: Resource Depletion and Choice in Context,” Journal of Marketing Research , 46 ( June ), 344 – 55 . Google Scholar CrossRef Search ADS Pocheptsova Anastasiya , Peterson Francine , Etkin Jordan ( 2015 ), “Using Products to Pursuit Multiple Goals: The Effect of Positive Mood on Perceived Means Instrumentality,” Journal of Consumer Psychology , 25 2 , 296 – 303 . Google Scholar CrossRef Search ADS Pyone Jin Seok , Isen Alice M. ( 2011 ), “Positive Affect, Intertemporal Choice, and Levels of Thinking: Increasing Consumers’ Willingness to Wait,” Journal of Marketing Research , 48 ( June ), 532 – 43 . Google Scholar CrossRef Search ADS Ratchford Brian T. ( 1982 ), “Cost-Benefit Models for Explaining Consumer Choice and Information Seeking Behavior,” Management Science , 28 ( February ), 197 – 212 . Google Scholar CrossRef Search ADS Redelmeier Donald A. , Shafir Eldar ( 1995 ), “Medical Decision Making in Situations that Offer Multiple Alternatives,” JAMA , 273 ( January ), 302 – 5 . Google Scholar CrossRef Search ADS Schrift Rom Y. , Netzer Oded , Kivetz Ran ( 2011 ), “Complicating Choice,” Journal of Marketing Research , 48 2 , 308 – 26 . Google Scholar CrossRef Search ADS Schwarz Norbert , Bless Herbert ( 1991 ), “Happy and Mindless, but Sad and Smart? The Impact of Affective States on Analytic Reasoning,” in Emotion and Social Judgments , ed. Forgas Joseph Paul , Elmsford, New York : Pergamon Press , 55 – 71 . Schwarz Norbert , Clore Gerald L. ( 1983 ), “Mood, Misattribution, and Judgments of Well-Being: Informative and Directive Functions of Affective States,” Journal of Personality and Social Psychology , 45 ( September ), 513 – 23 . Google Scholar CrossRef Search ADS Schwarz Norbert , Clore Gerald L. ( 1996 ), “Feelings and Phenomenal Experiences,” in Social Psychology: Handbook of Basic Principles , ed. Kruglanski Arie W. , Higgins E. Tory , New York : Guilford , 385 – 407 . Sela Aner , Berger Jonah ( 2012 ), “Decision Quicksand: How Trivial Choices Suck Us In,” Journal of Consumer Research , 39 2 , 360 – 70 . Google Scholar CrossRef Search ADS Sela Aner , Berger Jonah , Liu Wendy ( 2009 ), “Variety, Vice, and Virtue: How Assortment Size Influences Option Choice,” Journal of Consumer Research , 35 6 , 941 – 51 . Google Scholar CrossRef Search ADS Sela Aner , Berger Jonah , Nardini Gia ( 2017 ), “How Tradeoffs Shrink Attribute Hierarchy,” working paper, Department of Marketing, University of Florida, Gainesville, FL 32611. Shiv Baba , Huber Joel ( 2000 ), “The Impact of Anticipating Satisfaction on Consumer Choice,” Journal of Consumer Research , 27 2 , 202 – 16 . Google Scholar CrossRef Search ADS Shugan Steven M. ( 1980 ), “The Cost of Thinking,” Journal of Consumer Research , 7 2 , 99 – 111 . Google Scholar CrossRef Search ADS Stigler George J. ( 1961 ), “The Economics of Information,” Journal of Political Economy , 69 ( June ), 213 – 25 . Google Scholar CrossRef Search ADS Townsend Claudia , Kahn Barbara E. ( 2014 ), “The Visual Preference Heuristic: The Influence of Visual versus Verbal Depiction on Assortment Processing, Perceived Variety, and Choice Overload,” Journal of Consumer Research , 40 5 , 993 – 1015 . Google Scholar CrossRef Search ADS Tversky Amos , Shafir Eldar ( 1992 ), “Choice under Conflict: The Dynamics of Deferred Decision,” Psychological Science , 3 ( November ), 358 – 61 . Google Scholar CrossRef Search ADS Watson David , Clark Lee A. , Tellegen Auke ( 1988 ), “Development and Validation of Brief Measures of Positive and Negative Affect: The PANAS Scales,” Journal of Personality and Social Psychology , 54 6 , 1063 – 70 . Google Scholar CrossRef Search ADS Wegener Duane T. , Petty Richard E. ( 1994 ), “Mood Management across Affective States: The Hedonic Contingency Hypothesis,” Journal of Personality and Social Psychology , 66 6 , 1034 – 48 . Google Scholar CrossRef Search ADS Wegener Duane T. , Petty Richard E. , Smith Stephen M. ( 1995 ), “Positive Mood Can Increase or Decrease Message Scrutiny: The Hedonic Contingency View of Mood and Message Processing,” Journal of Personality and Social Psychology , 69 1 , 5 – 15 . Google Scholar CrossRef Search ADS Weitzman Martin L. ( 1979 ), “Optimal Search for the Best Alternative,” Econometrica: Journal of the Econometric Society , 47 ( May ), 641 – 54 . Google Scholar CrossRef Search ADS Wyer Robert S. , Clore Gerald L. , Isbell Linda M. ( 1999 ), “Affect and Information Processing,” Advances in Experimental Social Psychology , 31 , 1 – 77 . Google Scholar CrossRef Search ADS Xu Jing , Jiang Zixi , Dhar Ravi ( 2013 ), “Mental Representation and Perceived Similarity: How Abstract Mindset Aids Choice from Large Assortments,” Journal of Marketing Research , 50 ( August ), 548 – 59 . 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 This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Journal of Consumer ResearchOxford University Press

Published: Dec 11, 2017

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 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

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

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off