NINA BELEI, KELLY GEYSKENS, CAROLINE GOUKENS, SURESH RAMANATHAN, and JOS LEMMINK∗ Marketers commonly assume that health claims attached to otherwise unhealthful food stimulate consumption because such claims offer justification for indulgence and reduce guilt. This article proposes a generalized theory of healthful indulgences, identifying when and why people overconsume versus regulate food intake in response to health claims. Four studies demonstrate that not all health claims are created equal. The authors suggest that the nature of the food attributes the claims emphasize—namely, functional versus hedonic—determines the extent of consumption of the indulgence. Health claims featuring functional attributes (e.g., “extra antioxidants”) trigger high levels of health-goal accessibility, which, together with simultaneously accessible indulgence goals attached to the indulgence, results in goal conflict. This conflict leads to reduced consumption of the food. In contrast, health claims featuring hedonic attributes (e.g., “low fat”) render health goals less accessible while accentuating the pleasure dimension of the food, resulting in lower goal conflict and increased consumption of the food. Implications for the food industry and public policy makers are discussed. Keywords: health claims, goal conflict, self-control, consumption
The Best of Both Worlds? Effects of Attribute-Induced Goal Conflict on Consumption of Healthful Indulgences Food companies have long recognized the value of offering snacks that tout the healthfulness of the products in a
bid to overcome consumers’ reluctance to eat unhealthful foods as they attempt to live healthy, well-balanced lives. Alarmed by approximately 1.7 billion overweight people (Deitel 2003; World Health Organization 2011), health organizations are advocating greater availability of more wholesome food alternatives, and consumers themselves are demanding food that combines “the best of both worlds”: scrumptious food with a healthful twist that takes away the guilt (Palmer 2008). Thus, it is not surprising that the food industry has increased its focus on providing consumers with a range of so-called healthful indulgences. A chocolate bar with extra antioxidants (Hershey’s), a vanilla muffin enriched with vitamins (VitaMuffins), creamy omega-3 peanut butter (Jif), low-sodium buttered popcorn (Jolly Time), reduced-sugar preserves (Smucker’s), low-fat potato chips (Lay’s), and low-fat cookies (Chips Ahoy) are only some of the many examples of indulgences claiming to provide health benefits. Sales of healthful snacks now outpace traditional snacks by a ratio of 3:1 (Palmer 2008), offering great opportunities for the food and beverage industry. Conceptually, healthful indulgences share two important characteristics. First, they represent an “improved” version of foods generally perceived as unhealthful, consumed primarily for reasons of taste, pleasure, and indulgence.
*Nina Belei is Assistant Professor of Marketing, Institute for Management Research, Radboud University Nijmegen (e-mail: n.belei@fm. ru.nl). Kelly Geyskens is Assistant Professor of Marketing, School of Business and Economics, Maastricht University (e-mail: k.geyskens@ maastrichtuniversity.nl). Caroline Goukens is Associate Professor of Marketing, School of Business and Economics, Maastricht University (e-mail:
[email protected]). Suresh Ramanathan is David R. Norcom ’73 Endowed Professor of Marketing, Mays Business School, Texas A&M University (e-mail:
[email protected]). Jos Lemmink is Canon-Océ Endowed Chair in Business Services Innovation, Professor of Marketing, Maastricht University (e-mail: j.lemmink@ maastrichtuniversity.nl). The authors are grateful to Brian Sternthal, Raj Raghunathan, Nidhi Agrawal, and Claudia Jasmand for their invaluable feedback and comments on previous versions of the article. They also acknowledge the helpful and constructive input of the three anonymous JMR reviewers during the review process. Financial support from the European Union (Marie Curie Individual Fellowship Grant 254931), Netherlands Organization for Scientific Research (Veni Individual Grant 45110-009), and Marketing Science Institute (Grant 4-1743) is gratefully acknowledged. The first three authors contributed equally and are mentioned in alphabetical order. Chris Janiszewski served as associate editor for this article.
© © 2012, American Marketing Association ISSN: 0022-2437 (print), 1547-7193 (electronic)
900
Journal of Marketing Research Vol. XLIX (December 2012), 900–909
Consumption of Healthful Indulgences Second, they carry a claim or label either explicitly or implicitly suggesting that the food is more healthful (and, thus, “better for you”) than conventional versions. Typically, such claims stress either the presence (or increased amount) of food attributes perceived as beneficial for health (e.g., vitamins, calcium, antioxidants) or the absence (or decreased amount) of attributes regarded as detrimental to health (e.g., sugar, fat, cholesterol). Prior research indicates that consumers may increase their consumption of such healthful indulgences because they are guilt reducing and, thus, easy to justify (Okada 2005; Wansink and Chandon 2006). Consumers’ tendency to overgeneralize health claims often results in misleading inferences on the amount of calories contained in the food and, thus, in overestimation of the appropriate amount to eat (Wansink 2004; Wansink and Chandon 2006). Despite the growing prevalence of healthful indulgences in the market, surprisingly little research has focused on the different types of health claims attached to indulgences and their effects on food intake. To the best of our knowledge, the only study to demonstrate the consumption effects associated with health claims is that of Wansink and Chandon (2006), who show that low-fat claims lead to increased consumption of snacks (arguably) due to guilt reduction and ease of justification. In this research, we examine whether the nature of the health claim attached to an indulgence affects consumption of the indulgence. More specifically, we compare health claims that stress a functional attribute (e.g., antioxidants, cholesterol) with health claims that stress a hedonic attribute (e.g., fat). Prior research is agnostic about the nature of the attributes that health claims emphasize. We draw on research demonstrating an inverse relationship between hedonic (e.g., fun, enjoyable) and functional (e.g., nourishing, healthful) attributes (Raghunathan, Naylor, and Hoyer 2006) and propose that seemingly similar health claims result in opposite consumption patterns depending on whether the attribute the claim emphasizes is functional or hedonic. Thus, when a consumer eats chocolate labeled as either high in antioxidants (i.e., a functional attribute) or low in fat (i.e., a hedonic attribute), we expect decreased versus increased consumption, respectively, compared with the consumption of a conventional chocolate without a health claim. Across four studies, we provide evidence that only health claims stressing functional attributes (hereinafter, “functional health claims”) help control the consumed amount by triggering high levels of health-goal accessibility. Because a hedonic indulgence goal is also salient, given the nature of the product category, stressing a functional attribute results in high conflict between the opposing health and indulgence goal. This conflict induces an aversive state, resulting in reduced consumption of the indulgence. In contrast, health claims featuring hedonic attributes (hereinafter, “hedonic health claims”) seem to direct attention to the hedonic qualities of the food, rendering a health goal temporarily less accessible. Such increased focus on the pleasure dimension of the food together with low accessibility of health goals results in low conflict between health and indulgence goals. This low level of goal conflict stimulates increased consumption of the indulgence. However, importantly, when a conflict is created externally (e.g., by means of priming), this increased consumption is blocked.
901 CONCEPTUAL BACKGROUND In their attempts to combine a healthful lifestyle with the pleasures of indulgences, consumers increasingly seek products that offer both a superior taste and health benefits. The increasing availability of healthful indulgences designed to satisfy people’s cravings for everyday treats accompanied by reduced guilt seems to provide an optimal solution. Consumers no longer need to cut back on indulgence, but instead can engage in satisfying, smartsnacking consumption behaviors; manufacturers are able to increase their market shares while providing real value to consumers; and society at large benefits from more healthful eating patterns that potentially reduce the growing problems related to overweight and obesity. Unfortunately, this promising picture painted is far from accurate. Emerging evidence suggests that healthful indulgences may not be beneficial to consumers. Research has shown that people often use information-processing heuristics that provide them with easier-to-justify decisions, allowing avoidance of potential negative emotions (Payne and Bettman 2000). The use of heuristics means that only a subset of relevant information is processed, which increases the probability of systematic decision errors. Such ease-of-justification processing has been documented especially in situations involving decisions of whether to indulge in, or refrain from, unhealthful but tempting food. For example, research has demonstrated people’s tendency to overgeneralize the benefits associated with nutrition-based claims, resulting in erroneous inferences about attributes of the product that are not substantively related to the attribute the claim actually emphasizes (Polivy et al. 1986; Schuldt and Schwarz 2010; Wansink and Chandon 2006). These inappropriate generalizations can have important implications on how consumers perceive food items. Andrews, Netemeyer, and Burton (1998) find that consumers perceive foods promoted with “no cholesterol” and “healthy” claims (e.g., margarine) as containing lower levels of fat. Schuldt and Schwarz (2010) recently demonstrate that consumers judge cookies claiming to be “organic” as containing fewer calories. As a consequence, health claims attached to inherently unhealthful, energy-dense foods might provide consumers with an implicit license to eat, which would ultimately be reflected in overconsumption of such foods. Despite general agreement in the literature that consumers tend to overgeneralize nutrient-based food claims, few studies have tested the behavioral implications of such overgeneralization. Wansink and Chandon (2006) were the first to find that low-fat claims have a countervailing effect, increasing overall consumption by up to 50%. Arguably, low-fat food induces consumers to falsely infer that the food also contains fewer calories, making it more acceptable to eat. Combining these findings with consumers’ tendency to overgeneralize other nutrition-based labels, resulting in misleading health halos (e.g., Chandon and Wansink 2007), might spark the conclusion that all health claims attached to unhealthful foods provide similar ease of justification. Thus, it might be expected that all indulgences framed as healthful stimulate food intake instead of triggering self-control. In contrast with the view that health claims function as consumption-stimulating cues, an alternative perspective is
902
JOURNAL OF MARKETING RESEARCH, DECEMBER 2012
that consumers subscribe to the “unhealthy = tasty” intuition (Raghunathan, Naylor, and Hoyer 2006). According to this view, an increasing degree of functionality (e.g., safe, healthful) is simultaneously associated with a decrease in hedonic potential (e.g., fun, pleasure), thus suggesting that hedonic and functional attributes are negatively correlated. Accordingly, the more healthful a food is portrayed, the less enjoyable it is perceived to be. This implies that consumers would perceive any inherently unhealthful food stressing health claims as relatively poorer in taste than a standard option and, thus, as less attractive. Consequently, we might predict decreased consumption of healthful indulgences. We present a more nuanced perspective on these contrasting views on when and why people overconsume versus regulate food intake in response to health claims. More specifically, we suggest that the consumed amount of indulgences framed as healthful depends on the nature of the attribute the claims emphasize. Building on the work of Raghunathan, Naylor, and Hoyer (2006), we suggest that health claims featuring functional food attributes that have a strong connotation of “health” (e.g., low cholesterol, vitamins, antioxidants, calcium) should make the concept of health highly accessible in consumers’ minds. In contrast, a health claim highlighting hedonic food attributes strongly associated with the food’s tastiness (e.g., “fat,” “sugar”; Raghunathan, Naylor, and Hoyer 2006; see also Wertenbroch 1998) should direct attention to the hedonic qualities of the food, rendering a health goal less accessible. Although the type of attribute emphasized in a health claim can result in different levels of health-goal accessibility, we expect an indulgence goal to be highly accessible when people are exposed to tempting food, regardless of the nature of the attribute stressed in the health claim. Research has shown that, in general, people categorize food according to a good/bad dichotomy, and they universally perceive tempting food, such as chocolate, ice cream, and pizza, as “bad” in terms of wholesomeness (Chernev 2011). Bad but tempting food options are likely to activate appetitive and, thus, indulgence-related goals, leading to increased desires and craving for such options (Fedoroff, Polivy, and Herman 2003; Raghunathan, Naylor, and Hoyer 2006). Therefore, we anticipate that such food will by itself trigger high levels of indulgence-related goal accessibility, regardless of the nature of the health claim present. Consequently, these differences in health- and indulgencegoal accessibility may engender different levels of goal conflict. Specifically, we expect a high level of goal conflict for functional health claims and a low level for hedonic health claims. Goal conflict is an aversive state that consumers usually try to resolve by dissociating themselves from the object causing the conflict (Ramanathan and Williams 2007); thus, we posit that functional health claims lead to decreased consumption of the food associated with such conflict. In the case of low goal conflict, no goal conflict needs to be resolved. Instead, the focus on pleasure triggered by hedonic food attributes may stimulate increased food intake compared with a food without such a claim. We tested these predictions in four experimental studies following the approach Spencer, Zanna, and Fong (2005) advocate. In Study 1, we show that seemingly similar health claims attached to unhealthful food result in different consumption patterns. Claims emphasizing a functional food
attribute (e.g., antioxidants, low cholesterol) induce a significant decrease in consumption compared with a control condition, whereas claims highlighting a hedonic food attribute (e.g., fat) trigger a significant increase in consumption. Study 2 demonstrates that functional versus hedonic food attributes elicit different levels of health-goal accessibility, whereas indulgence goals are equally accessible across conditions. These findings support our prediction of different levels of goal conflict evoked by functional versus hedonic attributes, with a high level of goal conflict in the functional health claim condition and a low level of goal conflict in the hedonic health claim condition. In Study 3, we show that the consumption-enhancing effects of hedonic health claims can be blocked if a conflict is created by an external activation of the health goal. Finally, in Study 4, we extend the findings of Study 3 to cues in the consumption context that subtly remind participants of the functional versus hedonic nature of eating, showing that cues referring to the functional nature of eating moderate the consumption-enhancing effect of hedonic health claims. STUDY 1 The objective of Study 1 was to test our hypothesis that health claims differing in the nature of the attributes emphasized (i.e., hedonic vs. functional) have differential effects on foods’ consumption. We expected functional health claims attached to indulgences to trigger reduced consumption of the foods bearing these claims, whereas we expected hedonic health claims attached to indulgences to stimulate increased consumption. Design and Procedure We randomly assigned 109 undergraduate students to one of three conditions. In the functional-attribute condition, participants evaluated a chocolate product package that featured the following health-related claim: “Antioxidants—Health from the cacao bean.” In the hedonicattribute condition, participants evaluated a chocolate package with a low-fat label. In the control condition, participants evaluated a version of the package without health-related cues. We selected “antioxidants” as a representative of a functional attribute because the claim is increasingly prominent on various food packages and is known for its clear functional rather than hedonic benefits on health. We selected “fat” as a representative of a hedonic attribute for our health claim manipulation because of the preeminence of low-fat health claims in the food market and because of the word’s association with taste and food enjoyment (Raghunathan, Naylor, and Hoyer 2006). A pretest (N = 36) showed that the low-fat claim reminded participants of the hedonic aspects of eating rather than the importance of being healthy (t(18) = 3037, p < 001). The opposite was true for the antioxidant claim: This claim reminded participants of the importance of being healthy rather than the hedonic aspects of eating (t(16) = −2027, p < 005). Furthermore, both health claims were perceived as equally believable (t(34) = 090, not significant [n.s.]), equally attractive (t(34) = 042, n.s.), and equally convincing (t(34) = 105, n.s.). All participants completed a battery of 15 questions pertaining to the attractiveness of the chocolate package (e.g., “The design of the package’s front side is appealing,”
Consumption of Healthful Indulgences “The information on the package’s front side is credible”). We then invited the participants to taste the chocolate they had seen on the product package. To increase the variance in food intake, we used chocolate pearls (approximately 1 g each) instead of the brand’s regular seashell shapes (approximately 11 g each). We informed participants that presenting relatively small portions of the food in such taste tests was common so the company had provided chocolate pearls, which were the same as the seashells, just smaller and more convenient to test. We placed two transparent bowls of chocolate (one with dark chocolate pearls [700 g] and one with milk chocolate pearls [700 g]) in front of the participants. By providing them with two types of chocolate, we increased the chance that most participants would like the test product. Participants evaluated the product on several seven-point (“agree/disagree”) items related to taste, color, and structure. Results After the taste test and unbeknownst to the participants, the experimenter weighed how many grams of the chocolate pearls each participant had consumed. We created a summated variable for the consumption of the milk and dark chocolate pearls. An analysis of variance (ANOVA) of the health claim manipulation on the quantity consumed revealed a significant main effect (F(2, 106) = 11037, p < 0001). In line with our expectations and previous research on the effects of low-fat claims (Wansink and Chandon 2006), participants in the hedonic-attribute condition consumed significantly more chocolate than participants in the control condition (Mlow-fat = 28031, SD = 14049; Mcontrol = 20046, SD = 16052; F(1, 106) = 6027, p < 001). More important, compared with the control condition, participants in the functional-attribute condition consumed significantly fewer chocolate pearls (Mantioxidants = 13045, SD = 7061; F(1, 106) = 4043, p < 005). It could be argued that the antioxidant claim, which made the chocolate seem more healthful, also made the chocolate seem less tasty (Raghunathan, Naylor, and Hoyer 2006). However, if so, this effect should also have held for the low-fat claim, which was not borne out by the increased consumption in this case. To test whether the health claims influenced the taste perceptions of the chocolate, we conducted an additional ANOVA, which revealed no significant differences in the perceived taste of the chocolates across the three conditions (F(2, 106) < 1). Furthermore, the ANOVA revealed no differences in people’s intention to purchase the chocolate across conditions (F(2, 106) < 1). Discussion The results of Study 1 provide initial support for our hypothesized effects of health claims on actual food consumption. Although all health claims attached to unhealthful foods may be viewed as easy to justify, our results show that ostensibly similar health claims can lead to different consumption patterns: A functional health claim decreases consumption whereas a hedonic health claim increases consumption compared with a control condition. Because the low-fat claim highlights the reduced amount of a hedonic attribute (i.e., fat), whereas the antioxidants claim stresses the increased amount of a functional attribute (i.e., antioxidants), it could be argued that this difference in framing stimulated decreased versus increased food
903 intake. Therefore, we ran a replication study (N = 97) with an additional condition using a claim that stressed the reduced amount of a functional attribute (i.e., low cholesterol). The results revealed a significant main effect of the health claim manipulation (F(3, 93) = 9085, p < 0001) and showed that compared with the control condition, consumption increased when we exposed participants to the low-fat health claim (Mlow-fat = 31028, SD = 27025; Mcontrol = 21082, SD = 10044; F(1, 93) = 4056, p < 005) and decreased when we exposed them to the antioxidants claim (Mantioxidants = 11053, SD = 7091; F(1, 93) = 5061, p < 005). Furthermore, exposure to the low-cholesterol claim decreased consumption of the chocolate to the same extent as the antioxidant claim (Mlow-cholesterol = 11002, SD = 5055; F(1, 93) = 6029, p < 001), whereas the difference between the antioxidant and the low-cholesterol claim was not significant (F(1, 93) < 1). These findings rule out a framing explanation as well as the related idea of loss aversion, which would suggest that the absence of hedonic attributes is more desirable than the presence of functional attributes. STUDY 2 The objective of Study 2 was to provide insight into the underlying processes involved with different consumption patterns. In particular, we wanted to test our hypothesis that functional attributes in a health claim result in higher levels of health-goal activation than hedonic attributes, while having no differential effects on indulgence-goal activation. We used a lexical decision task in which faster recognition of target words (health and indulgence) signified the activation of goals (Fishbach, Friedman, and Kruglanski 2003; Geyskens et al. 2008). Design and Procedure We randomly assigned 63 students to one of three conditions (“antioxidants,” “low-fat,” or “control”) in which they performed the same package evaluation task as in Study 1. After removing all materials, we prepared participants for the lexical decision task. After a two-second warning screen, including a fixation cross, a stimulus word appeared on the computer screen in front of the participants. They responded as quickly and accurately as possible by pressing a key to indicate whether the stimulus was a word (left “Shift” key) or a nonword (right “Shift” key). Response times in milliseconds and accuracy were recorded. After a trial round of ten trials (5 neutral and 5 pseudowords), participants saw 66 letter strings presented randomly: 34 nonwords, 16 neutral words unrelated to health or indulgence, 8 target health-related words (e.g., “healthful,” “vitamins”), and 8 indulgence-related words (e.g., “desire,” “wanting”). We counterbalanced the order of the health and indulgence words between participants to control for possible spillover effects. Results To prepare the data for analysis, we used only correct responses (Fazio 1990) and removed response latencies greater than or lower than three standard deviations from the means (8%). We averaged response times to the target and neutral words and log-transformed the summated variables (for ease of interpretation, however, we report the means of the response times in milliseconds). We created
JOURNAL OF MARKETING RESEARCH, DECEMBER 2012
a dummy variable indicative of the order in which the target words appeared; however, testing for order showed no significant effects, and therefore we dropped this variable from further analyses. To test the effects of the antioxidants and low-fat health claim manipulation on response times to the health and indulgence words, we conducted a reaction time (within; neutral, health, indulgence) × health claim (between; control, low-fat, antioxidants) mixed ANOVA. The results revealed a significant interaction effect (F(2, 60) = 3070, p < 005). Planned contrasts show that in line with our predictions on indulgence-goal accessibility, no differences in reaction times to indulgence words occurred across conditions (F(2, 60) < 1), suggesting similar accessible indulgence goals in the functional health claim condition (M = 633096), the control condition (M = 638095), and the hedonic health claim condition (M = 658006). No differences occurred for response times to the neutral words across conditions (F(2, 60) < 1). However, consistent with our hypothesis on health-goal accessibility, planned contrasts show that the antioxidants claim made health-related words relatively more accessible than the low-fat claim (Mantioxidants = 638072, SD = 73079; Mlow-fat = 703025, SD = 102029; F(1, 60) = 10085, p < 001), while a marginally significant difference occurred in comparison with the control condition (Mcontrol = 662035, SD = 80038; F(1, 60) = 3010, p < 010). This suggests that the health goal was facilitated in the antioxidants claim condition. In the antioxidants claim condition, the health-related words (F(1, 60) = 3051, p < 007) and indulgence-related words (F(1, 60) = 4089, p < 005) were activated equally, compared with the neutral words, indicating experienced goal conflict in this functional health claim condition. Although activation of the health goal did not differ between the low-fat claim and the control condition (F(1, 60) = 1010, n.s.), further analyses revealed that the ordering (functional health claim < control < hedonic health claim) was significant, as indicated by a significant linear trend of health-goal accessibility across conditions (F(1, 60) = 5071, p < 005). No quadratic trend emerged (F(1, 60) < 1). In the low-fat claim condition, the health goal was significantly less activated than the neutral words (F(1, 60) = 5026, p < 005) and the indulgence goal (F(1, 60) = 10091, p < 001). This suggests that the health goal was inhibited, and thus participants experienced little to no goal conflict. Figure 1 presents these results. Discussion The results of the lexical decision task support our hypothesis that attributes emphasized in health claims differing in their functional versus hedonic nature result in different levels of health-goal accessibility. Although we find similar indulgence-goal accessibilities across conditions, only the health claim with the functional, health-related attribute (i.e., antioxidants) simultaneously triggered high levels of health-goal accessibility, suggesting high conflict between health and indulgence goals. In contrast, health claims featuring hedonic attributes (i.e., fat) entail a much lower level of goal conflict than the functional claim condition. We posit that the low accessibility of the health goal in the hedonic health claim condition is the result of active management of the two opposing goals, with the hedonic
Figure 1 STUDY 2: REACTION TIMES TO HEALTH AND INDULGENCE WORDS ACROSS CONDITIONS
Lexical Decision Time (in Milliseconds)
904
720 700
Neutral words Health words Indulgence words
680 660 640 620 600 580 Antioxidants
Control
Low Fat
attribute leading to an enhanced focus on the pleasure associated with consuming the food and a concomitant decrease in accessibility of the health goal (Grant, Malaviya, and Sternthal 2004; Laran and Janiszewski 2009; Liberman, Förster, and Higgins 2007). STUDY 3 Thus far, we have demonstrated that functional health claims lead to decreased consumption (Study 1) and create conflict by the simultaneous activation of a health and hedonic indulgence goal (Study 2). In contrast, hedonic health claims lead to increased consumption (Study 1), with low levels of health-goal accessibility easing the way for a sole focus on the pleasure associated with consuming the food (Study 2); consequently, no conflict arises for hedonic health claims. A logical next step would be to conduct a study that measures both goal activation and consumption and to investigate whether the behavioral effects are driven by this differential goal activation. However, as Spencer, Zanna, and Fong (2005) argue, measuring the activation of the underlying concept can prime the concept for the participants and thus prevent the observation of the behavioral effect. Therefore, we opted to manipulate, instead of measure, the level of health/hedonic goal activation. In line with our argumentation, it could be argued that the consumption increase in the hedonic-attribute condition would be blocked if a conflict is created by an explicit external activation of the health goal. In this study, we therefore examined how the activation of a health versus indulgence goal before the claim exposure influences our previous results. We applied a priming technique, resulting in three conditions: a control condition without health- or indulgence-goal activation (neutral prime), a condition with health-goal activation (health prime), and a condition with indulgence-goal activation (indulgence prime). The aim of priming is to activate a concept in long-term memory (usually below the awareness threshold) by exposing participants to words related to that concept (Bargh and Chartrand 1999). With this priming technique, we simulated the activation of the health/indulgence goal.
Consumption of Healthful Indulgences One hundred forty-nine undergraduate students participated in partial fulfillment of a course requirement and were randomly assigned to one of the six cells of a 2 (claim: hedonic vs. functional) × 3 (prime: neutral vs. health vs. indulgence) design. Participants first received a “language test” that primed them with neutral, healthrelated, or indulgence-related words. The language test was a scrambled-sentences task (Bargh and Chartrand 2000). Each sentence consisted of five words, and participants were instructed to construct a grammatically correct fourword sentence. In the health-goal-activation condition, 8 of the 20 sentences in the test contained a word related to health (e.g., “vitamins,” “healthful”). In the indulgencegoal-activation condition, these sentences contained a word related to indulgence (e.g., “indulge,” “pleasure”). We chose the words to ensure that the health-/indulgencerelated words would be used in the sentence composition. The neutral prime condition replaced the goal-related words with neutral words. The participants then evaluated a chocolate product package, identical to the one in Study 1. This package featured a claim highlighting the presence of antioxidants (functionalattribute condition), a claim focusing on the product’s low-fat content (hedonic-attribute condition), or no claim (control condition). All participants completed a battery of 15 questions pertaining to the attractiveness of the chocolate package. Similar to Study 1, we then invited all participants to taste the chocolate they had seen on the product package. Again, we used the chocolate pearls instead of the brand’s regular seashell shapes. We placed one transparent bowl filled with a mixture of milk and dark chocolate pearls (700 g) in front of the participants and asked them to evaluate the product on several seven-point (“agree/disagree”) items related to taste, color, and structure. Results After the taste test and unbeknownst to the participants, the experimenter weighed how many grams of the chocolate pearls each participant had consumed. An ANOVA testing the effects of the prime (neutral vs. health vs. indulgence) and health claim (functional vs. hedonic attribute) on the consumed quantity of chocolate pearls (in grams) revealed a significant main effect of the health claim (F(1, 143) = 10059, p < 001). Furthermore, a planned interaction contrast marginally supported our prediction that differences in consumption between functional and hedonic health claims would be substantially different in the health prime condition compared with the neutral and indulgence prime conditions (F(1, 143) = 3054, p < 007). Replicating the results of Study 1, participants in the neutral prime condition consumed more chocolate pearls in the hedonicattribute claim condition (Mlow-fat = 26063, SD = 2306) than in the functional-attribute claim condition (Mantioxidants = 14028, SD = 11011; F(1, 143) = 5092, p < 005). Participants primed with the indulgence goal showed the same consumption pattern, with participants in the hedonic-attribute claim condition (Mlow-fat = 26092, SD = 22012) consuming more chocolate pearls than those in the functional-attribute claim condition (Mantioxidants = 14037, SD = 12012; F(1, 143) = 8007, p < 001). In line with our prediction, the consumption increase in the hedonic-attribute claim condition
was blocked when participants were primed with health. More specifically, participants in the health prime conditions ate an equal amount in the hedonic-attribute condition (Mlow-fat = 15070, SD = 13036) as in the functional-attribute condition (Mantioxidants = 14024, SD = 13015; F411 1435 < 1). Figure 2 presents these findings. Discussion The findings of Study 3, in which we manipulated goal activation, are consistent with our reasoning that food situations triggering a conflict between indulgence and health goals help consumers control their food intake. Priming people with an indulgence goal does not alter the previously found effect of health claims on consumption. Unhealthful food by itself triggers high levels of indulgence-related goal accessibility (Fedoroff, Polivy, and Herman 2003), so the prime has only limited effect. Conversely, priming people with a health goal changes the previously found consumption-increasing effect of hedonic claims. Hedonic claims do not trigger a health goal, and therefore no conflict with the indulgence goal arises. However, when a health goal is primed, a conflict arises such that a similar decrease in consumption to that in the case of functional health claims occurs. These findings provide additional evidence for our reasoning that the previously found consumption-inhibiting effects of functional health claims can be explained by the conflict caused by the simultaneous activation of an indulgence goal and a health goal. Furthermore, the results show that consumptionenhancing effects of hedonic health claims can be blocked by creating a conflict externally. This implies that any cue in the marketing environment that reminds consumers of health concerns can protect them from the consumptionenhancing effects of hedonic health claims. Therefore, in Study 4, we tested this hypothesis and exposed participants to subtle health/hedonic environmental cues. In addition, we used another product category (i.e., nuts) and another functional health claim (i.e., “enriched omega-3”) to generalize our findings. Figure 2 STUDY 3: THE INTERACTION OF HEALTH CLAIM WITH PRIME ON CONSUMED AMOUNT OF CHOCOLATE
Chocolate Consumption (Grams)
Design and Procedure
905
30 Antioxidants Low fat
25 20 15 10 5 0 Neutral Prime
Health Prime
Indulgence Prime
906
JOURNAL OF MARKETING RESEARCH, DECEMBER 2012 STUDY 4
Figure 3
In Study 4, we exposed participants to subtle cues that referred to either the functional nature or the hedonic nature of eating. In line with the findings of Study 3, we expected that subtle cues (i.e., other than explicit health claims) referring to the functional nature of eating would moderate the consumption-enhancing effect of hedonic health claims because they create a conflict with the hedonic goal triggered by the claim.
STUDY 4: THE INTERACTION OF HEALTH CLAIM WITH CONSUMPTION CONTEXT ON CONSUMED AMOUNT OF NUTS
Ninety-eight undergraduate students participated in partial fulfillment of a course requirement. Participants were randomly assigned to one of the four cells of a 2 (health claim: hedonic vs. functional) × 2 (consumption context: hedonic [i.e., tasty] vs. functional [i.e., healthful]) design. We told the participants that they would take part in a study intended to evaluate the taste of a new brand of roasted nuts. However, the cover page of the booklet providing the instruction for the taste test differed between the two consumption context conditions. In the functional consumption context, the cover page featured the slogan “IF YOU NEED A HEALTHFUL BREAK.” In the hedonic consumption context, the cover page featured the slogan “IF YOU NEED A TASTY BREAK.” Each participant then received a nontransparent bowl filled with 400 grams of roasted nuts. In the hedonic health claim condition, a “low-fat” label was affixed to the bowl. In the functional health claim condition, a label reading “enriched omega-3” was affixed. A pretest (N = 36) confirmed the functional nature of the omega-3 claim: This claim reminded participants of the importance of being healthy rather than the hedonic aspects of eating (t(18) = −2017, p < 005). Furthermore, participants perceived both health claims as equally believable (t(36) = 072, n.s.), equally attractive (t(36) = 026, n.s.), and equally convincing (t(36) = 1029, n.s.). Participants then evaluated the product on several seven-point (“agree/disagree”) items related to taste, color, and structure. Results After the taste test and unbeknownst to the participants, the experimenter weighed how many grams of the nuts each participant had consumed. An ANOVA testing the effects of the health claim (hedonic vs. functional) and the consumption context (hedonic vs. functional) on the consumed quantity of nuts (in grams) revealed a significant main effect of the health claim (Mlow-fat = 12018, SD = 9083; Momega-3 = 5099, SD = 3087; F(1, 94) = 18063, p < 001) and a marginally significant main effect of consumption context (Mhealthful = 7097, SD = 6081; Mtasty = 10025, SD = 9010; F(1, 94) = 2097, p < 01). More important, however, these effects were qualified by a significant interaction effect (F(1, 94) = 4011, p < 005). As Figure 3 depicts, participants in the hedonic consumption context consumed more low-fat nuts (Mlow-fat = 15011, SD = 10076) than enriched omega-3 nuts (Momega-3 = 5077, SD = 3052), whereas participants in the functional consumption context consumed the same amount of low-fat nuts (Mlow-fat = 9060, SD = 8029) as enriched omega-3 nuts (Momega-3 = 6022, SD = 4026; F(1, 94) = 2067, n.s.). Thus, in
Nuts Consumption (Grams)
Design and Procedure
16 14
Enriched omega-3 Low fat
12 10 8 6 4 2 0 Healthy
Tasty
the functional, health-related consumption context, participants consumed less of the low-fat nuts (Mlow-fat = 9060, SD = 8029) than in the hedonic consumption context (Mlow-fat = 15011, SD = 10076; F(1, 94) = 7002, p < 001). Discussion The results of Study 4, in which we manipulated the consumption context, were highly supportive of the goal conflict explanation as suggested by the findings of the previous studies. Compared with participants who experienced the taste test in a hedonic consumption context, those who experienced the taste test in a functional consumption context were better able to refrain from indulging in the low-fat food. These results imply that promoting low-fat products explicitly as a healthful alternative might backfire for marketers because any health cue present in the environment introduces a conflict leading to decreased consumption amounts. GENERAL DISCUSSION Healthful indulgences seem to provide a solution to consumers’ attempts to combine a healthy lifestyle with the pleasures of indulgences. However, recent research has revealed a backfire effect of presenting inherently unhealthful food as more healthful, as consumers tend to overgeneralize the benefits associated with such foods, leading them to consume significantly more of the healthful indulgence (e.g., low-fat M&Ms) than they would of the conventional version (e.g., regular M&Ms). The current research examined whether the consumption-increasing effects established for low-fat health claims (Wansink and Chandon 2006) can be generalized to other health claims as well. Our results provide strong and converging evidence that not all health claims are created equal, in that they lead to opposite consumption patterns depending on whether the food attributes the claims stress are of a hedonic or a functional nature. Study 1 showed that health claims emphasizing a hedonic food attribute induce a significant increase in consumption compared with a control condition, whereas claims
Consumption of Healthful Indulgences highlighting a functional food attribute result in a significant decrease in consumption. The mechanism underlying these findings seems to be one based on differences in health-goal accessibility following exposure to different types of food attributes, which, together with highly accessible indulgence goals triggered by the hedonic nature of the food itself, result in different levels of goal conflict (Study 2). Specifically, we find that claims featuring hedonic attributes activate a health goal to a much lower extent than claims featuring functional attributes, whereas no differences occur in indulgence-goal activation across all conditions. The differences in health-goal and indulgence-goal accessibility suggest that functional attributes stressed in health claims create conflict between the competing goals, which is much less apparent for health claims featuring hedonic food attributes. Studies 3 and 4 further support this goal conflict explanation. Indeed, the consumptionenhancing effects of hedonic health claims can be blocked by externally creating a conflict by means of a goal priming technique (Study 3) and subtle health cues in the consumption environment (Study 4). Importantly, Study 4 shows that any cue in the marketing environment that raises health concerns can create an internal conflict and thus decreases consumption behavior of indulgences. This means that manufacturers should be aware not only of their own label strategies but also of other manufacturers’ strategies. That is, if a competing indulgent food brand claims to be more functional in nature, the same conflict is created by which we would predict similar inhibiting effects on consumption. We tested this hypothesis in a follow-up study (N = 52), in which we applied the same procedure as in Study 1: Participants were exposed to a 7UP advertisement that featured either a functional health claim (“7UP Cherrie Antioxidants—Healthy Happy Life”) or a neutral claim (“7UP Cherrie—Happy Life”). Afterward, when we asked participants to test another regular soft drink without a health claim, they drank less after being exposed to the functional health claim (M = 98010) than after being exposed to the neutral claim (M = 142001; F(1, 50) = 4021, p < 005). This finding further supports the generalizability of our findings and the strength of the effects of health cues on consumption of indulgences. The findings also suggest that indulgences featuring functional health claims can prime a generalized health orientation that helps consumers control consumption of any other indulgence. Theoretically, these differences in goal activation between hedonic and functional health claims are in line with research documenting primacy of affirmation (Grant, Malaviya, and Sternthal 2004), according to which affirmations (e.g., the attributes emphasized in a claim) rather than their tags (e.g., “low,” “with,” “extra”) are processed. More specifically, when product benefits are expressed as negations (e.g., “not difficult to use,” “not fattening”; where “not” is the negator tag and “difficult to use” or “fattening” are affirmations), consumers’ initial and automatic responses tend to consider primarily the affirmation. That is, people tend to elaborate on the affirmation (i.e., “fat”) and not on the negator (e.g., “low”), a finding that has been documented in various streams of research (for a review, see Grant, Malaviya, and Sternthal 2004). Our findings support this view by showing that people seem to elaborate on the affirmation (e.g., “cholesterol”) rather than on the
907 negator (e.g., “low”), thus making the functional versus hedonic nature of the attribute the key determinant for how much of the food they eventually consume. Following this line of reasoning, we would predict that the consumptionincreasing effects of low-fat claims generalize to all hedonic claims, even to hedonic claims that are more difficult rather than easier to justify. To test this prediction, we conducted another follow-up study (N = 65), in which we asked participants to take part in a taste test of pudding. As in the previous studies, we presented them with a bowl of pudding (400 g) and asked them to evaluate the pudding on a few dimensions. Again, we manipulated the label affixed to the bowl: regular versus high fat. Note that for the latter, the nature of the attribute is hedonic, while the claim itself is negative because of the tag “high.” Counterintuitively, participants consumed more of the high-fat pudding (M = 107086 g) than the regular pudding (M = 75027 g; F(1, 63) = 4002, p < 005). These data offer preliminary support for our reasoning that the nature of the attribute (hedonic vs. functional) is an important factor regarding how much a person consumes. Our findings add to existing research in several important ways. The conceptual framework this research proposes calls into question the conjecture that the semantic representation of health claims automatically results in overgeneralizations of the benefits of the unhealthful foods to which they are attached. We identified the nature of the attributes a claim emphasizes as a key determinant for whether consumption of the food carrying the claim increases or decreases compared with a control group. Thus, our research answers the call for more in-depth research on different types of nutrition claims and their effects on food intake (Wansink and Chandon 2006) by identifying the conditions under which consumption of tempting foods carrying different types of claims increases or decreases (e.g., Finkelstein and Fishbach 2010). Moreover, our research contributes to a deeper understanding of the underlying processes involved in competing goals, especially under conditions of self-control conflict in the eating domain, in which indulgence goals potentially compete with health goals. As we argued previously, prior research provides conflicting predictions about whether exposure to healthful food labels activates health goals (Shah and Kruglanski 2003) or whether it satisfies or inhibits such goals (Liberman, Förster, and Higgins 2007; Wilcox et al. 2009). Our framework introduces a previously neglected factor—that is, the nature of the attribute emphasized in a health claim—as a potential determinant for the direction of influence. Furthermore, our findings extend the ease-of-justification view by unpacking the meaning of ease. Although we question the assumption that there should be increased consumption of all healthful indulgences because they are all equally easy to justify compared with conventional indulgences, we show that ease does occur for hedonic claims, such as low fat, because they stress indulgence without activating the health goal. Accordingly, conflict between these two competing goals is low, resulting in limited selfregulation and increased consumption. Note, however, that ease in this case is the specific result of low levels of experienced goal conflict rather than enhanced elaboration processes leading to justification simplifications. In contrast,
908
JOURNAL OF MARKETING RESEARCH, DECEMBER 2012
functional health claims trigger health associations coexperienced with the hedonic indulgence goal. These two competing goals establish a situation of high conflict and result in self-regulation manifested by reduced consumption of the indulgence causing the conflict. Even more important, we show that the consumption resulting from the low conflict between the health and indulgence goal for the hedonic claim can be reduced when an external activation of the health goal leads to increased conflict between the two competing goals. We thus offer a generalized theory of healthful indulgences, which underscores the importance of studying the complex dynamics underlying self-control dilemmas (Ramanathan and Hofmann 2011) and emphasize important implications, which we discuss subsequently. On a more general note, the portrayed effects of functional claims on consumption of hedonic products cannot be readily explained by prior research. That is, research on consumers’ evaluation of food options that represent conflicting health and indulgence goals has focused on calorie estimation and consumption intentions rather than on real consumption behavior. For example, Chernev and Gal (2010) find that people believe that a meal comprising a hamburger and a salad has 100 calories less than when they evaluate the hamburger alone. Likewise, Chandon and Wansink (2007) find that people perceive Subway meals as containing 21.3% fewer calories than same-calorie McDonald’s meals. These findings imply that perceptual biases in deriving quantitative estimates might be of a more cognitive nature, while consumption is more routed in motivational processes. Indeed, while our effects on consumption are blocked by external goal-activation processes, quantitative estimates disappear when consumers consciously consider arguments contradicting the health claim. Thus, parallel processes might play a differential role after consumers are exposed to health claims. This observation highlights the importance of understanding the effects of health claims by examining their impact on both purchase and consumption. That is, the decreasing consumption effects of functional health claims might be mitigated by more consumption occasions or by more consumers. Further research should investigate these parallel effects. In addition, further research could test the generalizability of our results. That is, further examining the conflict created by the product claim and the goal the consumer envisions would be worthwhile. In this research, we focused on food products that are consumed for reasons of fun and pleasure. We found that any claim that reminds consumers of health (i.e., a goal conflicting with the nature of the consumption situation) creates a conflict, resulting in consumption-decreasing effects. Therefore, research could investigate whether a similar conflict is created when a product is consumed for reasons of health while carrying a claim referring to hedonics. For example, would a similar conflict be created when a product such as vitamin water claims to have a delicious taste? Although prior research has paid much attention to the relative weight that consumers place on functional- versus hedonic-attribute dimensions (Chitturi, Raghunathan, and Mahajan 2007; Dhar and Wertenbroch 2000; Okada 2005), our research shows that functionality and hedonics may not be as easy to combine as previously believed. Further research on this interplay between functionality and hedonics would certainly be worthwhile.
Our findings have broad relevance to the food industry and suggest that the nature of attributes emphasized in health claims can have significant effects on consumption. In particular, our findings imply important negative repercussions of the popular strategy of offering more healthful alternatives of inherently unhealthful foods when the attributes the claim stresses are of a functional nature. Functional food attributes emphasized in health claims enhance consumers’ self-control of food intake, which reverses the desired consumption increase food manufacturers seek. In contrast, and in line with previous findings (i.e., Wansink and Chandon 2006), exposure to healthful indulgences featuring hedonic food attributes (e.g., lower levels of fat) seems to lead to increased consumption. Therefore, food manufacturers might benefit from stressing hedonic food attributes. However, our findings also imply that promoting low-fat products explicitly as a healthful alternative might backfire because any health cue present in the environment introduces a conflict leading to decreased consumption amounts. For public policy makers, our findings provide some comfort that not all marketing efforts intended to promote more wholesome alternatives of unhealthful, tempting food necessarily have negative consequences for consumers’ eating patterns. Although we replicate the detrimental effects of low-fat references on food intake, we also demonstrate the power of health claims featuring functional food attributes in triggering more controlled consumption, ultimately leading to more health-conscious eating behaviors for unhealthful foods. Reframing hedonic claims in functional terms (e.g., low-fat foods described in terms of cholesterol or hydrogenated oils) helps consumers control their consumption. Moreover, our findings imply that consumers might be better off if they are surrounded by cues referring to healthfulness even when they consume indulgences with hedonic health claims or regular indulgences. As Studies 3 and 4 show, these cues to the health goal trigger a conflict with the activated indulgence goal and help consumers control their consumption. REFERENCES Andrews, Craig J., Richard G. Netemeyer, and Scot Burton (1998), “Consumer Generalization of Nutrient Content Claims in Advertising,” Journal of Marketing, 62 (October), 62–75. Bargh, John A. and Tanya L. Chartrand (1999), “The Unbearable Automaticity of Being,” American Psychologist, 54 (7), 462–79. and (2000), “The Mind in the Middle: A Practical Guide to Priming and Automaticity Research,” in Handbook of Research Methods in Social and Personality Psychology, Harry T. Reis and Charles M. Judd, eds. New York: Cambridge University Press, 253–85. Chandon, Pierre and Brian Wansink (2007), “The Biasing Health Halos of Fast-Food Restaurant Health Claims: Lower Calorie Estimates and Higher Side-Dish Consumption Intentions,” Journal of Consumer Research, 34 (October), 301–314. Chernev, Alexander (2011), “Semantic Anchoring in Sequential Evaluations of Vices and Virtues,” Journal of Consumer Research, 37 (February), 761–74. and David Gal (2010), “Categorization Effects in Value Judgments: Averaging Bias in Evaluating Combinations of Vices and Virtues,” Journal of Marketing Research, 47 (August), 738–47.
Consumption of Healthful Indulgences Chitturi, Ravindra, Rajagopal Raghunathan, and Vijay Mahajan (2007), “Form Versus Function: How the Intensities of Specific Emotions Evoked in Functional Versus Hedonic Trade-Offs Mediate Product Preferences,” Journal of Marketing Research, 44 (November), 702–714. Deitel, Mervyn (2003), “Overweight and Obesity Worldwide Now Estimated to Involve 1.7 Billion People,” Obesity Surgery, 13 (3), 329–30. Dhar, Ravi and Klaus Wertenbroch (2000), “Consumer Choice Between Hedonic and Utilitarian Goods,” Journal of Marketing Research, 37 (February), 60–71. Fazio, Russell H. (1990), “A Practical Guide to the Use of Response Latencies in Social Psychological Research,” in Review of Personality and Social Psychology, Vol. 11, C. Hendrick and M.S. Clark, eds. Newbury Park, CA: Sage Publications, 74–97. Fedoroff, Ingrid, Janet Polivy, and Peter C. Herman (2003), “The Specificity of Restraint Versus Unrestraint Eaters’ Responses to Food Cues: General Desire to Eat, or Craving for the Cued Food?” Appetite, 41 (1), 7–13. Finkelstein, Stacey R. and Ayelet Fishbach (2010), “When Healthy Food Makes You Hungry,” Journal of Consumer Research, 37 (3), 357–67. Fishbach, Ayelet, Ronald S. Friedman, and Arie W. Kruglanski (2003), “Leading Us Not into Temptation: Momentary Allurements Elicit Overriding Goal Activation,” Journal of Personality and Social Psychology, 84 (2), 296–309. Geyskens, Kelly, Siegfried Dewitte, Mario Pandelaere, and Luk Warlop (2008), “Tempt Me Just a Little Bit More: The Effects of Prior Food Temptation Actionability on Goal Activation and Consumption,” Journal of Consumer Research, 35 (December), 600–610. Grant, Jung Susan, Prashant Malaviya, and Brian Sternthal (2004), “The Influence of Negation on Product Evaluations,” Journal of Consumer Research, 31 (3), 583–91. Laran, Julian and Chris Janiszewski (2009), “Behavioral Consistency and Inconsistency in the Resolution of Goal Conflict,” Journal of Consumer Research, 35 (April), 967–84. Liberman, Nira, Jens Förster, and E. Tory Higgins (2007), “Completed Versus Interrupted Priming: Reduced Accessibility from Post-Fulfillment Inhibition,” Journal of Experimental Social Psychology, 43 (March), 258–64. Okada, Erica Mina (2005), “Justification Effects on Consumer Choice of Hedonic and Utilitarian Goods,” Journal of Marketing Research, 42 (February), 43–53. Palmer, Sharon (2008), “Healthy Indulgence: A Best of Both Worlds Approach to Eating,”Today’s Dietitian, 10 (September), 62.
909 Payne, John W. and James R. Bettman (2000), “Preferential Choice and Adaptive Strategy Use,” in Bounded Rationality: The Adaptive Toolbox, Gerd Gigerenzer and Reinhard Selten, eds. Cambridge, MA: MIT Press, 123–46. Polivy, Janet, C. Peter Herman, Rick Hackett, and Irka Kuleshnyk (1986), “The Effects of Self-Attention and Public Attention on Eating in Restrained and Unrestrained Subjects,” Journal of Personality and Social Psychology, 50 (6), 1253–60. Raghunathan, Rajagopal, Rebecca Walker Naylor, and Wayne D. Hoyer (2006), “The Unhealthy = Tasty Intuition and Its Effects on Taste Inferences, Enjoyment, and Choice of Food Products,” Journal of Marketing, 70 (October), 170–84. Ramanathan, Suresh and Wilhelm Hofmann (2011) “Sticky Desires or Tricky Self-Control? Dynamics of Attentional and Affective Processes in Response to Temptation,” working paper, Mays Business School, Texas A&M University. and Patti Williams (2007), “Immediate and Delayed Emotional Consequences of Indulgence: The Moderating Influence of Personality Type on Mixed Emotions,”Journal of Consumer Research, 34 (August), 212–23. Schuldt, Jonathan P. and Norbert Schwarz (2010), “The ‘Organic’ Path to Obesity? Organic Claims Influence Calorie Judgments and Exercise Recommendations,” Judgment and Decision Making, 5 (June), 144–50. Shah, James Y. and Arie W. Kruglanski (2003), “When Opportunity Knocks: Bottom-Up Priming of Goals by Means and Its Effects on Self-Regulation,” Journal of Personality and Social Psychology, 84 (June), 1109–22. Spencer, Steven J., Mark P. Zanna, and Geoffrey T. Fong (2005), “Establishing a Causal Chain: Why Experiments Are Often More Effective Than Mediation Analyses in Examining Psychological Processes,” Journal of Personality and Social Psychology, 89 (6), 845–51. Wansink, Brian (2004), “Environmental Factors That Increase the Food Intake and Consumption Volume of Unknowing Consumers,” Annual Review of Nutrition, 24, 455–79. and Pierre Chandon (2006), “Can ‘Low-Fat’ Nutrition Labels Lead to Obesity?” Journal of Marketing Research, 43 (November), 605–617. Wertenbroch, Klaus (1998), “Consumption Self-Control via Purchase Quantity Rationing,” Marketing Science, 17 (4), 317–37. Wilcox, Keith, Lauren Block, Gavan Fitzsimons, and Beth Vallen (2009), “Vicarious Goal Fulfillment: When the Mere Presence of a Healthy Option Leads to an Ironically Indulgent Decision,” Journal of Consumer Research, 36 (October), 380–93. World Health Organization (2011), “Obesity and Overweight: Fact Sheet No 311,” (May 12), (accessed January 18, 2012), [available at http://www.who.int/mediacentre/factsheets/fs311/en/].
Copyright of Journal of Marketing Research (JMR) is the property of American Marketing Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
