Persuasive research paper

profiletrx450rmb
unhealthy__tasty.pdf

The Unhealthy = Tasty Intuition and Its Effects on Taste Inferences, Enjoyment, and Choice of Food Products Author(s): Rajagopal Raghunathan, Rebecca Walker Naylor and Wayne D. Hoyer Reviewed work(s): Source: Journal of Marketing, Vol. 70, No. 4 (Oct., 2006), pp. 170-184 Published by: American Marketing Association Stable URL: http://www.jstor.org/stable/30162121 . Accessed: 22/11/2012 03:40

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]

.

American Marketing Association is collaborating with JSTOR to digitize, preserve and extend access to Journal of Marketing.

http://www.jstor.org

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

Rajagopal Raghunathan, Rebecca Walker Naylor, & Wayne D. Hoyer

The Unhealthy = Tasty Intuition and Its Effects on Taste Inferences, Enjoyment, and Choice of Food

Products Across four experiments, the authors find that when information pertaining to the assessment of the healthiness of food items is provided, the less healthy the item is portrayed to be, (1) the better is its inferred taste, (2) the more it is enjoyed during actual consumption, and (3) the greater is the preference for it in choice tasks when a hedonic goal is more (versus less) salient. The authors obtain these effects both among consumers who report that they believe that healthiness and tastiness are negatively correlated and, to a lesser degree, among those who do not report such a belief. The authors also provide evidence that the association between the concepts of "unhealthy" and "tasty" operates at an implicit level. The authors discuss possibilities for controlling the effect of the unhealthy = tasty intuition (and its potential for causing negative health consequences), including controlling the volume of unhealthy but tasty food eaten, changing unhealthy foods to make them less unhealthy but still tasty, and providing consumers with better information about what constitutes "healthy."

Assessing the health consequences of food consump- tion habits has emerged as an important topic of inquiry in recent years (Oakes 2005; Oakes and

Slotterback 2005; Seiders and Petty 2004; Verbeke 2006; Wansink 2004a, b; Wansink and Huckabee 2005). From an evolutionary perspective, the easy availability of food pre- sents a unique quandary for developed countries and societies. On the one hand, it has virtually obliterated prob- lems resulting from malnourishment. On the other hand, it has generated problems stemming from the overconsump- tion of food. For example, obesity is growing at an unprece- dented rate, not just in the United States, where nearly two- thirds of the population are overweight or obese, but in other developing nations as well (Seiders and Petty 2004).

To many consumers and consumer activists, organiza- tions (e.g., lobbyists for grocery manufacturers) and restau- rants (e.g., McDonald's, Hardee's) that are interested in pro- moting unhealthy eating habits are to blame for consumers' propensity to overconsume (Brownell and Horgen 2003;

Rajagopal Raghunathan is Assistant Professor of Marketing (e-mail: [email protected]), and Wayne D. Hoyer is The James L. Bayless/William S. Farish Fund Chair for Free Enterprises and Professor of Marketing (e-mail: [email protected]), McCombs School of Business, University of Texas at Austin. Rebecca Walker Naylor is Assistant Professor of Marketing, Moore School of Busi- ness, University of South Carolina (e-mail: [email protected]). The authors thank Susan Broniarczyk, Kristin Diehl, Julie Irwin, Andrew Gershoff, Jonathan Levav, and attendees at the 2005 Marketing Young Scholars Conference for their helpful suggestions on the article.

To read or contribute to reader and author dialogue on this article, visit http://www marketingpower. com/jmblog.

0 2006, American Marketing Association ISSN: 0022-2429 (print), 1547-7185 (electronic)

Nestle 2003). Consequently, lawsuits have been filed against fast-food companies such as McDonald's (Parloff 2003). Attorneys representing these companies have argued that consumers are aware of the ingredients in their offer- ings and that they understand the weight and health conse- quences of overconsumption. In their view, consumers are ultimately responsible for the negative consequences of their own actions.l

The objective of this research is to examine the role of a subtle (and currently unrecognized) factor in the overcon- sumption of food that is perceived as unhealthy. What if people consume food that is considered unhealthy not despite its perceived unhealthiness but rather because of it? That is, what if part of the attractiveness of food lies in its perceived unhealthiness? This can happen if consumers intuitively believe that the unhealthier the food, the tastier it is. The operation of such a belief would increase the chances that people will overconsume food portrayed as less (versus more) healthy because they expect that such food will taste better. Consistent with this idea, we propose that the perceived unhealthiness of food has the ironic effect of enhancing its attractiveness. Consequently, under cir- cumstances in which a hedonic (e.g., enjoyment) goal is more (versus less) salient, people will choose options they perceived as unhealthier, even if no information about their tastiness relative to other options is available. Across the experiments we conduct in this research, we obtain support for this proposition.

1The fast-food chains appear to be winning the battle, at least for now. In October 2005, the U.S. House of Representatives passed a bill to block lawsuits blaming restaurants and food com- panies for causing obesity or weight gain (Zhang 2005).

Journal of Marketing 170 Vol. 70 (October 2006,170-184

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

Our research generates two theoretically interesting and substantively meaningful insights. First, from a theoretical standpoint, we document both direct and indirect support for our main thesis, namely, that people subscribe to the unhealthy = tasty intuition and that this intuition affects inferences about food choice, actual enjoyment of foods, and choice of food products. These notions are particularly intriguing because there is little rigorous evidence that healthiness and tastiness are indeed inversely related to each other. Second, our research suggests that the influence of the unhealthy = tasty intuition can be implicit; as such, even people who may not report believing that unhealthy food is tastier make judgments and choices as if they sub- scribed to such a view.

We structure the rest of the article as follows: We begin by briefly commenting on the potential sources for the unhealthy = tasty intuition. Then, we derive predictions of how the intuition affects three variables of theoretical and substantive interest; (1) taste inferences of food products when information pertaining to their healthiness is pro- vided, (2) differences in enjoyment of the same food prod- uct when it is presented as being more versus less healthy, and (3) choice between two snack options as a function of differences in the propensity to pursue a hedonic goal. We report results from four controlled experiments that test these hypotheses and conclude with a discussion of the implications of our findings.

Theory Sources of the Unhealthy = Tasty Intuition Intuitions or lay beliefs and theories may be generated internally, through personal experience and self-observation (Ross and Nisbett 1991), or externally, from environmental cues (Morris, Menon, and Ames 2001). We posit that both types of sources underlie the unhealthy = tasty intuition.

Internal source. We posit that the unhealthy = tasty intu- ition is generated internally on the basis of a belief in a compensatory relationship between the "wholesomeness" of stimuli and their "hedonic potential." That is, we believe that the unhealthy = tasty intuition is a specific manifesta- tion of a more general principle, according to which there is an inverse relationship between things that are "healthy," "nourishing," and "good for you" and those that are "enjoy- able," "fun," and "exciting." Evidence for an assumed com- pensatory relationship between the wholesomeness and hedonic potential of stimuli emerged across a separate series of experiments we conducted. For example, the results from one such experiment revealed that compared with an automobile portrayed as attractive and fun to drive, one that was portrayed as unattractive and less enjoyable to drive was inferred to be safer, even though the more attrac- tive automobile was objectively safer (as per the Insurance Institute of Highway Safety). We speculate that a belief in the more general unwholesome = fun intuition is rooted in religious messages, such as the Protestant work ethic, according to which a person is morally obligated to accord greater priority to necessities over luxuries (e.g., Berry

1994; Weber 1998; see also Maslow 1962). Implicit in this view are the ideas that comfort and luxury do not come eas- ily and that a person needs to work hard to earn the right to indulge (e.g., Kivetz and Simonson 2002). Thus, the Protes- tant work ethic makes a clear distinction between stimuli or activities that are of a "serious," "worthy," and "work- oriented" nature and those that are of a "trivial," "frivolous," and "fun-related" nature and suggests that stimuli/activities fall neatly into one or the other category but rarely into both. In the context of food, this translates into a belief that healthy food is less tasty.

External source. Operating in concert with the internal source for the unhealthy = tasty intuition is an external one. Through the mass media or personal communication, people are constantly exposed to views that are compatible with the intuition. Several recent articles in popular maga- zines (e.g., "Is Fat the Next Tobacco?" which appeared in Fortune [Parloff 2003]), popular books (e.g., Fast Food Nation, by Eric Schlosser [2001]), and movies (e.g., Mor- gan Spurlock's Super Size Me, which appeared in theaters nationwide in 2004) explicitly or implicitly suggest an inverse relationship between healthiness and tastiness. Con- sumers are also likely to have been exposed to the unhealthy = tasty view through personal communication. For example, parents and guardians appear to find value in propagating such a message because it encourages children to consume hedonically unpleasant food that is perceived as salubrious (e.g., eating broccoli) and discourages them from overconsuming hedonically appealing fare that is poten- tially harmful (e.g., candy).

In summary, we propose that the unhealthy = tasty intu- ition is generated internally through its consistency with the more general unwholesome = fun intuition and/or exter- nally through repeated exposure to views that are compati- ble with it.2 We believe that these sources help generate the hypothesis that unhealthiness is inversely related to tasti- ness. After this association has been generated, psychologi- cal mechanisms help perpetuate it, as we discuss next.

Influence of the Unhealthy - Tasty Intuition on Decision Making Figure 1 depicts the influence of the unhealthy = tasty intu- ition on three conceptually separate stages of the decision-

2It is worthwhile to mention an alternative source for the intu- ition: direct experience. A person may arrive at the intuition through lay inferencing (Kelly and Thibaut 1969), a process that involves personal observation that unhealthier food is tastier. There are two reasons we do not believe that this is a valid source for the intuition. First, lay inferencing presupposes objectivity on the part of the inference maker. However, we know that because of the prevalence of external messages consistent with the unhealthy = tasty intuition, consumers are unlikely to be objective judges of the correlation between healthiness and tastiness. Sec- ond, to assess accurately the true correlation between healthiness and tastiness, it is first necessary to assign relatively stable health- mess and tastiness ratings to various types of food. However, rele- vant findings indicate that ratings of the healthiness and tastiness of food likely depend heavily on contextual factors (Drewnowski 1997; Oakes 2005).

The Unhealthy = Tasty Intuition 1171

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

making process. Research has shown that consumers gener- ate inferences about missing product attributes on the basis of lay theories, that is, beliefs and intuitions about the world (Broniarczyk and Alba 1994; Sujan and Dekleva 1987). Indeed, inferences about missing product attributes are dri- ven by intuitions even in the face of contradictory cues. For example, in a study in which participants were asked to make inferences about a camera's durability, the partici- pants relied more heavily on the intuition that durability and warranty are positively correlated than on the actual corre- lation (in the presented data) between durability and shutter speed, an association that is not supported by intuition (Bro- niarczyk and Alba 1994).

Influence of intuition on taste inferences. Consistent with these findings, we predict that as a consequence of belief in the unhealthy = tasty intuition (see Figure 1, Stage 1),

H1: The healthier a food item is perceived to be, the lower is its inferred tastiness.

Influence of intuition on enjoyment. Inferences about missing attributes are likely to act as hypotheses that can bias judgments about these attributes (Deighton 1984; Hoch and Ha 1986; Shiv, Carmon, and Ariely 2005). The opera- tion of this hypothesis confirmation bias was captured in Shiv, Carmon, and Ariely's (2005) experiments, which were designed to demonstrate the placebo effects of marketing actions. For example, in their first experiment, they find that

consumers who pay a discounted price for an energy drink that purportedly increases mental acuity derive less benefit from the drink (measured in terms of number of mental puzzles solved) than do participants who pay full price. Consistent with the operation of the hypothesis confirma- tion bias, the discounted price serves as a hypothesis about the (lack of) effectiveness of the drink, which is confirmed on consumption of the drink (and reflected in the subse- quent puzzle-solving task). In the context of the current research, we expect that consumers will judge the same food product as more tasty when it is portrayed as less healthy. We state this prediction more formally as follows (see Figure 1, Stage 2):

Hz: When actual tastiness of a food product is controlled for, the healthier the food is portrayed to be, the lower is its judged tastiness.

Influence of intuition on choice. Product choice is often a function of consumption goals (Bettman, Luce, and Payne 1998). In the context of food choice, researchers have sug- gested that, in general, the hedonic goal of enjoyment is salient and important (Dhar and Simonson 1999; Food Mar- keting Institute 1996; Shiv and Fedorikhin 1999; Tepper and Trail 1998). For example, Shiv and Fedorikhin's (1999) experiments examine how a person's mood can influence how he or she resolves the trade-off between the enjoyment and the health benefits of food offerings. Likewise, Dhar and Simonson's (1999) experiments examine conditions that lead to "highlighting" (a focus on either an enjoyment

FIGURE 1 Model of the Influence of the Unhealthy = Tasty Intuition on Judgments and Decisions

172 / Journal of Marketing, October 2006

Stage 1 Stage 3

Choice

Salience of hedonic goal

Stage 2

Actual enjoyment

Belief in the unhealthy = tasty intuition

Taste inferences

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

or a health goal in choosing an entree and a dessert) versus "balancing" (a focus on both goals). Given that enjoyment is likely to be a significant goal when people choose food products, the unhealthy = tasty intuition suggests that con- sumers with a more (versus less) salient hedonic goal will have a relatively greater preference for options they per- ceived as unhealthy (see Figure 1, Stage 3).

H3: Given a choice between two food products that vary in healthiness, people with a more (versus less) salient enjoyment goal will choose the less healthy option.

Implicit and explicit influence of intuition. Thus far, we have suggested that consumers' taste inferences, the judged tastiness of food, and product choice will be influenced in a manner consistent with the unhealthy = tasty intuition. Here, we explore whether explicitness of belief in the intu- ition moderates this general pattern. If we find that even consumers who do not explicitly report a belief in the intu- ition make decisions as if they did, important theoretical and substantive implications would emerge. From a theo- retical standpoint, such a finding would suggest that the influence of the intuition can occur outside the awareness of consumers making consumption decisions in the context of food. From a substantive standpoint, the finding would point to the difficulty of controlling the influence of the intuition.

The possibility that even people who do not report a belief in the intuition will make decisions as if they believed in it is consistent with the mechanism of hidden covariation detection (e.g., Lewicki 1986; Lewicki and Hill 1989; Lewicki, Hill, and Czyzewska 1997). Lewicki and col- leagues find that people can learn the relationship between variables implicitly, that is, without being able to articulate their learning. In one experiment (Lewicki 1986, Experi- ment la), participants were exposed to information about some traits (e.g., kindness, capability) of women such that these traits covaried with their hair length. Later, the same participants were asked to assess these traits in another set of women with long and short hair. Participants' judgments of these targets were consistent with the covariations between hair lengths and traits in the initial stimuli. Impor- tantly, even participants who were unable to articulate these covariations explicitly made judgments as if they had learned them. Thus, people are able to acquire information nonconsciously about covariations in a data set without being able to report such learning explicitly (Lewicki and Hill 1989, p. 240).

Consistent with Lewicki and colleagues' (e.g., Lewicki 1986; Lewicki and Hill 1989; Lewicki, Hill, and Czyzewska 1997) work, we expect that as a result of exposure to the many external sources that suggest a negative correlation between healthiness and tastiness, consumers will have implicitly learned to associate unhealthiness with tastiness. Thus, we expect that even consumers who do not explicitly report believing that unhealthier food is tastier will make judgments and decisions as if they did. However, because there is evidence that explicitly (versus implicitly) learned and articulated relationships are likely to have significantly stronger influences on judgments (Allen and Janiszewski 1989), we expect that the influence of the unhealthy = tasty

intuition will be stronger among people who report explicit belief in it. Thus:

H4: The pattern of results we predict in H1 -H3 holds for both people who report an explicit belief in the unhealthy = tasty intuition and those who do not, but it is stronger among the former group.

Experiment 1: Implicitness of Belief in Intuition We designed Experiment 1 to establish that people implic- itly subscribe to the unhealthy = tasty intuition. To address this objective, we used the implicit association test (IAT; Greenwald, McGhee, and Schwartz 1998; Greenwald, Nosek, and Banaji 2003), an established tool for demon- strating the type of implicit association we wish to demonstrate.

Participants and Procedure One hundred thirty-eight undergraduate students partici- pated in this experiment (which we conducted using IBM computers and the Direct RT software) for course extra credit. Participants were told that their task was to catego- rize correctly stimuli shown on the middle of their screens. We selected stimuli from one of the following four cate- gories (see the Appendix) for the categorization tasks: (1) pictures of unhealthy foods, (2) pictures of healthy foods, (3) words associated with the tastiness or enjoyment of food (e.g., tasty, delicious), and (4) words associated with the lack of taste or enjoyment (e.g., flavorless, unpalatable). In line with established protocol (Greenwald, McGhee, and Schwartz 1998), participants completed seven blocks of tri- als, five of which were practice blocks designed to familiar- ize participants with the target stimuli and the categoriza- tion labels. The critical blocks for analysis were Block 4, in which the category labels were "Healthy Foods/Enjoyable" versus "Unhealthy Foods/Not Enjoyable," and Block 7, in which the category labels were "Unhealthy Foods/ Enjoyable" and "Healthy Foods/Not Enjoyable" (see Table 1).

Stimuli from all four categories (healthy food images, unhealthy food images, enjoyment words, and lack of enjoyment words) were presented for classification in the critical blocks for 48 trials across both blocks. An implicit belief in the intuition that unhealthy foods are tastier (and that healthier foods are less tasty) would be reflected in faster responses when stimuli were to be categorized in the condition in which the classification task was congruent with the intuition (i.e., when participants paired unhealthy foods with words associated with tastiness and enjoyment) than in the incongruent condition (i.e., when participants paired healthy foods with words associated with tastiness and enjoyment). Explicit measures of belief in the unhealthy = tasty intuition were collected (using paper-and- pencil measures) before the IAT procedure to compare par- ticipants' explicit and implicit attitudes. We assessed explic- itness of participants' beliefs in the intuition by eliciting their agreement with the following two items (1= "strongly disagree," 9 = "strongly agree"): (1) "Things that are good

The Unhealthy . Tasty Intuition 1173

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

TABLE 1 Experiment 1: Sequence of Trial Blocks in IAT

Number of Items Assigned to Left-Key Items Assigned to Right-Key Block Trials Function Response Response 1 20 Practice Healthy food images Unhealthy food images

2 28 Practice Good taste/enjoyment words Bad taste/lack of enjoyment words

3 24 Practice Healthy food images + good Unhealthy food images + bad taste/enjoyment words taste/lack of enjoyment words

4 48 Critical test block Healthy food images + good Unhealthy food images + bad taste/enjoyment words taste/lack of enjoyment words

5 40 Practice Unhealthy food images Healthy food images

6 24 Practice Unhealthy food images + good Healthy food images + bad taste/enjoyment words taste/lack of enjoyment words

7 48 Critical test block Unhealthy food images + good Healthy food images + bad taste/enjoyment words taste/lack of enjoyment words

for me rarely taste good," and (2) "There is no way to make food healthier without sacrificing taste."

Data Preparation

Following the revised IAT scoring algorithm (Greenwald, Nosek, and Banaji 2003), we subjected all data to the fol- lowing criteria: (1) We eliminated trial response latencies greater than 10,000 milliseconds (of the 13,248 trials in the critical blocks, 2 were above this threshold and thus were eliminated); (2) we excluded participants whose response times were less than 300 milliseconds on more than 10% of the critical trials (we discarded 2 of the participants, result- ing in a final sample of 136 participants); (3) we included all response latencies, even those for false responses; and (4) we computed a difference score between the two critical blocks of trials ([congruent: unhealthy food + enjoyable and healthy foods + not enjoyable] - [incongruent: healthy foods + enjoyable and unhealthy foods + not enjoyable] divided by the pooled standard deviation of response laten- cies across both blocks). The resultant measure is the IAT D effect (Greenwald, Nosek, and Banaji 2003).

Results and Discussion Response latencies were significantly lower when the stim- uli were presented in the congruent (versus incongruent) condition. Mean response time when participants were asked to classify stimuli into the congruent categories (pair- ing unhealthiness with enjoyment) was 767.76 millisec- onds, compared with 1137.91 milliseconds when partici- pants were asked to classify stimuli into the incongruent categories (pairing healthiness with enjoyment); this differ- ence in response time was significant (D = .648, SD = .497, t(135) = 17.19, p < .0001, d =1.42). That participants were able to respond significantly faster when asked to group unhealthy foods with words describing tastiness and enjoy- ment than with words describing lack of taste and lack of enjoyment suggests that participants have a stronger implicit association between unhealthiness and tastiness than they do between healthiness and tastiness.

To test whether the unhealthy = tasty intuition influ- ences the response times of even the participants who did not report explicit belief in the intuition, we first averaged the items used to measure explicitness of belief to form an explicitness-of-belief index. We then assessed the mean dif- ference in response time for the congruent versus incongru- ent condition at each level of reported explicitness of belief in the intuition (see Irwin and McClelland 2001). The results revealed that this difference was significant at every level of belief. That is, at all reported levels of explicit dis- agreement and agreement with the intuition, the difference in response latencies when participants were asked to pair healthiness with enjoyment versus when they were asked to pair unhealthiness with enjoyment was significantly differ- ent from zero (all is > 3.23, all ps < .01). Participants at all levels of belief were able to classify unhealthiness with enjoyment more quickly than they were able to classify healthiness with enjoyment.

Thus, the results of Experiment 1 are consistent with our hypothesis that consumers implicitly subscribe to the view that healthiness and tastiness are inversely related to each other, regardless of their level of explicit belief in this relationship. As such, we expect that even consumers who do not report agreeing that healthiness and tastiness are inversely related to each other will make judgments and decisions as if they did; we tested this hypothesis in the next three experiments.

Experiment 2: Taste Inferences of Crackers

The results from Experiment 1 demonstrated that people implicitly associate unhealthiness with tastiness. Thus, we expect that the influence of the unhealthy = tasty intuition on decision making can take place outside of awareness such that even participants who do not report believing that healthiness and tastiness are inversely related to each other make decisions as if they believe in such a relationship. In general, the objective of Experiment 2 was to test this pre-

174 I Journal of Marketing, October 2006

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

diction in the context of inference making. Specifically, the objective of this experiment was twofold: First, we wanted to assess whether a food product portrayed as unhealthy would be inferred to taste better (H1) than one portrayed as healthy. Second, we wanted to assess whether such an infer- ence is made both by people who explicitly report believing that tastiness is inversely related to healthiness and, to a lesser degree, by those who do not (H4).

Participants, Stimuli, and Procedure One hundred ten undergraduate students (51 men and 59 women) took part in this experiment for course extra credit. Participants were told that the purpose of the experiment was to assess consumers' taste perceptions of three new brands of cheddar-flavored snack crackers. To lend credibil- ity to the cover story, participants were told that a national cracker manufacturer was interested in introducing these brands of crackers into the local market. The brands were portrayed as varying in terms of unhealthiness. After being exposed to nutritional information for the cracker versions, participants were asked to infer the tastiness of the three versions.

Unhealthiness manipulation. The most natural way to manipulate unhealthiness perceptions is to portray offerings as differing in the content of an ingredient (e.g., fat) that is generally accepted as being unhealthy. However, altering the content of fat (or any other ingredient) can alter taste perceptions directly (i.e., without altering perceptions of unhealthiness). As a result, varying content of a specific ingredient would confound our prediction with the compet- ing possibility that the ingredient may independently enhance inferred tastiness of the crackers. To overcome this problem, we chose to manipulate perceptions of unhealthi- ness by varying the purported content of "bad" versus "good" fat, while controlling for the crackers' total fat con- tent. This allowed us to manipulate the perceived unhealthi- ness of the crackers selectively while controlling for any (potential) effects due to a direct association of fat content with taste.

The unhealthiness manipulation unfolded as follows: Before seeing the nutritional information for each cracker version, participants were presented with the following information about the characteristics of bad (saturated) and good (unsaturated) fat:

Medical opinion suggests that consuming unsaturated fat (good fat) can raise the level of high-density lipoproteins (HDL), which carry cholesterol from the heart to the liver and thereby eliminate excess cholesterol; in contrast, con- suming saturated fat (bad fat) can raise the level of low- density lipoproteins (LDL), which carry cholesterol from the liver to the rest of the body and thereby lead to blocked coronary arteries. (adapted from Harvard School of Public Health Nutrition Source 2004)

Then, participants were told that each of the three new brands of crackers contained 13 grams of total fat per serv- ing. One brand (the healthy alternative) was portrayed as containing 11 grams of good fat and 2 grams of bad fat, and another (the unhealthy alternative) was portrayed as con- taining 2 grams of good fat and 11 grams of bad fat. A third, "control" version was portrayed as containing 6.5 grams of

both types of fat. All participants were exposed to informa- tion about all three versions of the crackers in a counterbal- anced manner.

Dependent and other variables. After exposure to infor- mation about the fat content of the three cracker versions, participants were asked to rate "How tasty do you think these crackers would be?" and "How much do you think you would enjoy eating them?" on a ten-point scale (1 = "not at all," 10 = "very") for each cracker version. To test whether the influence of the unhealthy = tasty intuition can occur implicitly, we assessed explicitness of participants' beliefs in the intuition by eliciting participants' agreement with the same two statements we used in Experiment 1. Finally, par- ticipants provided information on their liking for cheddar- flavored snack crackers and also indicated their perceptions of the unhealthiness of the three cracker versions.

Results

We dropped 4 participants from the analyses for indicating that they did not like any type of cheddar-flavored snack cracker (a rating of 1 on the scale that assessed liking). This left 106 participants (48 men and 58 women).

Manipulation check. We first wanted to assess whether our unhealthiness manipulation worked as intended. We regressed perceived unhealthiness of crackers on their pur- ported bad fat content. The results revealed a positive monotonic relationship between purported bad fat content and perceived unhealthiness (f3 = +.39, t(101) = 9.32, p < .001), revealing that the unhealthiness manipulation worked as we intended.

Taste inferences. The items used to measure tastiness and enjoyment of the crackers were highly correlated (Cronbach's a = .86); thus, we averaged them to form a tastiness index. The items used to measure explicitness of belief in the intuition were also highly correlated (Cron- bach's u = .73); thus, we averaged them to form an explicitness-of-belief index, as in Experiment 1. Using the taste inference index of the crackers as the dependent variable, we conducted a mixed analysis of variance (ANOVA) with three levels of portrayed unhealthiness of the cracker (healthy, control, unhealthy) as a repeated factor and explicitness of belief in the intuition as a continuous between-subjects factor. The results revealed a marginal main effect of unhealthiness (F(2,103) = 2.44, p = .09). As we predicted, mean inferred tastiness was highest for the cracker portrayed as containing the highest amount of bad fat (M = 7.33) and lowest for the cracker portrayed as con- taining the lowest amount of bad fat (M = 5.16), with the control cracker falling in between (M = 6.38). Although the differences between the inferred tastiness of the unhealthy and the control cracker and those between the healthy and the control cracker were not significant (F(1, 104) = 1.53 and 1.84, respectively, both ps > .12), the focal contrast between tastiness ratings of the unhealthy and healthy crackers was significant (F(1, 104) = 17.02, p < .001). These results offer evidence consistent with HI and indicate that consumers infer the taste of food products when infor- mation relevant to assessing their healthiness is provided in a manner consistent with the unhealthy = tasty intuition.

The Unhealthy = Tasty Intuition 1175

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

Role of explicitness of belief. The results of the ANOVA also revealed a significant unhealthiness x explicitness-of- belief interaction (F(2, 103) = 12. 12, p < .001), suggesting that the inference that crackers with more bad fat tasted bet- ter was made to differing degrees, depending on partici- pants' strength of belief in the unhealthy = tasty intuition. To test whether the unhealthy = tasty intuition influenced the taste inferences of participants who did not report explicit belief in it, we assessed the mean difference in inferred tastiness ratings of the unhealthy versus healthy cracker at each level of reported explicitness of belief in the intuition (Irwin and McClelland 2001). The results revealed that this difference was significant at every level of belief. That is, as we depict in Figure 2, at all reported levels of disagreement and agreement with the intuition, the differ- ence in inferred tastiness of the unhealthy versus healthy cracker was significantly different from zero (all is > 2.68, all ps < .01).3

Discussion The results from Experiment 2 lead us to conclude that the unhealthy = tasty intuition influences taste inferences. Although participants in the experiment did not taste any of the crackers, and the total amount of fat was the same across the different versions of the cracker, they readily assumed that crackers with more bad fat would taste better. Furthermore, we also found that though the effect of por- trayed healthiness on taste inferences was particularly strong among those who reported an explicit belief in the unhealthy = tasty intuition, we also obtained the effect among participants who reported disagreement with the intuition. This suggests that consumers may implicitly sub- scribe to the view that healthiness and tastiness are inversely correlated, thus conceptually corroborating the results we obtained in Experiment 1.

Experiment 3: Enjoyment of Mango Lassi

We wanted to expand on our findings in several ways in Experiment 3. First, we attempted to address a potential

3To ensure that our results hold when the unhealthy = tasty intu- ition is measured as such (i.e., that unhealthy implies tasty rather than that healthy implies not tasty), we conducted an additional study with 287 undergraduate students that reversed the wording of the explicitness-of-belief measure. That is, participants were asked to rate their agreement with the statement "The more unhealthy a food, the better it tastes" rather than the statement "There is no way to make food healthier without sacrificing taste." The results were nearly identical to those we found in Experiment 2. Participants rated the unhealthy cracker (M = 7.31) as signifi- cantly tastier than the healthy cracker (M = 5.33; F(1, 286) = 163.71, p < .0001), and this difference in taste ratings interacted with explicitness of belief (F(1, 285) =10.99, p < .001), such that higher levels of belief were associated with greater differences in taste ratings. Furthermore, when we used Irwin and McClelland's (2001) method, the difference in rated taste between the unhealthy and the healthy cracker was significantly different from zero at every level of explicit belief in the intuition (all is > 2.57, all ps < .05) except at the lowest level. At the lowest level of explicit belief, this difference approached significance (t =1.55, p = .12).

limitation arising from the within-subjects design used in Experiment 2. Because participants were asked to infer the tastiness of three different cracker versions that varied only in terms of portrayed healthiness, it could be argued that the participants may have guessed our hypotheses. Although the significance of the two-way unhealthiness x explicitness-of-belief interaction suggests that demand effects cannot fully explain our findings, we wanted to act conservatively and conceptually replicate these findings using a between-subjects manipulation of unhealthiness. Second, we wanted to demonstrate that the portrayed healthiness of food affects actual (versus inferred) enjoy- ment (see Figure 1, Stage 2). Such a finding would indicate that the unhealthy = tasty intuition acts as a hypothesis that participants confirm after sampling stimuli (e.g., Hoch and Ha 1986; Shiv, Carmon, and Ariely 2005). Third, we wanted to assess whether the implicit nature of the influ- ence of the unhealthy = tasty intuition can be replicated in the context of actual enjoyment. Such a finding would fur- ther attest to the robustness of H4. Finally, we tested the generalizability of our results by using a different (1) prod- uct category, (2) unhealthiness manipulation, and (3) exper- imental setup.

Forty adults were invited to a housewarming party at the residence of one of the authors of this article. In the invita- tion to the event (sent by e-mail), invitees were informed that refreshments would be served at the party. The refresh- ment item targeted for testing our predictions was Mango Lassi (an East Indian delicacy similar to a milkshake). We used the context of the housewarming party, which pro- vided a more realistic setting than the laboratory, to test our predictions.

FIGURE 2 Experiment 2: Difference in Inferred Tastiness of an Unhealthy Versus a Healthy Cracker at Each Level of Explicitness of Belief in the Unhealthy =

Tasty Intuition

0 2 4 6 8 10 Belief

1761 Journal of Marketing, October 2006

Belief

Belief

Badgood rating difference Linear (Badgood rating difference)

7.00

6.00

5.00

4.00

3.00

2.00

1.00

.00

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

Participants, Stimuli, and Procedure

At the party, invitees were told that snacks had been ordered from a newly opened Indian restaurant (Hyderabad Masala) and that the operators of this restaurant were interested in catering mainly to an American audience. To lend credibil- ity to the cover story, during appropriate moments at the party, the host nonchalantly expressed that "when the restaurant operators realized that I was going to have a pre- dominantly American crowd at the party, they wondered if the guests would be willing to fill out a short questionnaire about the food. I didn't see any harm in agreeing, especially since they promised to give you a 25% discount on your next visit to the restaurant" A coupon for "25% off the next meal at Hyderabad Masala" was made available to the invi- tees. To lend further credibility to the cover story, a confed- erate (a graduate student of East Indian origin) was hired to play the role of an agent representing the restaurant. The confederate handed out the one-page questionnaire that pro- vided information about the following food items offered at the party and elicited participants' judgments of the tasti- ness and enjoyment of the following items using a ten-point scale (1 = "not at all," 10 = "very"): (1) Idlis, (2) Samosas, and (3) Mango Lassi.4 These items were placed on a dining table in the formal dining area. The Idlis and Samosas were served on paper plates, and the Mango Lassi was served in prefilled 12-ounce plastic cups.

Because the invitees arrived at various times during a fairly large (four hours) time interval and because the din- ing area was physically distant from the family room, where people congregated, the invitees tended to be separated from the other guests at the time of sampling the food items and filling out the questionnaire. This minimized potential response bias effects due to social influence (Raghunathan and Corfman 2006). To encourage candidness further, the confederate (who was always present next to the food) stressed to every invitee, "Since we [the restaurant opera- tors] are interested in catering to the American taste, we request that you respond as candidly as possible. Please don't worry about hurting our feelings."

Unhealthiness manipulation. Across all participants, the Idli was portrayed as a healthy item, and the Samosa was portrayed as an unhealthy one. These two items were not the focus of this experiment, but we included them to lend credibility to the cover story and to mask the main target of this experiment (the Mango Lassi). For roughly half of the participants, the Mango Lassi was portrayed as a healthy item ("made out of real mango pulp and milk; generally considered very healthy"), and for the other half, it was por- trayed as an unhealthy one ("made out of real mango pulp and milk; generally considered unhealthy"). Note that this manipulation (unlike the one we used in Experiment 2) serves to alter perceptions of unhealthiness directly (rather than manipulate it by altering the content of an ingredient considered unhealthy).

4Idli is a traditional South Indian dish made by steaming a mix- ture of mainly lentils and rice. Samosas are deep-fried vegetable dumplings. Lassi refers to sweet drinks made with a base of yogurt.

Measurement of explicitness of belief in the intuition. Two days after the party, the host e-mailed the invitees indi- vidually, attaching an article that had recently appeared in the local newspaper. The newspaper article suggested that the "low-carb fad" was waning. The purported purpose of the e-mail was to alert the recipients to the content of the "interesting" article, but the true purpose was to gauge the invitees' explicitness of belief in the unhealthy = tasty intu- ition. Using the news article as a backdrop, the host assessed the invitees' beliefs in the relationship between healthiness and tastiness by asking the following question, "On a scale of one to nine, how much would you agree with the following statement: `Food that is unhealthy generally tastes better'?" We took invitees' ratings as a measure of the explicitness of their beliefs in the unhealthy = tasty intu- ition; higher numbers indicated greater explicitness of belief.

Results

About a week after the party, the invitees were individually e-mailed and asked if they had suspected that the true pur- pose of the questionnaire administered during the party was different from the purported one. We dropped data from one individual who reported suspicion.

Taste inferences. Tastiness and enjoyment of the Mango Lassi were highly correlated with each other (Cronbach's a = .73); thus, we averaged them to form an enjoyment index. Using this index as the dependent variable, we con- ducted a mixed ANOVA with two levels of unhealthiness of Mango Lassi (healthy, unhealthy) as a between-subjects factor and explicitness of belief in the intuition as a contin- uous between-subjects factor. The results from this analysis appear in Figure 3. Consistent with our prediction (H2), the mean enjoyment ratings were higher for participants in the condition in which the milkshake was portrayed as being unhealthy (M = 5.13) than for participants in the condition in which the milkshake was portrayed as being healthy (M=3.30;F(1, 35)=24.29,p<.001).

Role of explicitness of belief Echoing findings from Experiment 2, the ANOVA also revealed a marginal health- iness x explicitness-of-belief interaction (F(1, 35) = 3.91, p = .06), indicating that the relationship between enjoyment of the milkshake and its perceived healthiness depended on participants' explicitness of belief in the unhealthy = tasty intuition. Follow-up analyses revealed that the effect of unhealthiness on enjoyment was significant both at one standard deviation above the mean belief level (t(1, 35) = 4.84, p < .001) and at one standard deviation below it (t(1, 35) = 2.00, p = .05), indicating that portrayed healthiness of food significantly influenced participants' enjoyment of the food, even among those who did not report agreeing with the unhealthy = tasty intuition. These results provide further support for our hypothesis that the influence of the intuition on consumption decisions can take place outside of awareness.

Discussion Viewed in conjunction with the results from Experiments 1 and 2, the results from Experiment 3 indicate that taste

The Unhealthy = Tasty Intuition 1177

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

FIGURE 3 Experiment 3: Mango Lassi Enjoyment Ratings x Explictness of Belief That Food That Is Unhealthy

Generally Tastes Better

Healthy Con ition Unhealthy Condition

inferences made on the basis of perceived unhealthiness are confirmed on sampling the food, implying that a mecha- nism akin to the hypothesis confirmation bias may underlie the prevalence of the unhealthy = tasty intuition. Further- more, the finding that even participants who reported dis- agreeing that healthiness needs to be sacrificed in order to enhance tastiness enjoyed the milkshake more when it was portrayed as unhealthy (versus healthy) indicates that the influence of the unhealthy = tasty intuition can be implicit, thus conceptually replicating results from Experiments 1 and 2. Finally, the results were consistent with our predic- tions even though we used a different (1) design (between- versus within-subjects), (2) product category, and (3) unhealthiness manipulation; this attests to the robustness of the influence of the unhealthy = tasty intuition.

Experiment 4: Choice of Crackers The results from Experiments 2 and 3 suggest that people's belief in the unhealthy = tasty intuition is sufficiently strong to influence taste inferences and enjoyment of food. The objective in Experiment 4 was to further explore whether the intuition would influence choice decisions (see both Figure 1, Stage 3, and H3). We know from previous research that when consumers are faced with the task of choosing among food products, better-tasting options have a higher likelihood of being chosen when a hedonic goal is relatively more salient (e.g., Dhar and Simonson 1999; Shiv and Fedorikhin 1999). Extrapolating from these results, we expect that participants will prefer an option portrayed as being more (versus less) unhealthy when a hedonic goal is

relatively more salient because according to our hypothesis, people will infer this option to taste better (and thus will judge it as a better candidate for fulfilling the hedonic goal).

Participants, Stimuli, and Procedure Two hundred ninety-three undergraduate students (107 men, 184 women, 2 unknown gender) participated in this experiment for course extra credit. They were first told that a national manufacturer of crackers was interested in assessing which of two versions of a cheddar-flavored snack cracker is preferred in choice tasks. Subsequently, information about good and bad fat (identical to the infor- mation we used in Experiment 1) was provided before par- ticipants chose between a cracker version with 11 grams of bad fat and 2 grams of good fat (unhealthy version) and one with 2 grams of bad fat and 11 grams of good fat (healthy version).

Before indicating their choice, approximately half of the participants were told to imagine that they were "craving something really tasty" and that "they wanted to reward themselves with a nice snack." The purpose of these state- ments was to prime a hedonic, or enjoyment, goal. The remaining participants were not exposed to such a prime; they were simply told to imagine that they were "in the mood for a snack." The dependent variable was which cracker version participants selected in a forced-choice task. We elicited participants' explicitness of belief in the unhealthy = tasty intuition using the same questions we used in Experiments 1 and 2. Consistent with H3, we expected that the relative preference for the unhealthy (ver- sus healthy) cracker version would be higher among partici- pants who were primed with the hedonic goal.

Results and Discussion

Manipulation check and choice. As we expected, par- ticipants perceived the cracker with the higher content of good (versus bad) fat content as healthier (Ms = 6.13 and 3.05; t(291) = 1067.71, p < .01). To test our prediction that the relative preference for the unhealthy (versus healthy) option would be higher among those primed with a hedonic goal, we conducted a chi-square test. As we predicted, the proportion of participants who chose the unhealthy option was significantly greater among those who were primed with the hedonic goal (M = 21%) than among those who were not primed with this goal (M = 8%; x2 = 15.88, p < .001; .001; = .23).5

Role of explicitness of belief. To test the role of explicit- ness of belief, we conducted an ANOVA using the prime and explicitness of belief in the intuition as the predictor variables and choice as the dependent variable. The results revealed a nonsignificant prime x explicitness-of-belief interaction (F(1, 258) = .59, p > .44), indicating that, regardless of whether the participants reported agreeing

5The proportion of participants who chose the unhealthy cracker was low even in the hedonic prime condition, presumably because participants did not want to indicate that they were interested in consuming a cracker that was explicitly portrayed as being unhealthy.

178 I Journal of Marketing, October 2006

Low belief that food that is unhealthy generally tastes better

High belief that food that is unhealthy generally tastes better

6.00

5.00

4.00

3.00

2.00

1.00

.00

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

with the unhealthy = tasty intuition, they made choice deci- sions as if they did.

Thus, the results from Experiment 3 conceptually repli- cate those we obtained previously, providing evidence that the unhealthy = tasty intuition affects not just inferences and judgments but actual choice decisions as well. We found that when a hedonic goal was made temporarily more salient, participants were more likely to select the unhealthy cracker, presumably because they inferred that it would taste better. Furthermore, the lack of significance of the prime x explicitness-of-belief interaction suggests that the influence of the intuition can take place outside the aware- ness of consumers, indicating that controlling for the influ- ence of the intuition may be difficult.

General Discussion Understanding the effects of food consumption habits on health has emerged as a topic of significance in recent years. The question of why consumers overconsume food despite the negative consequences of doing so has espe- cially piqued the interest of academics (e.g., Kozup, Creyer, and Burton 2003; Oakes 2005; Oakes and Slotterback 2005; Seiders and Petty 2004; Smith 2004; Tepper and Trail 1998; Urala and Lahteenmaki 2004; Verbeke 2005; Wansink 2004 a, b; Wansink and Chandon 2006; Wansink and Huckabee 2005). Previous findings are consistent with the idea that people overconsume food that is considered unhealthy either because they are unaware of the negative conse- quences of doing so (presumably because of poor marketing and/or presentation of nutritional information) or because they lack the willpower to resist such foods. In what we believe is the first rigorous empirical demonstration of its kind, our findings indicate that in addition to these factors, a more subtle force may be at work; namely, consumers over- consume foods that they perceive as unhealthy because they assume that such food tastes better.

The results from Experiment 1, in which we used the well-established IAT procedure (Greenwald, McGhee, and Schwartz 1988), demonstrated not only that people sub- scribe to the unhealthy = tasty intuition but also that this intuition operates at an implicit level. Both participants who reported an explicit belief in the intuition and those who did not took longer to categorize stimuli in the incongruent (versus congruent) condition, which suggests that the implicit subscription to the unhealthy = tasty intuition is robust.

The results from three other experiments confirmed our predictions that when information pertaining to the assess- ment of the healthiness of food items is provided, foods per- ceived as less healthy are (1) inferred to taste better, (2) enjoyed more during actual consumption, and (3) preferred in a choice task when a hedonic goal is more (versus less) salient. We obtained these patterns of results even among those who disagreed that healthiness is inversely related to tastiness, which suggests that the influence of the unhealthy = tasty intuition can take place outside of aware- ness; this result conceptually replicates Experiment 1's results, which demonstrate that subscription to the unhealthy = tasty intuition is implicit.

Theoretical Implications

Validity of the intuition and reasons for its propagation. In light of our findings, an important question that emerges is whether the unhealthy = tasty intuition is valid, that is, whether the healthiness and tastiness of food items are indeed negatively correlated with each other in the real world. Although the intuition appears to enjoy widespread subscription (e.g., Keller, Sternthal, and Tybout 2002), there is little scientific evidence to support the view that tastiness and healthiness are negatively correlated with each other. Indeed, from an evolutionary standpoint, evidence points to the opposite, namely, that tastiness and healthiness are posi- tively correlated with each other. For example, in general, it is accepted among researchers (e.g., Drewnowski 1997; Smith 2004) that certain food groups (e.g., carbohydrates, fat) are perceived as tastier precisely because these foods have proved effective for survival; in these researchers' view, tastiness has served humans well as a proxy for healthiness.

Why then do negative health consequences result from consuming the food groups (e.g., carbohydrates, fat) that are, evolutionarily speaking, considered healthy? Medical evidence suggests that it is the overconsumption, not mod- erate levels of consumption, of these food groups that leads to negative health consequences (e.g., Drewnowski 1997).6 Indeed, consuming fat and carbohydrates in moderation has been found to have salutary effects on health (Better Health Channel 2004). Consistent with this view, findings indicate that overconsumption of any food group can result in nega- tive health consequences.

There is yet another reason to question the validity of the unhealthy = tasty intuition. Note that to assess the actual relationship between healthiness and tastiness, it is first necessary to measure the healthiness and tastiness of van- ous food options accurately. We believe that this is a decep- tively difficult task because neither judgments of healthi- ness nor those of tastiness are likely to be stable across time or across people. Oakes's (2005) findings demonstrate how judgments of healthiness are notoriously susceptible to con- textual influences. In one study, for example, Oakes finds that a lower-calorie snack portrayed as being "disreputable" in terms of health (i.e., a Snickers miniature bar with 47 calories) was inferred to promote greater weight gain than one portrayed as being more "reputable" in terms of health (i.e., a half cup of green peas with 62 calories). Thus, par- ticipants in this study relied on their subjective perceptions of the snacks' healthiness rather than on more diagnostic information (the snacks' caloric value) to infer their poten- tial for weight gain. There is reason to believe that judg- ments of tastiness may be similarly susceptible to contex- tual influences. For example, Carpenter and Nakamoto's (1989) findings suggest that what people happen to taste

6Needless to say, underconsumption of nutrients can also lead to negative health consequences. Indeed, malnourishment is the world's deadliest cause of premature death (World Health Organi- zation 2002). However, because this research is restricted to the context of the United States, where food is plentiful, we restrict our discussion to negative consequences of overconsumption.

The Unhealthy = Tasty Intuition 1179

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

first (e.g., Coke or Pepsi) may become their preferred option. An overview of other relevant findings suggests that there is little agreement (among adults) about tastiness rat- ings of food (see, e.g., Drewnowski 1997, p. 240).

Nonconscious influence. It is well known that con- sumers' decisions are influenced by their lay theories and intuitions even when the lay theories are invalid (Broniar- czyk and Alba 1994; Shiv, Carmon, and Ariely 2005). What is noteworthy in the current research is that our predictions were confirmed across two different types of food products, three types of judgment tasks, and two conceptually differ- ent types of unhealthiness manipulations. The generality of our findings across these experimental differences hints at the prevalence and strength with which people believe in the unhealthy = tasty intuition. Even more intriguing, how- ever, is the indication of the implicit nature of the influence of the intuition. The results from Experiment 1 confirmed that people may have implicitly learned to correlate unhealthiness with tastiness. The results from Experiments 2 and 3 offer conceptual replication by showing that even participants who reported disagreeing with the view that healthiness and tastiness are inversely correlated behaved as if they did. These findings indicate that the influence of the intuition may lie outside the awareness of consumers mak- ing judgments about what to consume. At a conceptual level, this notion shares similarities with Lewicki and col- leagues' (e.g., Lewicki 1986; Lewicki and Hill 1989; Lewicki, Hill, and Czyzewska 1997) work on nonconscious learning. The idea that a person's judgments and decisions may be influenced by factors outside his or her awareness is also echoed in various other findings, such as mere expo- sure effects (Zajonc 1980), truth effects (Hawkins and Hoch 1992), and assimilation/contrast effects (e.g., Schwarz and Bless 1992).

Managerial, Consumer, and Public Policy Implications Although the obesity problem afflicting people in the United States may be due, in part, to the ubiquity of food high in fat content, our findings indicate that even if options with less fat were available, consumers may choose to con- sume foods high in fat because of a misguided notion that they need to eat unhealthy to eat tasty or that eating healthy means having food that is not tasty. Evidence from food sci- entists suggests that despite current concerns about healthy eating, health is secondary to taste in selecting many food products (for a discussion of taste versus health in con- sumer selection of corn chips, see Tepper and Trail 1998). There is also evidence in recent research on functional food -a new category of products that promise consumers improvements in targeted physiological functions, such as lowering cholesterol (Urala and Lahteenmaki 2004)--that consumers are unwilling to compromise on taste even for the sake of the health benefits these foods offer (Verbeke 2006). This lesson was perhaps most starkly revealed in the interesting story of the McLean, a low-fat burger that McDonald's tried to market in the early 1990s. Although the burger was actually preferred to burgers that contained more fat in blind taste tests, it was a miserable market fail- ure (Gladwell 2001).

Given our evidence that the influence of the unhealthy = tasty intuition is strong and may take place outside the con- sumer's awareness, this intuition could be a major con- tributing factor to obesity in consumers. Therefore, a key question is, What (if anything) can be done to counteract its potentially devastating impact on consumer food choice decisions and, subsequently, on consumer health?

More research is being directed at understanding both how to market products such that consumers make better food choices (Andrews, Netemeyer, and Burton 1998; Kozup, Creyer, and Burton 2003; Seiders and Petty 2004; Wansink and Huckabee 2005) and how to help consumers make changes in their environment to help control their food consumption (Wansink 2004a). These streams of research offer guidance to the food industry, which has been accused of contributing to the obesity epidemic, and to con- sumers who want to make healthier choices that are still enjoyable, despite the pervasiveness of a belief in the unhealthy = tasty intuition. Possibilities for controlling the effect of the unhealthy = tasty intuition (and its potential for causing negative health consequences) include (1) control- ling the volume of unhealthy but tasty food that is eaten, (2) changing unhealthy foods to make them less unhealthy but still tasty, and (3) providing consumers with better informa- tion about what constitutes "healthy." We address each of these possibilities in the remainder of this discussion.

Limiting consumption volume of unhealthy foods. If consumers infer that unhealthy foods are tastier, they will possess a strong desire to eat foods that are relatively unhealthy (e.g., foods that are high in fat or high in calo- ries). A possible solution to this problem would be for both marketers and consumers to take steps to ensure that if these foods are going to be eaten, they are eaten in reason- able quantities. Such steps are important because, as we mentioned previously, medical evidence suggests that it is the overconsumption, not moderate levels of consumption, of foods considered unhealthy that leads to negative health consequences (e.g., Drewnowski 1997). Therefore, mea- sures aimed at restricting consumption of foods considered unhealthy (but tasty) to moderate levels can go a long way toward counteracting the potential negative health conse- quences of a belief in the unhealthy = tasty intuition.

Responsible marketers and individual consumers can take such steps to control consumption volume of unhealthy foods. For example, Wansink and Huckabee (2005) suggest that marketers can guide consumers' volume decisions by changing packaging (e.g., offering multipacks with smaller individual servings, offering premium-priced smaller pack- aging), changing the volume of implied serving sizes in packaging information (e.g., designating a serving as three cookies versus six), and providing consumers with easier ways to assess how much they are consuming (e.g., placing markers on the outsides of packaging to indicate a serving size, placing indicator lines inside packaging). Consumers can also make changes to their personal environment to help reduce consumption.

Wansink (2004a) offers many practical alterations that can be made to both the eating and the food environments, including pre-serving portions when snacking while dis- tracted (e.g., while watching television); repackaging foods

180 I Journal of Marketing, October 2006

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

into small containers to suggest smaller consumption norms; never eating directly from a package; and using smaller plates, bowls, and eating utensils. Such steps can allow marketers to continue to market unhealthy foods in responsible ways and allow consumers to continue to enjoy the foods they consider tasty, albeit in moderation. Although many of these changes seem small, Wansink and Huckabee (2005) point out that 80% of the population gain weight because of a calorie excess of less than 50 calories a day. The changes suggested here could allow consumers to enjoy the foods they consider tasty in smaller, controlled quantities, perhaps enough to avoid consumption of those additional 50 calories.

Reformulating unhealthy foods. Another possibility for controlling the effects of the unhealthy = tasty intuition is to change the composition of unhealthy foods. One alternative is to reformulate high-energy-density foods (i.e., foods high in calories) to lower their energy density (i.e., calorie con- tent) by replacing some of the fat with water, fiber filler, or vegetables (Wansink and Huckabee 2005). Although evi- dence from our studies suggests that people believe that food cannot be made healthier without sacrificing taste (and, thus, that these lower-energy-density foods would be less appetizing), taste tests indicate otherwise. Rolls, Ello- Martin, and Tohill's (2004) research suggests that con- sumers who eat lower-energy-density foods eat the same volume as they would of the unmodified food (resulting in fewer calories consumed), rate themselves as equally satis- fled as those who eat full-calorie dense foods, and, most important in our research context, do not perceive the foods as tasting worse (Kral and Rolls 2004; Rolls 2005; Wansink and Huckabee 2005). Research suggests that up to 20% of the fat in a high-energy-density food can be replaced with low-density items (e.g., fruits, vegetables) without con- sumers noticing a difference in taste (Rolls, Ello-Martin, and Tohill 2004).

Such findings suggest that many consumers are not knowledgeable enough about food composition to realize that the unhealthy foods they love can be successfully refor- mulated to be just as tasty without being as unhealthy. Because such modifications could have a large impact on the number of calories consumed, they could be very bene- ficial if consumers are educated to believe that the healthier version of their favorite unhealthy food is just as tasty. However, because of the pervasiveness of the unhealthy = tasty intuition, such a strategy could backfire (i.e., despite research evidence suggesting the taste had not truly changed, consumers could respond to such a modification as they did in Experiment 3, in which they allowed the knowledge that a food is now healthier to color their actual taste perceptions).

To combat this problem, an obvious solution would be to stress both taste and health for these items, with the taste dimension being the dominant dimension. By stressing the taste dimension first, companies could work on breaking down the unhealthy = tasty intuition and increasing the probability of choice. Marketing strategies involving sam- pling, credible sources, and opinion leaders would be criti- cal. First, sampling would be a key factor in gaining accep- tance for these new, healthier products because consumers

will not believe that they are tasty unless they have tried the product themselves. A free sample represents a low-risk way to experience the taste. These samples could be distrib- uted through the mail or, more appropriately, handed out in health clubs or at sporting/fitness events.

Second, well-respected health or fitness experts could be employed to endorse these products. To change engrained beliefs, a credible source is needed. Consumers are more likely to accept a message about a product when the source is considered trustworthy and viewed as an expert on the topic (e.g., Goldsmith, Lafferty, and Newell 2000; Sternthal, Dholakia, and Leavitt 1978). This technique could be par- ticularly effective if these sources are opinion leaders. For example, health club employees and trainers are often looked to by consumers for tips on exercise and diet. A mar- keting or public policy program could educate and persuade these people on the tastiness and health benefits of the new products. They could then be important transmitters of this information to consumers. Thus, this message would be more likely to be accepted and acted on.

Another possibility might be to stimulate "buzz" for certain products among consumers. If marketers could con- vince a key group of consumers in the word-of-mouth net- work that their food products taste good and are part of a healthy diet, this group could influence and excite other groups of consumers. A similar phenomenon occurred in the fast and widespread acceptance of the Atkins diet, in which consumers were willing to give up foods high in car- bohydrates for quick weight loss.

Reeducating consumers about what is "healthy." A final avenue for counteracting the effects of the unhealthy = tasty intuition is to educate consumers about what constitutes healthy. Determining what is healthy or unhealthy is a deceptively difficult task because, as we discussed previ- ously, healthiness is not an inherent quality of food; rather, it depends on the conditions (e.g., quantity, genetic makeup of the person consuming it) under which it is consumed. Nevertheless, consumers can be educated to approach nutri- tion in a more balanced fashion than they currently do. In Oakes's (2005) study, which we described previously, con- sumers overemphasized fat content when they assessed the healthiness of foods and relied on a categorization system (of foods as bad versus good) using stereotypes about types of food/ingredients to determine how much weight a food would cause someone to gain. Had they relied instead on a more complete nutritional assessment of the foods, they would have found that the food they considered healthy on the basis of fat content (peas) would actually lead to more weight gain than the food they considered unhealthy (Snickers) because of the caloric content of the two foods.

What can be done to convince consumers to consider all types of nutritional information? A possibility is to develop an easy-to-use nutrition index that combines the various types of relevant information into a single number, which consumers could then use to make easy comparisons across different types of foods. This would enable consumers to choose healthier food with minimal effort and confusion. Furthermore, reeducating consumers to evaluate the health- iness of a food could actually change consumers' percep- tions about what constitutes healthy and unhealthy.

The Unhealthy =Tasty Intuition 1181

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

Limitations and Further Research

An interesting implication of our main hypothesis (that people subscribe to the unhealthy = tasty intuition) is that regardless of which food group is associated with unhealth- iness, food items that are higher (versus lower) in that ingredient will be inferred to be tastier. Thus, for example,

in contexts in which carbohydrates are considered unhealthy (as is likely to be the case among those following the Atkins diet), food items higher (versus lower) in carbo- hydrates will be (1) inferred to taste better, (2) enjoyed more, and (3) preferred when a hedonic goal is more (ver- sus less) salient. Indeed, the results from Experiment 3, in which no mention was made of the particular ingredients

APPENDIX Experiment 1: Stimuli Used in IAT

Word List

Unhealthy Foods Healthy Foods Enjoyable Not Enjoyable

Tasty Disliked Delicious Less tasty Yummy Unappealing

Appetizing Bland Flavorful Flavorless Appealing Unappetizing

Mouthwatering Unpalatable

1821 Journal of Marketing, October 2006

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

that made the Mango milkshake more or less healthy, offer conceptual support for this hypothesis.

An important related empirical question that is still unanswered by the current research is whether other manipulations of unhealthiness would produce similar results in terms of taste inferences, enjoyment, and choice. For example, would altering the content of preservatives (assuming that they are considered unhealthy) alter percep- tions of tastiness in a manner consistent with the unhealthy = tasty intuition? If not, identifying the manipula- tions that lead to the types of results we document in this research would be important for both theoretical and sub- stantive reasons. Similarly, further research should ascertain the boundaries of the content of unhealthy ingredients within which results consistent with the unhealthy = tasty intuition will be obtained. Is there a limit beyond which increasing unhealthiness does not lead to enhanced infer- ences of tastiness? Levin and Gaeth's (1988) results suggest that there is. They find that beef that was portrayed as 25%

fat was expected to taste worse than beef that was portrayed as 75% lean. These results indicate that there is a range within which increasing the portrayed unhealthiness of food leads to enhanced perceptions of tastiness but also a range beyond which it does not.

A final and interesting extension of our work would be to assess whether and to what extent socioeconomic and cultural factors moderate our effects. In contexts in which food is not plentiful (e.g., in developing countries, such as India or China, and in underdeveloped countries, such as Somalia or Cambodia), people may believe tastiness is posi- tively correlated with healthiness, as evolution intended (Drewnowski 1997; Smith 2004). If so, we may expect to find a pattern of results that is opposite to that which we find in this research. Similarly, in epicurean cultures (e.g., France), in which food is valued primarily for its tastiness, it is possible that the perception of an inverse relationship between healthiness and tastiness is held with less conviction.

REFERENCES Allen, Chris T. and Chris A. Janiszewski (1989), "Assessing the

Role of Contingency Awareness in Attitudinal Conditioning with Implications for Advertising Research," Journal of Mar- keting Research, 26 (February), 30-43.

Andrews, J. Craig, Richard G. Netemeyer, and Scot Burton (1998), "Consumer Generalization of Nutrient Content Claims in Advertising," Journal of Marketing, 62 (October), 62-75.

Berry, Christopher J. (1994), The Idea of Luxury. Cambridge, UK: Cambridge University Press.

Better Health Channel (2004), "Sugar," (accessed February 12, 2004), [available at http://www.betterhealth.vic.gov.aulbhcv2/ bhcarticles.nsf/pages/Sugar?OpenDocument].

Bettman, James R., Mary Frances Luce, and John W. Payne (1998), "Constructive Consumer Choice Processes," Journal of Consumer Research, 25 (December), 187-217.

Broniarczyk, Susan M. and Joseph W. Alba (1994), "The Role of Consumers' Intuitions in Inference Making," Journal of Con- sumer Research, 21 (December), 393-407.

Brownell, Kelly D. and Katherine Battle Horgen (2003), Food Fight: The Inside Story of the Food Industry, America's Obesity Crisis, and What We Can Do About It. New York: McGraw- Hill.

Carpenter, Gregory S. and Kent Nakamoto (1989), "Consumer Preference Formation and Pioneering Advantage," Journal of Marketing Research, 26 (August), 285-98.

Deighton, John (1984), "The Interaction of Advertising and Evi- dence," Journal of Consumer Research, 11 (December), 763-70.

Dhar, Ravi and Itamar Simonson (1999), "Making Complemen- tary Choices in Consumption Episodes: Highlighting Versus Balancing," Journal of Marketing Research, 36 (February), 29-44.

Drewnowski, Adam (1997), "Taste Preferences and Food Intake," Annual Review of Nutrition, 17, 237-53.

Food Marketing Institute (1996), Trends in the United States: Con- sumer Attitudes and the Supermarket. Chicago: Food Market- ing Institute.

Gladwell, Malcolm (2001), "The Trouble with Fries: Fast Food Is Killing Us. Can It Be Fixed?" The New Yorker, (March 5), (accessed February 12, 2004), [available at http://www. gladwell.com/2001/2001_03_05_a_fries.htm].

Goldsmith, Ronald E., Barbara A. Lafferty, and Stephan J. Newell (2000), "The Impact of Corporate Credibility on Consumer

Reactions to Advertisements and Brands," Journal of Advertis- ing, 29 (3), 43-54.

Greenwald, Anthony G., Debbie E. McGhee, and Jordan L.K. Schwartz (1998), "Measuring Individual Differences in Implicit Cognition: The Implicit Association Test,' Journal of Personality and Social Psychology, 74 (6),1464-80.

Brian Brian A. Nosek, and Mahzarin R. Banaji (2003), "Under- standing and Using the Implicit Association Test: I. An Improved Scoring Algorithm," Journal of Personality and Social Psychology, 85 (2),197-216.

Harvard School of Public Health Nutrition Source (2004), "Fats and Cholesterol: The Good, the Bad, and the Healthy Diet," (accessed February 12, 2004), [available at http://www.hsph. harvard.edu/nutritionsource/fats.html].

Hawkins, Scott A. and Stephen J. Hoch (1992), "Low Involvement Learning: Memory Without Evaluation,' Journal of Consumer Research, 19 (September), 212-25.

Hoch, Stephen J. and Young-won Ha (1986), "Consumer Learn- ing: Advertising and the Ambiguity of Product Experience," Journal of Consumer Research, 13 (September), 221-33.

Irwin, Julie R. and Gary H. McClelland (2001), "Misleading Heuristics and Moderated Multiple Regression Models," Jour- nal of Marketing Research, 38 (February), 100-109.

Keller, Kevin Lane, Brian Sternthal, and Alice Tybout (2002), "Three Questions You Need to Ask About Your Brand," Har- vard Business Review, 80 (September), 80-85.

Kelly, Harold H. and John W. Thibaut (1969), "Group Problem Solving," in The Handbook of Social Psychology, 2d ed., Gard- ner Lindsey and Elliot Aronson, eds. Reading, MA: Addison- Wesley, 1-101.

Kivetz, Ran and Itamar Simonson (2002), "Earning the Right to Indulge: Effort as a Determinant of Customer Preferences Toward Frequency Program Rewards," Journal of Marketing Research, 39 (May), 155-70.

Kozup, John C., Elizabeth H. Creyer, and Scot Burton (2003), "Making Healthful Food Choices: The Influence of Health Claims and Nutrition Information on Consumers' Evaluations of Packaged Food Products and Restaurant Menu Items," Jour- nal of Marketing, 67 (April), 19-34.

Kral, Tanja V. and Barbara J. Rolls (2004), "Energy Density and Portion Size: Their Independent and Combined Effects on Energy Intake," Physiology and Behavior, 82 (1),131-38.

The Unhealthy . Tasty Intuition 1183

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

Levin, Irwin P. and Gary J. Gaeth (1988), "How Consumers Are Affected by the Framing of Attribute Information Before and After Consuming the Product," Journal of Consumer Research, 15 (December), 374-78.

Lewicki, Pawel (1986), "Processing Information about Covaria- tions That Cannot Be Articulated," Journal of Experimental Psychology: Learning, Memory, and Cognition, 12 (January), 135-46.

and and Thomas Hill (1989), "On the Status of Nonconscious Processes in Human Cognition: Comment on Reber," Journal of Experimental Psychology: General, 118 (3), 239-41.

and and and Maria Czyzewska (1997), "Hidden Covaria- tion Detection: A Fundamental and Ubiquitous Phenomenon," Journal of Experimental Psychology: Learning, Memory, and Cognition, 23 (January), 221-28.

Maslow, Abraham (1962), Toward a Psychology of Being. Oxford: Van Nostrand.

Morris, Michael W., Tanya Menon, and Daniel R. Ames (2001), "Culturally Conferred Conception of Agency: A Key to Social Perception of Persona, Groups, and Other Actors," Journal of Personality and Social Psychology, 5 (April), 169-82.

Nestle, Marion (2003), Food Politics: How the Food Industry Influences Nutrition and Health. Berkeley: University of Cali- fornia Press.

Oakes, Michael E. (2005), "Beauty or Beast: Does Stereotypical Thinking About Foods Contribute to Overeating?" Food Qual- ity and Preference, 16 (July), 447-54.

and and Carole S. Slotterback (2005), "Too Good to Be True: Dose Insensitivity and Stereotypical Thinking of Foods' Capacity to Promote Weight Gain," Food Quality and Prefer- ence,l6 (December), 675-81.

Parloff, Roger (2003), "Is Fat the Next Tobacco?" Fortune, (Sep- tember), 50-54.

Raghunathan, Rajagopal and Kim Corfman (2006), "Is Happiness Shared Doubled and Sadness Shared Halved? Social Influence on Enjoyment of Hedonic Experiences," Journal of Marketing Research, 43 (August), 386-94.

Rolls, Barbara (2005), The Volumetrics Eating Plan: Techniques and Recipes for Feeling Full on Fewer Calories. New York: HarperCollins.

Julie Julie A. Ello-Martin, and Beth C. Tohill (2004), "What Can Intervention Studies Tell Us About the Relationship Between Fruit and Vegetable Consumption and Weight Man- agement?" Nutrition Reviews, 62 (1),1-17.

Ross, Lee and Richard E. Nisbett (1991), The Person and the Sit- uation: Perspectives of Social Psychology. New York: McGraw-Hill.

Schlosser, Eric (2001), Fast Food Nation: The Dark Side of the All-American Meal. Boston: Houghton Mifflin.

Schwarz, Norbert and Herbert Bless (1992), "Constructing Reality and Its Alternatives: An Inclusion/Exclusion Model of Assimi- lation and Contrast Effects in Social Judgment," in The Con- struction of Social Judgments, Leonard L. Martin and Abraham

Tesser, eds. Hillsdale, NJ: Lawrence Erlbaum Associates, 217-45.

Seiders, Kathleen and Ross D. Petty (2004), "Obesity and the Role of Food Marketing: A Policy Analysis of Issues and Reme- dies," Journal of Public Policy & Marketing, 23 (Fall), 153-69.

Shiv, Baba, Ziv Carmon, and Dan Ariely (2005), "Placebo Effects of Marketing Actions: Consumers May Get What They Pay For," Journal of Marketing Research, 42 (November), 383-93.

and and Alexander Fedorikhin (1999), "Heart and Mind in Conflict: The Interplay of Affect and Cognition in Consumer Decision Making," Journal of Consumer Research, 26 (December), 278-92.

Smith, Trenton G. (2004), "The McDonald Equilibrium: Advertis- ing, Empty Calories, and the Endogenous Determination of Dietary Preferences," Social Choice and Welfare, 23 (3), 383-413.

Sternthal, Brian, Ruby R. Dholakia, and Clark Leavitt (1978), "The Persuasive Effect of Source Credibility: A Situation Analysis," Public Opinion Quarterly, 42 (Fall), 285-314.

Sujan, Mita and Christine Dekleva (1987), "Product Categoriza- tion and Inference Making: Some Implications for Compara- tive Advertising," Journal of Consumer Research, 14 (Decem- ber), 372-78.

Tepper, Beverly J. and Amy C. Trail (1998), "Taste or Healthy: A Study on Consumer Acceptance of Corn Chips," Food Quality and Preference, 9 (4), 267-72.

Urala, Nina and Lisa Lahteenmaki (2004), "Attitudes Behind Con- sumers' Willingness to Use Functional Foods," Food Quality and Preference, 15 (October -December), 793-803.

Verbeke, Wim (2006), "Functional Foods: Consumer Willingness to Compromise on Taste for Health?" Food Quality and Prefer- ence,l7 (1-2),126-31.

Wansink, Brian (2004a), "Environmental Factors That Increase the Food Intake and Consumption Volume of Unknowing Con- sumers," Annual Review of Nutrition, 24, 455-79.

(2004b), (2004b), Marketing Nutrition. Champaign: University of Illinois Press.

and and Pierre Chandon (2006), "Can ̀ Low Fat' Nutrition Labels Lead to Obesity?" Journal of Marketing Research, 43 (November), forthcoming.

and and Mike Huckabee (2005), "De-Marketing Obesity," California Management Review, 47 (Summer), 6-18.

Weber, Max (1998), The Protestant Work Ethic and the Spirit of Capitalism, 2d ed. Los Angeles: Roxbury.

World Health Organization (2002), The World Health Report, (accessed January 21, 2005), [available at http:/lwww.who.int! whr/2002/en/] .

Zajonc, Robert B. (1980), "Feeling and Thinking: Preferences Need No Inferences," American Psychologist, 35 (February), 151-75.

Zhang, Jane (2005), "House Passes ̀Cheeseburger' Responsibility Bill," The Wall Street Journal, (October 20), A4.

184 I Journal of Marketing, October 2006

This content downloaded by the authorized user from 192.168.52.62 on Thu, 22 Nov 2012 03:40:05 AM All use subject to JSTOR Terms and Conditions

  • Article Contents
    • p. 170
    • p. 171
    • p. 172
    • p. 173
    • p. 174
    • p. 175
    • p. 176
    • p. 177
    • p. 178
    • p. 179
    • p. 180
    • p. 181
    • p. 182
    • p. 183
    • p. 184
  • Issue Table of Contents
    • Journal of Marketing, Vol. 70, No. 4 (Oct., 2006), pp. 1-210
      • Volume Information
      • Front Matter
      • Corporate Social Responsibility, Customer Satisfaction, and Market Value [pp. 1-18]
      • The Influence of Avatars on Online Consumer Shopping Behavior [pp. 19-36]
      • Creating a Market Orientation: A Longitudinal, Multifirm, Grounded Analysis of Cultural Transformation [pp. 37-55]
      • It Just Feels Good: Customers' Affective Response to Touch and Its Influence on Persuasion [pp. 56-69]
      • Marketing's Credibility: A Longitudinal Investigation of Marketing Communication Productivity and Shareholder Value [pp. 70-91]
      • Upgrades and New Purchases [pp. 92-102]
      • Influence Tactics for Effective Adaptive Selling [pp. 103-117]
      • How Biased Household Inventory Estimates Distort Shopping and Storage Decisions [pp. 118-135]
      • Factors Influencing the Effectiveness of Relationship Marketing: A Meta-Analysis [pp. 136-153]
      • Achieving Marketing Objectives through Social Sponsorships [pp. 154-169]
      • The Unhealthy = Tasty Intuition and Its Effects on Taste Inferences, Enjoyment, and Choice of Food Products [pp. 170-184]
      • Broadening the Scope of Reference Price Advertising Research: A Field Study of Consumer Shopping Involvement [pp. 185-204]
      • Back Matter