Professional Documents
Culture Documents
Get Your Science Out of Here: When Does Invoking Science in The Marketing of Consumer Products Backfire?
Get Your Science Out of Here: When Does Invoking Science in The Marketing of Consumer Products Backfire?
In this research, we propose that although consumers view the scientific process
as competent, they also perceive it as cold. Across 10 experimental studies, we
demonstrate that these lay beliefs impact consumers’ reactions to marketers tout-
ing the science behind their brands. Specifically, since hedonic attributes are as-
sociated with warmth, the coldness associated with science is conceptually disflu-
ent with the anticipated warmth of hedonic products and attributes, reducing
product valuation. In contrast, when products are positioned as utilitarian, invoking
science in marketing appeals has a positive effect, as the perceived competence
of the scientific process is more compatible with the competence associated with
utilitarian products. We further demonstrate that when the necessity of science to
create a hedonic product is made salient and thus more fluent, this backfire effect
is attenuated. Finally, we identify three theoretically and practically relevant indi-
vidual differences (endorsement of the lay belief, trust in scientists, and whether
the consumer works in a STEM field) that moderate the backfire effect of pairing
science with hedonically positioned products.
Keywords: perceptions of science, science appeals, lay belief, lay theory, hedonic
versus utilitarian, marketing communications
721
722 JOURNAL OF CONSUMER RESEARCH
increasing consumers’ product valuation and when they Raghunathan, Naylor, and Hoyer 2006; Zane, Smith, and
backfire. Reczek 2020) by introducing a novel lay belief about a do-
Over the past four decades, Americans have generally main that is becoming more relevant as consumers increas-
viewed science as beneficial (Krause et al. 2019), but, re- ingly rely on products produced by science even while
cently, a growing number of Americans report lower trust having mixed feelings about science itself. Finally, our
in science (American Academy of Arts and Sciences 2018; work also offers practical guidance to marketers on when
Gauchat 2012). There is therefore reason to believe that and why science appeals may backfire in the marketplace.
consumers may hold mixed beliefs about the use of the sci-
entific process in bringing products to market. In this work, SCIENCE APPEALS
we propose that although consumers view the scientific
products that the research on perceived artificiality of prod- 2006). In fact, past research has shown that when a prod-
ucts would define as inherently “unnatural” because they uct’s attributes conflict with a lay belief about the relation-
are produced using human intervention. Thus, we are inter- ship between those attributes, consumers engage in
ested in the backfiring of science appeals beyond reasons increased information search because the information is
of unnaturalness (although we empirically rule out artifici- discrepant from what they expect (Haws et al. 2017). We
ality as an alternative explanation for the effects we ob- propose that the lay belief that science is cold yet compe-
serve in our studies). tent creates expectations that can either match or mismatch
with products that are positioned as hedonic or utilitarian.
CONSUMER LAY BELIEFS ABOUT THE Consumers tend to associate hedonic brands, products, and
SCIENTIFIC PROCESS attributes (which focus on positive sensory and affective
2021), similar to the dampening effects on product evalua- scientific process is cold yet competent and the expectation
tions when something does not “feel right” (e.g., a non- that hedonic products are warm. This theorizing suggests
rounded number in Wadhwa and Zhang 2015). two moderators for this backfire effect: (1) altering the
Thus, when science is invoked to market warm, hedonic associations consumers have with a specific hedonic prod-
products, we predict consumers should be less interested in uct and (2) altering belief in the science is cold yet compe-
those products.3 On the other hand, when science is in- tent lay belief. We first focus on altering associations
voked to market utilitarian products, conceptual fluency consumers have with hedonic products. In our experiments,
should be increased because the functional, competent ori- we focus on product categories where science is not natu-
entation of utilitarian products is consistent with the lay be- rally highly salient, as these are the products for which we
lief that science is competent, resulting in increased predict science appeals will backfire. Specifically, we fo-
differ in the extent to which they naturally endorse a given share (study 1a) and willingness to pay (WTP) (study 1b)
lay belief (Cheng et al. 2017; Mukhopadhyay and Johar are reduced when a hedonic product is promoted with a sci-
2005; Raghunathan et al. 2006). We propose that individu- ence appeal (vs. a control appeal with no mention of sci-
als who do not explicitly endorse the lay belief that science ence). Study 1c further demonstrates that the backfire
is cold will not devalue hedonic products that invoke sci- effect we observe is uniquely driven by referencing sci-
ence since, without this belief, science appeals will not be ence, as an appeal referencing rigor (but not science) in a
conceptually disfluent. Such a finding would also provide product development process does not produce the same
further evidence for our proposed process. We therefore negative effect on product valuation. Study 2 provides evi-
predict: dence that the science backfire effect for hedonic products
is driven by conceptual disfluency (hypothesis 2). Studies
H4: Those who do not hold the lay belief that science is
F(1, 311) ¼ 1753.10, p < .001, gp2 ¼ 0.58), while the condition where the description of Option A included a ref-
warmth/coldness of the scientific process was significantly erence to science, Option A was chosen significantly less
below the midpoint (M ¼ 1.26, F(1, 311) ¼ 76.74, p < often compared to the control condition (21.18% vs.
.001, gp2 ¼ 0.06). Consistent with our theorizing, these 30.47%; b ¼ .25, 2(1) ¼ 5.71, p ¼ .017, u ¼ 0.11). In
results suggest that consumers view the scientific process terms of percentage change [(Science Control)/Control],
as competent but cold. adding a science appeal decreased the likelihood of choos-
ing cookie Option A by 30.48%. Thus, study 1a shows that
STUDIES 1A–1C a science appeal significantly reduces consumers’ choice
of the cookie.
Study 1a
study 1a with a different operationalization of product val- appeal referencing science does so. Additionally, this study
uation, demonstrating that consumers are willing to pay provides evidence that the science backfire effect goes be-
less for a case of hedonic cookies when they are promoted yond cognitive/affective matching because simply appeal-
using a science appeal. ing to a cognitive attribute (i.e., a rigorous development
process) was insufficient to yield a backfire effect com-
Study 1c pared to the control condition.
The goal of study 1c was to demonstrate that the back-
fire effect for hedonic positioning is unique to explicitly STUDY 2
referencing science (vs. a rigorous development process
In study 2 (preregistration https://aspredicted.org/blind.
that does not invoke science) because it is driven by
smoothies are artificial?,”7 “How natural do you think JTB 0.01]). To test the relative indirect effect size of taste ver-
smoothies are?” (reverse coded), and “To what extent do sus fluency in our parallel mediation model, we also con-
you think JTB smoothies have additives?”). Finally, we ducted a pairwise comparison using the “contrast”
measured the extent to which participants thought the command through the PROCESS macro (Hayes 2017).
smoothie was tasty, healthy, and conducive to weight loss When the indirect effects of taste and fluency were directly
(from Andre et al. 2019), as well as whether the slogan was contrasted, fluency mediated to a significantly greater de-
plausible, that is it “seem[ed] like a real slogan companies gree (95% CI: [0.53, 0.24]). See web appendix D for a
might use” (1 ¼ “Not at all”; 7 ¼ “Very much so”). We full reporting of means, ANOVAs, and mediation analysis
measured plausibility because we propose that the effect of for these additional variables.
invoking science on product valuation is driven by the con-
Discussion. Study 2 provides support for both hypothe-
FIGURE 1
Science vs. Control) 2 (Positioning: Hedonic vs. they or their romantic partner used birth control, completed
Utilitarian) between-subjects design. In one between- a data validity indicator, and responded to demographic
subjects factor, half the participants were told to imagine questions and several exploratory measures (reported in
they had the goal of finding a lubricant that enhanced sex- web appendix B).
ual pleasure (the hedonic condition), and the other half
were told to imagine they were searching for a lubricant Results and Discussion
that would add an additional layer of birth control protec-
tion (the utilitarian condition). Participants in the hedonic Purchase Intentions. We subjected the data to a two-
condition then saw an appeal that positioned the lubricant way ANOVA predicting purchase intentions from appeal
as hedonic (increasing sexual pleasure), whereas partici- type (control ¼ 1, science ¼ 1), product positioning
pants in the utilitarian condition saw an appeal that posi- (hedonic ¼ 1, utilitarian ¼ 1), and the two-way interac-
tioned the lubricant as utilitarian (increasing birth control tion. There was no main effect of appeal type (F(1, 819) ¼
protection). In the other between-subjects factor, half the 0.70, p ¼ .402, gp2 ¼ 0.001), but there was a main effect
participants saw an appeal that invoked science, and the of product positioning (F(1, 819) ¼ 6.69, p ¼.01, gp2 ¼
other half saw a control appeal that did not mention sci- 0.01), such that when the lubricant was positioned as utili-
ence. Participants then reported their likelihood of purchas- tarian, purchase intentions were higher (M ¼ 4.61) com-
ing this lubricant using the same purchase intentions item pared to when it was positioned as hedonic (M ¼ 4.33).
and conceptual fluency measures (a ¼ 0.88) as study 2. Most relevant to our hypotheses, there was a significant
Participants also reported the extent to which they trusted two-way interaction (F(1, 819) ¼ 13.54, p < .001, gp2 ¼
the lubricant brand using two items: (r ¼ 0.87; “How trust- 0.02; figure 2). Simple effects revealed that when the lubri-
worthy do you think the makers of Noma lubricant are?” cant was positioned as enhancing sexual pleasure, the con-
and “How honest do you think the makers of Noma lubri- trol appeal (M ¼ 4.48) led to higher purchase intentions
cant are?”), as well as how competent they thought the than the science appeal (M ¼ 4.17; F(1, 819) ¼ 4.05, p ¼
brand was (“How competent do you think the makers of .045, gp2 ¼ 0.01). In contrast, when the lubricant was posi-
Noma lubricant are?”). Participants next reported whether tioned as providing birth control, the science appeal
730 JOURNAL OF CONSUMER RESEARCH
FIGURE 3
CONCEPTUAL FLUENCY MEDIATES THE APPEAL TYPE BY PRODUCT POSITIONING INTERACTION ON PURCHASE INTENTIONS
Results
FIGURE 6
INVOKING SCIENCE WHEN MARKETING A HEDONIC BODY WASH BACKFIRES UNLESS CONSUMERS SEE SCIENCE AS
RELATIVELY WARM (A) OR HAVE HIGH LEVELS OF TRUST IN SCIENTISTS (B). (A) MODERATION BY ENDORSEMENT OF SCIENCE IS
COLD LAY BELIEF. (B) MODERATION BY TRUST IN SCIENTISTS.
(a) (b)
6
5
Purchase Intentions
Purchase Intentions
4
4
3
Control Control
Science Science
2
−4 −2 0 2 4 2 4 6 8
Endorsement of Lay Intuition Trust in Scientists
GENERAL DISCUSSION
Across 10 studies, we found evidence that invoking sci-
downstream effects of this lay belief. In doing so, we not detrimental in promoting various products. Our findings
only contribute to the literature on consumer lay theories suggest that it would behoove marketers to invoke science
(Cheng et al. 2017; Deval et al. 2013; McFerran and when promoting products that have utilitarian attributes.
Mukhopadhyay 2013; Raghunathan et al. 2006; Zane et al. Likewise, for product categories where it is already natu-
2020) but also to the growing body of literature exploring rally salient to consumers that science is involved in pro-
how consumer response to products varies as a function of duction (e.g., highly technical products), invoking science
what is communicated to consumers about the product’s should not have harmful consequences. For CPGs and
origin and production process (Smith, Newman, and Dhar other categories where scientific development is less
2016; Newman and Dhar 2014; Reich, Kupor, and Smith expected, however, our findings suggest that invoking sci-
2018). ence to market these products might backfire. Thus, when
factors may be at play here and that experiences/messaging DATA COLLECTION INFORMATION
related to the scientific process can influence beliefs about
scientists, and vice versa. For example, personal experi- Studies 1b, 1c, 2, 3a, 3b, 4, 5a, 5b; the three pilots dis-
ence with science (e.g., in school) as a cold, careful process cussed in the introduction; the study 2 pretest, the study 3b
may influence perceptions of scientists. Further, research pretest, the two study 4 pretests, and the study 5a–5c pre-
has found that those who have a particular job, role, or en- test were collected by all three authors using Amazon
gage in particular behaviors tend to be perceived as pos- Mechanical Turk. Study 5c was collected by all three
sessing traits associated with that behavior or role (i.e., authors using Prolific. The study 1a pretest was collected
Spontaneous Trait Inference; Carlston and Skowronski at the Fisher Behavioral Lab at The Ohio State University
2005; Newman and Uleman 1990). Conversely, portrayals with the assistance of a lab manager and research assistants
Cadbury Dairy Milk (2010), “Cadbury Advert—Cadbury Fiske, Susan T., Amy J. C. Cuddy, and Peter Glick (2007),
Chocolate Charmer Ad,” Last Accessed May 17, 2022, “Universal Dimensions of Social Cognition: Warmth and
https://www.youtube.com/watch?v=Bv7mETANB7A. Competence,” Trends in Cognitive Sciences, 11 (2), 77–83.
Carlston, Donal E. and John J. Skowronski (2005), “Linking ver- Fuchs, Christoph, Martin Schreier, and Stijn M. J. van Osselaer
sus Thinking: Evidence for the Different Associative and (2015), “The Handmade Effect: What’s Love Got to Do with
Attributional Bases of Spontaneous Trait Transference and It?,” Journal of Marketing, 79 (2), 98–110.
Spontaneous Trait Inference,” Journal of Personality and Funk, Cary (2020), “Key Findings about Americans’ Confidence
Social Psychology, 89 (6), 884–98. in Science and their Views on Scientists’ Role in Society,”
Chattalas, Michael and Hirokazu Takada (2013), “Warm versus Pew Research Center, Last Accessed May 17, 2022, https://
Competent Countries: National Stereotyping Effects on www.pewresearch.org/fact-tank/2020/02/12/key-findings-
Expectations of Hedonic versus Utilitarian Product about-americans-confidence-in-science-and-their-views-on-
Properties,” Place Branding and Public Diplomacy, 9 (2), scientists-role-in-society/.
Ideology and Congruent Appeals,” Journal of Consumer Naylor, Rebecca Walker, Courtney M. Droms, and Kelly L. Haws
Research, 40 (2), 350–67. (2009), “Eating with a Purpose: Consumer Response to
Klein, Kristina and Valentyna Melnyk (2016), “Speaking to the Functional Food Health Claims in Conflicting versus
Mind or the Heart: Effects of Matching Hedonic versus Complementary Information Environments,” Journal of
Utilitarian Arguments and Products,” Marketing Letters, 27 Public Policy & Marketing, 28 (2), 221–33.
(1), 131–42. Newman, George E. and Ravi Dhar (2014), “Authenticity Is
Kostyk, Alena, James M. Leonhardt, and Mihai Niculescu (2017), Contagious: Brand Essence and the Original Source of
“Simpler Online Ratings Formats Increase Consumer Trust,” Production,” Journal of Marketing Research, 51 (3), 371–86.
Journal of Research in Interactive Marketing, 11 (2), Newman, Leonard S. and James S. Uleman (1990), “Assimilation
131–41. and Contrast Effects in Spontaneous Trait Inference,”
Kramer, Thomas and Lauren Block (2011), “Nonconscious Personality and Social Psychology Bulletin, 16 (2), 224–40.
Effects of Peculiar Beliefs on Consumer Psychology and Parker, Jeffrey R., Iman Paul, Ryan Hamilton, Omar Rodriguez-
Smith, Rosanna K., George E. Newman, and Ravi Dhar (2016), Responses to Comparative Advertising,” Journal of
“Closer to the Creator: Temporal Contagion Explains the Consumer Research, 32 (4), 530–40.
Preference for Earlier Serial Numbers,” Journal of Consumer Tonietto, Gabriela N., Selin A. Malkoc, Rebecca Walker Reczek,
Research, 42 (5), 653–68. and Michael I. Norton (2021), “Viewing Leisure as Wasteful
Spencer, Steven J., Geoffrey T. Zanna, and Mark P. Fong (2005), Undermines Enjoyment,” Journal of Experimental Social
“Establishing a Causal Chain: Why Experiments Are Often Psychology, 97, 104198.
More Effective than Mediational Analyses in Examining Wadhwa, Monica and Kuangjie Zhang (2015), “This Number Just
Psychological Processes,” Journal of Personality and Social Feels Right: The Impact of Roundedness of Price Numbers
Psychology, 89 (6), 845–51. on Product Evaluations,” Journal of Consumer Research, 41
Spiller, Stephen A., Gavan J. Fitzsimons, John G. Lynch Jr, (5), 1172–85.
and Gary H. Mcclelland (2013), “Spotlights, Floodlights, Wang, Wenbo, Hean Tat Keh, and Lisa E. Bolton (2010), “Lay
and the Magic Number Zero: Simple Effects Tests in Theories of Medicine and a Healthy Lifestyle,” Journal of