THEODORE J.

NOSEWORTHY and REMI TRUDEL

Marketers struggle with how best to position innovative products that

are incongruent with consumer expectations. Compounding the issue,
many incongruent products are the result of innovative changes in prod-
uct form intended to increase hedonic appeal. Crossing various prod-
uct categories with various positioning tactics in a single meta-analytic
framework, the authors nd that positioning plays an important role in
how consumers evaluate incongruent form. The results demonstrate that
when a product is positioned on functional dimensions, consumers show
more preferential evaluations for moderately incongruent form than for
congruent form. However, when a product is positioned on experiential
dimensions, consumers show more preferential evaluations for congruent
form than for moderately incongruent form. Importantly, an increase in
perceived hedonic benets mediates the former, whereas a decrease in
perceived utilitarian benets mediates the latter. The mediation effects are
consistent with the view that consumers must rst understand a products
functionality before engaging in hedonic consumption.
Keywords: form and function, product incongruity, functional positioning,
experiential positioning, product design
Looks Interesting, but What Does It Do?
Evaluation of Incongruent Product Form
Depends on Positioning
Product positioning is fundamental to marketing strat-
egy, especially for new and innovative products. In new
product development, marketers can innovate by adding
novel functions, or they can innovate by altering the phys-
ical form of the product to increase aesthetic or hedonic
appeal. The difculty is that innovative changes in form
are often incongruent with consumer expectations. Toning
shoes (shoes that help develop your physique) are a recent
example of a new product that, because of a large rounded
sole, has a form that is incongruent with consumer expec-
tations. When competing shoemakers Skechers and Reebok
rst introduced their toning shoes in 2009, they chose very
distinct promotional strategies. Skechers promoted how its
Shape-ups strengthened users back, core, and leg muscles
*
Theodore J. Noseworthy is a doctoral student in marketing, Richard
Ivey School of Business, University of Western Ontario (e-mail:
[email protected]. Remi Trudel is Assistant Professor of Market-
ing, School of Management, Boston University (e-mail: [email protected]).
The authors thank Kyle Murray, Keri Kettle, and Towhidul Islam for their
feedback and support on previous versions of this article. Rik Pieters
served as associate editor for this article.
(functional benets), while in a commercial, Reebok pro-
moted its EasyTone shoes as the route to a nice booty,
stating that it helps you get better legs and a better butt
(experiential benets). Despite the substantive implications,
research has yet to explore the role of product position-
ing in shaping consumers evaluations of new innovative
products.
In this research, we support the novel prediction that
product form and function are fully integrated such
that incongruent form can lead consumers to question a
products functionality and cause asymmetric evaluations
depending on how the product is positioned. We nd that
when a product is positioned on functional dimensions,
consumers evaluate moderately incongruent form more
favorably than congruent form. Conversely, when a product
is positioned on experiential dimensions, consumers evalu-
ate moderately incongruent form less favorably than con-
gruent form. Importantly, at the root of this reversal we
nd a shift in perceived benets. Specically, an increase in
perceived hedonic benets mediates the former, whereas a
decrease in perceived utilitarian benets mediates the latter.
We make several advances in our work. First, we extend
existing theory on product incongruity by detailing how
2011, American Marketing Association
ISSN: 0022-2437 (print), 1547-7193 (electronic) 1008
Journal of Marketing Research
Vol. XLVIII (December 2011), 10081019
Evaluation of Incongruent Product Form 1009
functional and experiential positioning strategies differen-
tially inuence the relationship between form incongruity
and consumer evaluations. Second, the marketing litera-
ture has offered no clear consensus on whether consumers
actually prefer incongruity. Research has shown both pos-
itive (Meyers-Levy and Tybout 1989; Stayman, Alden,
and Smith 1992) and negative (Campbell and Goodstein
2001) consumer evaluations for incongruent products. In
this research, we advance both accounts by showing that
the benet of moderately incongruent form is subject to
how the product is positioned. Furthermore, we predict and
demonstrate a process-based explanation for the observed
asymmetries. Last, from a practical standpoint, we answer
calls for greater realism and relevance (Reibstein, Day, and
Wind 2009), particularly given the strategic importance of
product positioning. We do so by combining the disaggre-
gate data from four independent experiments into a single
meta-analytic framework. This allows for a rigorous assess-
ment of our results by ruling out alternative explanations,
crossing different product categories, and using a range of
product positioning techniques. We conclude by discussing
the theoretical and managerial implications of our ndings.
CONCEPTUAL BACKGROUND
Marketing scholars have long explored how consumers
process incongruent products (Meyers-Levy and Tybout
1989; Peracchio and Tybout 1996; Stayman, Alden, and
Smith 1992). Consistent with this body of work, we dene
an incongruent product as a good or service that devi-
ates from a normative expectation. A simple example
might be a round (rather than square) digital camera. This
stream of research has substantive implications in that
new products, especially innovative ones, are often incon-
gruent with consumers existing mental representations or
schemas. A signicant amount of research has led to a
theory of schema congruity or, more precisely, a mod-
erate incongruity effect (Meyers-Levy and Tybout 1989;
Peracchio and Tybout 1996; Stayman, Alden, and Smith
1992). The moderate incongruity effect is predicated on the
idea that consumers will evaluate a product that is mod-
erately incongruent more favorably than a congruent or
extremely incongruent product (Meyers-Levy and Tybout
1989). A moderately incongruent object can be success-
fully identied with minimal effort, and the act of discov-
ery leads people to enjoy it more (Mandler 1982). Mandler
(1982) argues that by successfully identifying an incongru-
ent object, people are able to resolve tensions that originate
from not being able to initially make sense of it. Though
supported in consumer research, the moderate incongruity
effect has been moderated by numerous contextual factors,
such as dogmatism (Meyers-Levy and Tybout 1989), prior
knowledge (Peracchio and Tybout 1996), prior category
affect/processing goals (Goodstein 1993), perceived risk
(Campbell and Goodstein 2001), and, most recently, the-
matic processing (Noseworthy, Finlay, and Islam 2010). As
a whole, these factors call into question whether consumers
truly prefer moderately incongruent products, particularly
given the complexity of real-world consumption. Thus,
there is a question of ecological validity when it comes to
claims of incongruity enhancing product evaluations.
Compounding the issue is the observation that not all
types of incongruity are the same. Products can have incon-
gruent form, making them perceptually incongruent, or
they can have incongruent function, making them concep-
tually incongruent. Though seldom mentioned in the litera-
ture on the moderate incongruity effect, researchers explor-
ing product ambiguity have long acknowledged the dif-
ference between perceptual and conceptual product cues
(Gregan-Paxton, Hoefer, and Zhao 2005; Noseworthy and
Goode 2011). Conceptual incongruity is almost certainly
ambiguous in that the product will typically adopt a known
function from more than one product category, and the
adoption dees a normative expectation. This type of incon-
gruity is quite common today with the inux of new
hybrid or boundary-spanning products. In contrast, percep-
tual incongruity need not be ambiguous; product form can
defy normative expectations by simply being novel. The
practice of adjusting product form is typically considered
a hedonic alteration, commonly used by marketers to elicit
interest (Chitturi, Raghunathan, and Mahajan 2007). The
reason this is important is that changes in product form
lead to perceptual incongruity. Unlike conceptual incon-
gruity, perceptual incongruity does not explicitly commu-
nicate functionality. The onus is on the consumer to infer
it or the marketer to communicate it.
Researchers have shown that particularly with what are
referred to as artifact categories (i.e., things built by
humans), as opposed to natural categories (i.e., natu-
rally occurring objects: birds, dogs, trees, and so on),
physical appearance is highly correlated with assumed
functionality (Matan and Carey 2001). That is, form and
function interact in such a way that functionality can
be inferred from an objects physical appearance (Malt
and Johnson 1992). Consumer research exploring this
interaction has conrmed that perceptual cues dominate
functional inferences (Gregan-Paxton, Hoefer, and Zhao
2005). Indeed, consumers tend to believe that manufactured
products are designed for a specic purpose, and rightly
so. What this means for the moderate incongruity effect
is that incongruent form may cause consumers to ques-
tion the products functionality. Although this prediction
has yet to be explored in the literature, it is not with-
out merit. Researchers have shown that a lack of typi-
cality can sometimes decrease hedonic emphasis (Veryzer
and Hutchinson 1998) while increasing utilitarian empha-
sis (Malt and Johnson 1992). Coupled with the observa-
tion that utilitarianism and hedonism are not polar ends of
a unidimensional continuum, but rather two distinct atti-
tude components (Okada 2005; Voss, Spangenberg, and
Grohmann 2003), it seems reasonable that form adjust-
ments will differentially inuence hedonic and utilitarian
attitudes.
In their attempts to combine form (hedonic dimensions)
with function (utilitarian dimensions), more researchers
are beginning to discover that the two components
have differential effects on product evaluations (Chitturi,
Raghunathan, and Mahajan 2008; Okada 2005; Veryzer
and Hutchinson 1998; Voss, Spangenberg, and Grohmann
2003). Most notably, there is evidence that consumers
attach greater importance to the hedonic dimensions of
consumption, but only if the functionality of the prod-
uct meets their expectations (Chitturi, Raghunathan, and
Mahajan 2007). This observation ts with the idea that
utilitarian dimensions dominate until consumers feel they
deserve to indulge (Kivetz and Simonson 2002). It also ts
with the idea that when functionality is in question, people
1010 JOURNAL OF MARKETING RESEARCH, DECEMBER 2011
will forgo hedonic appeal and opt for a functionally supe-
rior alternative (Berry 1994). Thus, if product form adjust-
ments lead to functional uncertainty, and if consumers need
to understand a products functionality before engaging in
hedonic consumption, form adjustments may only augment
hedonic appeal when functionality is known. This has yet
to be considered in the literature on the moderate incon-
gruity effect.
The moderate incongruity effect is predicated on the
assumption that incongruity can be arousing if not too
severe. A scan of the literature reveals that the most com-
mon characteristic among studies that support the moderate
incongruity effect is the practice of explicitly conveying
the functional features of the product. Yet not all adver-
tisements do this. Instead, when it comes to adjustments
in product form, many advertisements promote the hedonic
or experiential aspects of products (e.g., fast food, cloth-
ing, luxury cars). Why does this matter? The explanation
is in Mandlers (1982) original thesis. Mandler argues that
the act of discovery (i.e., guring out what the product is
or does) is fundamental to the moderate incongruity effect.
This raises a simple question: Would consumers show pref-
erential evaluations for moderate incongruity if the products
were not functionally positioned?
To date, the literature on the moderate incongruity effect
has been silent on the role of product positioning. There-
fore, we aim to extend this research with the proposition
that incongruent form will lead to functional uncertainty
if functional attributes are not explicitly communicated.
Specically, we predict that when it is experientially posi-
tioned, consumers will question the functional benet of
incongruent form and thus evaluate it less favorably than
congruent form. However, when it is functionally posi-
tioned, consumers will be aware of the functional benets
of incongruent form and thus evaluate it more favorably
than congruent form. These predictions underlie the funda-
mental idea that people must rst understand the functional-
ity of a product before engaging in the hedonic dimensions
(Chitturi, Raghunathan, and Mahajan 2007). This is partic-
ularly important given that form adjustments are believed
to augment hedonic appeal (Veryzer and Hutchinson 1998).
Thus, we begin Study 1 by testing the fundamental propo-
sition that when it comes to product form, the moderate
incongruity effect (i.e., the increase in evaluations for a
moderately incongruent product) may be subject to how the
product is positioned. Stated formally:
H
1
: Consumers will evaluate a moderately incongruent product
more (less) favorably than a congruent product when the
product is functionally (experientially) positioned.
STUDY 1: POSITIONING EFFECTS ON EVALUATIONS
OF INCONGRUENT PRODUCTS
Consistent with prior research in marketing, we dene
utilitarian benets as the functional, instrumental, and prac-
tical dimensions of a product, whereas hedonic benets
are the aesthetic, experiential, and otherwise enjoyable
dimensions (Chitturi, Raghunathan, and Mahajan 2007,
2008; Dhar and Wertenbroch 2000). Although we employ
this distinction when operationalizing product position-
ing, we acknowledge that the two are neither mutually
exclusive nor proportionally static. Most products, given
the right context, can be promoted as either experien-
tial or functional (Dhar and Wertenbroch 2000). Thus, in
extending previous work (Bhm and Pster 1996; Chitturi,
Raghunathan, and Mahajan 2007; Dhar and Werten-
broch 2000; Voss, Spangenberg, and Grohmann 2003),
Study 1 explores the novel proposition that people pre-
fer moderately incongruent form when a product is
functionally positioned but prefer congruent form when a
product is experiential positioned. Because the discrepan-
cies in the literature focus on congruent versus moderately
incongruent products, we make no formal predictions for
extreme incongruity.
Method
Participants and design. Participants (N = 150; 52%
female; M
Age
= 26.8) were recruited through newspaper
advertisements and public posters and paid $10 for tak-
ing part in the study. The ad manipulation and dependent
measures were administered electronically in a behavioral
lab. Participants were tested in groups of 1015 individ-
uals and were randomly assigned to one of six experi-
mental conditions in a 2 (product positioning: functional
brand vs. experiential brand) 3 (form congruity: congru-
ent vs. moderately incongruent vs. extremely incongruent)
between-subjects factorial design.
Stimuli: form congruity. We established the operational-
ization of product form by rst compiling photographs of
different containers for soft drinks. We chose the soft drink
category for two reasons: First, soft drinks have been fea-
tured extensively in the congruity literature (Campbell and
Goodstein 2001; Meyers-Levy and Tybout 1989; Stayman,
Alden, and Smith 1992). Second, Campbell and Goodstein
(2001, p. 443) not only manipulate the visual form of soft
drinks but also offer a precise description of their manipula-
tion of moderate incongruity (e.g., a plastic single-serving
bottle with a drinking valve on the lid). Thus, we tailored
the stimuli in line with prior work.
According to Mandler (1982), perceptions of incon-
gruity are judgments of t. Thus, we administered a pretest
(n = 60) to determine whether participants perceived the
selected containers as more (less) typical of their expec-
tations for the general soft drink category. We measured
perceived typicality with three seven-point items (not at
all unique/very unique, very unlikely/very likely, and
[matches expectations] not at all well/very well; Camp-
bell and Goodstein 2001). We then averaged the pretest
items to form an internally consistent measure (o = .73).
Pairwise comparisons veried that the congruent container
was viewed as more typical (M = 6.01) than the moder-
ately incongruent container (M = 3.85; F(1, 57) = 34.88,
p < .001), which in turn was viewed as more typical than
the extremely incongruent container (M = 1.87; F(1, 57) =
73.09, p < .001). Posttest analyses conrmed no effects of
gender or age on measures of perceived typicality, price
perception, or brand favorability (Fs < 1).
Mean differences notwithstanding, we conducted a sec-
ond pretest (n = 60) to qualify the three-level operational-
ization. According to Mandler (1982), a hallmark of the
moderate incongruity effect is that people canwith some
deliberationsuccessfully gure out what a moderately
incongruent object is, whereas extreme incongruity is too
severe a violation that the object tends to go unidentied.
Evaluation of Incongruent Product Form 1011
To test this, the three soft drink containers were stripped
of all category information (e.g., brand names, category
label, ad claims, imagery), and a ctitious brand was photo-
edited onto the products. Participants were simply asked
to identify the products. We used time latencies to explore
participants degree of deliberation.
In line with Mandlers (1982) work, a congruent object
should be identied quickly and successfully, a moderately
incongruent object should be identied successfully but
with deliberation, and an extremely incongruent object
should be identied rarely, regardless of deliberation.
A binary logistic regression (0 = incorrect, 1 = correct)
revealed that participants were statistically no more accu-
rate at categorizing the congruent soft drink (96% accurate)
than the moderately incongruent soft drink (86% accu-
rate; B = 1.21, SE = .94, p > .05). Participants were far
more accurate, however, at categorizing both the con-
gruent and moderately incongruent soft drinks than the
extremely incongruent soft drink (3% accurate; B = 5.89,
SE = 1.43, p < .001 and B = 4.68, SE = 1.20, p < .001,
respectively). Participants were faster at categorizing the
congruent soft drink than the moderately incongruent soft
drink (M = 1,832 msecs vs. M = 7,106 msecs; F(1, 38) =
11.07, p < .01) and faster at categorization the moderately
incongruent soft drink than the extremely incongruent soft
drink (M = 7,106 msecs vs. M = 1,0336 msecs; F(1, 38) =
7.89, p < .05). Combined, the typicality, categorization, and
latency results conrmed the three-level operationalization
of product form.
Stimuli: product positioning. We manipulated product
positioning through explicit promotional claims. In the
experiential condition, the soft drink listed attributes such
as beautiful, inspiring, and invigorating. In the func-
tional condition, the soft drink listed attributes such as
fully carbonated, easy to open, and priced compet-
itively. Nothing else varied between the two conditions.
We pretested the advertisements by having participants
(n = 48) list their thoughts about the product. Two unafl-
iated judges coded participants thoughts for whether they
were experiential (emphasizing the aesthetic or hedonic
dimensions of the advertisement) or functional (empha-
sizing the practical or utilitarian dimensions). The two
coders were consistent (r = .83). All outstanding disagree-
ments were resolved through discussion. We then con-
structed a functional-experiential index by subtracting the
number of functional thoughts from the number of expe-
riential thoughts and dividing the difference by the total
number of thoughts. Zero indicates an equal number of
functional and experiential thoughts, a positive number
indicates more experiential thoughts, and a negative num-
ber indicates more functional thoughts. Overall, participants
listed more functional thoughts when the advertisement
promoted functional attributes (M = .19) and more expe-
riential thoughts when it promoted experiential attributes
(M = .33; F(1, 60) = 34.01, p < .001).
Procedure and dependent measures. We randomly
assigned participants to an electronic instrument that con-
sisted of an online magazine with a pop-up web advertise-
ment (see Web Appendix A at http://www.marketingpower
.com/jmrdec11). Participants were told that the purpose of
the exercise was to explore the impact of modern soft drink
advertisements and that they were to review the advertise-
ment at their own pace before completing the accompa-
nied ten-item questionnaire. The soft drink category was
explicitly communicated before participants viewed the
advertisement to activate the appropriate schema (Meyers-
Levy and Tybout 1989). Of the randomized ten items,
seven (anchored by 1 = not at all, and 7 = extremely)
captured participants overall attitudes toward the product
(left a favorable impression, is likable, is appealing, is
desirable, is of good quality, interested in trial, is a high
performance product). Three items captured participants
perceived typicality (discussed in the product form pretest).
The questionnaire concluded with the thought-listing task
discussed in the positioning pretest.
Results
Manipulation check: product form. We averaged the
three typicality items to form an internally consistent mea-
sure (o= .76). An analysis of variance (ANOVA) revealed a
main effect of product form (F(2, 144) = 130.98, p < .001).
Pairwise comparisons indicated the congruent soft drink
was perceived as more typical (M = 5.41) than the moder-
ately incongruent soft drink (M= 3.22; F(1, 144) = 101.01,
p < .001), which in turn was perceived as more typical than
the extremely incongruent soft drink (M= 1.89; F(1, 144) =
83.99, p < .001). There was no effect of product positioning
(F < 1). More important, the interaction between product
positioning and product form on perceived typicality was
not signicant (F < 1). Thus, the product form manipula-
tion was perceived as intended, and it was not inuenced
by product positioning.
Manipulation check: product positioning. As in the
pretest, two unafliated judges coded participants thoughts
for whether they were experiential or functional, and then
we calculated a functional-experiential index (intercoder
reliability, r = .84). As we expected, participants listed
more functional thoughts when the target advertisement
promoted functional attributes (M = .08) and more expe-
riential thoughts when it promoted experiential attributes
(M = .12; F(1, 144) = 15.56, p < .005). No other effects
were recorded (Fs < 1.81).
Product evaluation . We averaged the seven evaluation
items (o = .88). A two-way ANOVA revealed a signicant
product positioning product form interaction on product
evaluations (F(2, 144) = 12.29, p < .001). As Figure 1 illus-
trates, the nature of the interaction was such that the mod-
erate incongruity effect occurred when the target soft drinks
were functionally positioned. Specically, participants eval-
uated the moderately incongruent soft drink more favor-
ably (M = 5.36) than the congruent soft drink (M = 4.32;
F(1, 144) = 9.22, p < .01) and the congruent soft drink more
favorably than the extremely incongruent soft drink (M =
2.05; F(1, 144) = 16.28, p < .005). In contrast, the moderate
incongruity effect did not emerge when the soft drinks were
experientially positioned. Participants evaluated the congru-
ent soft drink more favorably (M= 5.11) than both the mod-
erately incongruent soft drink (M = 4.43; F(1, 144) = 7.18,
p < .05) and the extremely incongruent soft drink (M= 2.41;
F(1, 144) = 34.78, p < .001). Therefore, consistent with our
prediction (H
1
), the increase in evaluations for moderately
incongruent form seemed contingent on how the product
was positioned.
1012 JOURNAL OF MARKETING RESEARCH, DECEMBER 2011
Figure 1
THE EFFECT OF PRODUCT FORM AND PRODUCT
POSITIONING ON PRODUCT EVALUATIONS
3.0
3.5
4.0
4.5
5.0
5.5
6.0
Functional Experiential
P
r
o
d
u
c
t


E
v
a
l
u
a
t
i
o
n
s
Product Positioning
Congruent
Moderate
Extreme
Discussion
We conducted Study 1 to test whether the moderate
incongruity effect is robust across different types of posi-
tioning. The results suggest maybe not. If the moderate
incongruity effect manifests only when a product is func-
tionally positioned, it seems plausible that this type of posi-
tioning helps consumers make sense of the product. Again,
a key concept of the moderate incongruity effect is that
consumers must be able to gure out the product. Thus,
it follows that when a moderately incongruent product is
experientially positioned, consumers struggle to make sense
of it. This line of reasoning ts the idea that experiential
positioning focuses on contextual ideals instead of product
features and thus provides little to no information about the
functionality of the product. If this is so, we should be able
to nd differences in how consumers perceive the product
depending on how it is positioned. Specically, if people
are unable to make sense of an incongruent product when it
is experientially positioned, they should question the prod-
ucts functionality (i.e., perceive less utilitarian benets),
leading to lower product evaluations. If, however, people
are able to make sense of an incongruent product when
it is functionally positioned, they should view incongruent
form as an aesthetic alteration (i.e., perceive more hedo-
nic benets), leading to higher product evaluations. This
ts the idea that adjustments in product form are often
viewed as aesthetically pleasing (Veryzer and Hutchinson
1998). Yet it also ts the idea that people must rst under-
stand the functional dimensions of a product before engag-
ing in the hedonic dimensions (Chitturi, Raghunathan, and
Mahajan 2007). In short, we predict that the relationship
between product form and positioning on product evalua-
tions is mediated by the perceived functional and hedonic
benets of the product. Stated formally:
H
2a
: Perceived utilitarian benets will mediate the relationship
between form incongruity and product evaluations when
the product is experientially positioned but not when it is
functionally positioned.
H
2b
: Perceived hedonic benets will mediate the relationship
between form incongruity and product evaluations when
Figure 2
RESEARCH MODEL
Product form
Product
positioning
Product
evaluations
Risk
perception
Price
perception
Functional vs.
utilitarian
benefits
Notes: Solid arrows represent the theoretical model. The dotted lines
represent alternative explanations that we empirically ruled out.
the product is functionally positioned but not when it is
experientially positioned.
With that said, an alternative explanation for the
observed asymmetries in evaluation might simply be
the moderate incongruity effect. Researchers have long
explored how moderate incongruity can enhance con-
sumers extremity of evaluation through the act of dis-
covery (Meyers-Levy and Tybout 1989; Peracchio and
Tybout 1996; Stayman, Alden, and Smith 1992). Further-
more, research has shown that when perceived risk is high,
incongruity can lead to a decline in product evaluations
(Campbell and Goodstein 2001). Thus, it could be that
this study merely supports the moderate incongruity effect
in the functional condition but that perceived risk miti-
gated the effect in the experiential condition. This is par-
ticularly plausible given that an experientially positioned
product is often considered more expensive than a func-
tionally positioned product. Thus, in manipulating product
positioning, we may have inadvertently manipulated per-
ceived risk. Although this line of reasoning does not negate
the importance of the results, there is a need to explore this
explanation. Similarly, another explanation that pertains to
price perception might simply be that the moderately incon-
gruent soft drink promoting functional attributes was con-
sidered more expensive and, thus, more desirable and the
moderately incongruent soft drink promoting experiential
attributes was considered cheap/gimmicky and, thus, less
desirable. Indeed, this would be a fatal confound. Figure 2
depicts both the predicted and the alternative explanations
within the core research model.
To explore the prediction that discrete attitude compo-
nents mediate the observed asymmetries in evaluations for
moderately incongruent form, as well to test the alter-
native explanations of perceived risk and price percep-
tion, we designed four replication studies. In an effort to
rule out any additional inuence of the product category
Evaluation of Incongruent Product Form 1013
or positioning technique, each study examined a different
product category and adopted a unique product position-
ing tactic. The core purpose was to offer robust replica-
tions while providing additional insight into the underlying
mechanisms driving asymmetric evaluations for moderately
incongruen form.
META-ANALYSIS: GENERALIZATION OF
POSITIONING EFFECTS
We designed Studies 2, 3, 4, and 5 to test our predictions
using different product categories and different positioning
manipulations. Because we were interested in determin-
ing the robustness of the effect, while ruling out alterna-
tive explanations, we conducted the four studies with the
goal of replication. For the sake of brevity, the methods
and results appear together. Because the four studies share
the same design, procedure, and dependent measures, we
combine the data into a single meta-analytic framework.
A meta-analysis serves two objectives. First, the meta-
analysis allows us to generalize our ndings across alter-
native positioning tactics and different product categories
by treating the study replication as a level-two random
effect. Second, it allows us to produce an overall effect size.
Although multilevel models are pervasive in the social sci-
ences, they are surprisingly absent in marketing research.
Thus, in addition to the theoretical and practical contribu-
tions, we also present a multilevel approach to handling
replication data. Going forward, we cover the manipulation
checks separately and then move into the meta-analysis for
the core dependent measures.
Method
Participants and design. Participants took part in Stud-
ies 25 in exchange for course credit. Because the pilot
study conrmed that evaluations for extremely incongruent
form did not vary by positioning, only the congruent and
moderately congruent manipulations were carried forward
from the pretests. This enabled us to keep consistent with
prior work, which focuses on whether people prefer mod-
erate incongruent form to congruent form (Campbell and
Goodstein 2001). Thus, in all four studies, we randomly
assigned participants to one of four experimental conditions
in a 2 (product positioning: functional vs. experiential)
2 (product form: congruent vs. moderately incongruent)
between-subjects factorial design.
WebAppendixes BandC(seehttp://www.marketingpower
.com/jmrdec11) present the product positioning manipula-
tions for Studies 35 (Study 2 was identical to Study 1 with
the addition of new measures). The studies used a combi-
nation of real and ctitious brands and varied the product
categories. Specically, we adopted the products chosen for
Study 3 from the European automaker Peugeot given its
availability of standard models as well as several concept
cars that differ from conventional norms. We adopted the
products chosen for Study 4 from the German manufacturer
Minox because of the breadth of its digital camera line.
As in Study 1 (Zija soft drinks), Study 5 adopted a brand
name that was unassociated with the product category (Dali
wristwatches). We adopted all the products, except for the
wristwatches, from a recent article (Noseworthy, Cotte, and
Lee 2011). Thus, the stimuli used in the replications were
remarkably consistent with prior work. Nevertheless, we
secured the product form manipulations through the same
two-stage pretest described in Study 1 (Web Appendix D).
In addition to varying the product category, we varied the
positioning tactics. Specically, Study 2 manipulated prod-
uct positioning through promotional claims (N = 112; 53%
female; M
Age
= 20.9); in the functional positioning condi-
tion, the advertisement promoted aspects such as fully car-
bonated, easy to open, and priced competitively, and
in the experiential condition, the advertisement promoted
aspects such as beautiful, invigorating, and inspiring.
Study 3 manipulated product positioning through associa-
tive imagery (N = 100; 55% female; M
Age
= 21.1); in the
functional positioning condition, the advertisement accom-
panied images of the vehicle from different angles, the
interior, and a picture of the dashboard, and in the experi-
ential condition, the advertisement accompanied images of
a sunset highway, a luxury resort, and diamonds. Study 4
manipulated product positioning through explicit benet
statements (N = 92; 50% female; M
Age
= 22.3); in the func-
tional positioning condition, the advertisement emphasized
the utilitarian, instrumental, and practical dimensions of
the product, and in the experiential positioning condition,
the advertisement emphasized the aesthetic, hedonic, and
enjoyable dimensions. Finally, Study 5 manipulated prod-
uct positioning through brand association (N = 100; 53%
female; M
Age
= 27.3); in the functional positioning condi-
tion, the advertisement associated the product with com-
mon brands such as Timex, Seiko, and Casio, targeted to
everyday consumers, and in the experiential positioning
condition, the advertisement associated the product with
luxury brands such as Rolex, Cartier, and Armani, targeted
to afuent consumers. We secured the product position-
ing manipulations through thought-listing tasks (see Web
Appendix E at http://www.marketingpower.com/jmrdec11).
Across all studies, there was no effect of gender or age on
perceived typicality, price perception, brand awareness, or
brand favorability (Fs < 1).
All replications followed the exact same procedures as
in Study 1. The only difference was that we expanded the
10-item questionnaire to 24 items. Specically, to explore
the prediction of discrete attitude components, we adopted
10 seven-point semantic differential items from Voss,
Spangenberg, and Grohmann (2003). Five of the items
captured participants perceived utilitarian benets (effec-
tive/ineffective, helpful/unhelpful, functional/not func-
tional, necessary/unnecessary, and practical/impracti-
cal), and the other ve captured participants per-
ceived hedonic benets (fun/not fun, dull/exciting, not
delightful/delightful, not thrilling/thrilling, and enjoy-
able/unenjoyable). To explore the alternative explanation
of perceive risk, we included 4 seven-point items (not
at all risky/extremely risky, not at all concerned/highly
concerned, very unimportant/very important, and not at
all worried/very worried; Campbell and Goodstein 2001).
To explore the potential confound of price perception, the
randomized questionnaire was followed by an open-ended
question that asked participants to estimate the price of the
target product. The only exception was Study 2, in which
the retail price was given to participants at the beginning
of the study and they were subsequently asked to recall it
at the end. We administered this approach to offer a sce-
nario in which price was known before evaluation and to
determine whether knowing the price would render different
1014 JOURNAL OF MARKETING RESEARCH, DECEMBER 2011
results from those in Study 1. All studies concluded with
a thought task to test whether the positioning manipulation
was perceived as intended.
Results
Manipulation check: product form. The internal consis-
tency of the typicality items ranged from o = .76 to .83.
All four studies observed a main effect of product form.
In Study 2, participants perceived the congruent Zija soft
drink as more typical (M= 5.21) than the moderately incon-
gruent Zija soft drink (M = 3.37; F(1, 108) = 75.06, p <
.001). In Study 3, participants perceived the congruent
Peugeot car as more typical (M = 5.24) than the moder-
ately incongruent Peugeot car (M = 3.62; F(1, 96) = 78.36,
p < .001). In Study 4, participants perceived the congruent
Minox camera as more typical (M = 5.39) than the mod-
erately incongruent Minox camera (M = 3.19; F(1, 88) =
104.46, p < .001). Finally, in Study 5, participants perceived
the congruent Dali wristwatch as more typical (M = 5.34)
than the moderately incongruent Dali wristwatch (M= 3.38;
F(1, 96) = 143.86, p < .001). In all four studies, there was
neither a main effect of product positioning on perceived
typicality (Fs < 1) nor an interaction between product posi-
tioning and product form (Fs < 1). Thus, the product form
manipulations were perceived as intended.
Manipulation check: product positioning. The intercoder
reliability across replications ranged from r = .81 to .94. All
studies observed a main effect of product positioning on
the functional-experiential index. In Study 2, participants
recorded more functional thoughts when the advertisement
listed attributes such as fully carbonated, easy to open,
and priced competitively (M = .34) and more experien-
tial thoughts when the advertisement listed attributes such
as beautiful, invigorating, and inspiring (M = .14;
F(1, 108) = 14.06, p < .001). In Study 3, participants
recorded more functional thoughts when the advertisement
accompanied images of the vehicle from different angles,
the interior, and a picture of the dashboard (M = .09) and
more experiential thoughts when the advertisement accom-
panied images of a sunset highway, a luxury resort, and
diamonds (M = .20; F(1, 96) = 6.21, p < .05). Similarly,
in Study 4, participants recorded more functional thoughts
when the advertisement emphasized the instrumental and
practical dimensions of the camera (M = .04) and more
experiential thoughts when the advertisement emphasized
the aesthetic and enjoyable dimensions of the camera (M=
.06; F(1, 88) = 9.54, p < .01). Finally, in Study 5, partici-
pants recorded more functional thoughts when the adver-
tisement associated the wristwatch with brands such as
Timex, Seiko, and Casio, targeted to everyday consumers
(M= .08) and more experiential thoughts when the adver-
tisement associated the wristwatch with brands such as
Rolex, Cartier, and Armani, targeted to afuent consumers
(M = .16; F(1, 96) = 5.99, p < .05). There was neither a
main effect of product form on the index (Fs < 1) nor a
product formproduct positioning interaction across the
four studies (Fs < 1.85). Therefore, the product positioning
manipulations were perceived as intended.
Meta-Analysis
We rst analyzed Studies 2, 3, 4, and 5 independently.
We conducted all ANOVAs on a 2 (product positioning)2
(form congruity) between-subjects design. Table 1 reports
treatment means for the core dependent measures for Stud-
ies 25. As Table 1 shows, the results from the four studies
demonstrate the same pattern observed in Study 1. Thus,
we have effectively replicated the results using different
product categories and different positioning tactics. After
establishing the studies as replications, we now turn to the
meta-analysis.
To conduct the meta-analysis, we estimated a mixed lin-
ear model (restricted maximum likelihood; Studies 25;
N = 404). Table 2 presents the descriptives, correlations,
and average reliability estimates for each variable used
in the replications. It is worth noting that the aggregated
correlation between perceived utilitarian benets and per-
ceived hedonic benets was .21. The correlation ts with
the idea that hedonism and utilitarianism are not polar
ends along a unidimensional continuum (Okada 2005; Voss,
Spangenberg, and Grohmann 2003).
Product evaluations. Table 3 presents the mixed lin-
ear model results for product evaluations. The model was
specied such that study replication served as a level-
two random effect and product form (1 = congruent; 1 =
moderately incongruent) and product positioning (1 =
functional; 1 = experiential) served as level-one xed
effects. Specifying the model at two levels allowed for a
random coefcient at both the study level and the individ-
ual level. Overall, the variance component corresponding
to the random intercept was .29 and not signicant. This
indicates that the variance in product evaluations did not
signicantly vary by study. However, when the variance
components were partitioned across model levels, the intr-
aclass correlation coefcient was .27, meaning that roughly
27% of the variance in product evaluations was attributed
to study characteristics. This is not surprising given the use
of different product categories, but it does necessitate ruling
out treatment effects.
As Table 3 illustrates, product positioning and product
form did not signicantly affect product evaluations. How-
ever, the interaction term was signicant (B = .94, SE =
.09, p < .001). The interaction accounted for a signicant
increment in variance in product evaluations beyond the
effects of product positioning and product form (AR
2
= .17,
AF(1, 400) = 88.23, p < .001). Because it is possible that
the interaction varied by study replication, we modeled a
second model in which we treated the interaction term as a
random effect. Both the random study-level intercept (.
00
)
and the random interaction coefcient (.
01
) were not signi-
cant. Therefore, the interaction between product positioning
and product form was consistent across replications. The
meta-analysis clearly conveys the generalizability across
replications with an overall effect size (adjusted R
2
) of .18.
As a whole, participants evaluated the congruent products
more favorably when they were experientially positioned
(M = 5.32) than when they were functionally positioned
(M= 4.29; F(1, 400) = 52.02, p < .001). Conversely, partic-
ipants evaluated the moderately incongruent products more
favorably when they were functionally positioned (M =
5.05) than when they were experientially positioned (M =
4.19; F(1, 400) = 36.86, p < .001).
Evaluation of Incongruent Product Form 1015
Table 1
TREATMENT MEANS AND CELL COUNTS FOR CORE MEASURES IN STUDIES 25
Functional Positioning Experiential Positioning
Moderately Moderately
Congruent Incongruent Congruent Incongruent
Study 2: Soft Drinks
Product evaluation 4.31 (.92) 5.65 (.94) 5.06 (.86) 4.29 (.98)
Utilitarian benets 2.89 (.78) 2.84 (.88) 2.61 (.75) 1.64 (.74)
Hedonic benets 3.06 (.81) 5.38 (1.01) 4.42 (.96) 4.41 (1.13)
Perceived risk 1.77 (.83) 2.58 (.91) 2.02 (.82) 3.10 (.95)
Cell size 28 28 28 28
Study 3: Cars
Product evaluation 4.99 (.84) 5.77 (.78) 5.78 (.85) 4.73 (.91)
Utilitarian benets 3.87 (.94) 3.98 (.87) 3.92 (.86) 2.46 (.95)
Hedonic benets 5.24 (.80) 6.11 (.76) 5.68 (.81) 5.35 (.82)
Perceived risk 3.88 (.86) 4.11 (.83) 4.30 (.87) 4.88 (.86)
Cell size 25 25 25 25
Study 4: Digital Cameras
Product evaluation 3.49 (.85) 4.48 (.72) 4.46 (.83) 3.56 (.77)
Utilitarian benets 4.61 (.84) 4.54 (.87) 4.59 (.84) 3.45 (.79)
Hedonic benets 3.52 (.72) 4.55 (.73) 4.16 (.80) 3.81 (.73)
Perceived risk 2.39 (.87) 3.79 (.93) 3.02 (.89) 4.53 (.92)
Cell size 23 23 23 23
Study 5: Wristwatches
Product evaluation 4.32 (.81) 4.85 (.83) 5.29 (.83) 4.13 (.82)
Utilitarian benets 4.62 (.72) 4.89 (.71) 4.36 (.71) 3.38 (.69)
Hedonic benets 3.75 (.86) 5.20 (.83) 4.66 (.83) 4.36 (.81)
Perceived risk 2.80 (.92) 3.73 (.96) 2.92 (.80) 4.15 (1.03)
Cell size 25 25 25 25
Note: Standard deviations are in parentheses.
Table 2
MEANS, STANDARD DEVIATIONS, CORRELATIONS, AND AVERAGE RELIABILITY ESTIMATES FOR THE AGGREGATED VARIABLES
Variable M SD 1 2 3 4 5 6 7 8
1. Form congruity
2. Product positioning
3. Perceived typicality 4.25 1.25 .78

.02 (.78)
4. Utilitarian benets 4.13 1.07 .24

.28

.30

(.84)
5. Hedonic benets 4.60 1.26 .24

.01 .12

.21

(.89)
6. Product evaluations 4.70 1.11 .01 .04 .08 .27

.67

(.88)
7. Estimated price 20.78 48.35 .03 .15

.04 .10 .09 .07

8. Perceived risk 3.69 1.15 .38

.19

.21

.30

.24

.02 .11 (.79)

p < .05.

p < .01.

p < .001.
Notes: Values in parentheses are aggregated reliability estimates. Mean price is in thousands of dollars.
Mediation analysis: utilitarian benets. Given that we
replicated the asymmetric evaluations for moderately
incongruent form, we then switched our focus to test-
ing the underlying mechanism. We conducted mediation
analyses for each individual study independently. The
results were again consistent. For the sake of brevity, we
only report the meta-analysis mediation results. Individ-
ual mediation results are available in Web Appendix F
(http://www.marketingpower.com/jmrdec11).
To determine whether an emphasis on perceived utili-
tarian benets accounted for the decrease in evaluations
for moderately incongruent products that are experientially
positioned, we conducted a moderated mediation analy-
sis (Muller, Judd, and Yzerbyt 2005; Preacher, Rucker,
and Hayes 2007). In keeping with the meta-analysis, we
conducted three separate mixed linear models to test for
moderated mediation. Again, study replication served as
the level-two random effect and product form and prod-
uct positioning served as level-one xed effects. In the
rst model, product form and product positioning inter-
acted to predict product evaluations (the proposed out-
come variable; B = .69, SE = .11, p < .001). The sec-
ond model showed a product formproduct positioning
interaction on perceived utilitarian benets (the proposed
mediator; B = 1.21, SE = .09, p < .001). Finally, in the
third model, when we reran the product formproduct
positioning interaction on product evaluations control-
ling for perceived utilitarian benets, the interaction term
1016 JOURNAL OF MARKETING RESEARCH, DECEMBER 2011
Table 3
MIXED LINEAR MODEL WITH STUDY NUMBER AS A
LEVEL-TWO RANDOM EFFECT AND PRODUCT POSITIONING,
PRODUCT FORM, AND THEIR INTERACTION AS LEVEL-ONE
FIXED EFFECTS
Model 1 Model 2
Fixed Effects
Intercept j
00
4.18

(.28) 4.18

(.28)
Product positioning j
01
.08 (.09) .08 (.09)
Product from j
02
.02 (.09) .03 (.10)
Product positioning .94

(.09) .94

(.10)
product form j
03
Random Effects
Intercept .
00
.29 (.24) .29 (.24)
Product positioning .01 (.02)
product form .
01
Residual o
2
.81

(.05) .80

(.06)

p < .001.
Notes: N = 404. Standard errors are in parentheses. Unstandardized
coefcients are reported.
dropped to nonsignicance (B = .03, SE = .11, p = .97),
while the relationship between perceived utilitarian bene-
ts and product evaluations remained (B = .58, SE = .05,
p < .001). Importantly, the random study coefcient was not
signicant in any of the three models (ps > .10). To fur-
ther the interpretation, we conducted a bootstrapping anal-
ysis to examine the conditional indirect effects of product
positioning (Preacher, Rucker, and Hayes 2007; Model 2).
As we predicted, perceived utilitarian benets mediated the
relationship between product form and product evaluation
when the products were experientially positioned (indirect
B = .21, SE = .06, z = 3.59, p < .001) but not when they
were functionally positioned (z = .01, p = .83). This nd-
ing supports the prediction (H
2a
) that when products are
experientially positioned, moderately incongruent form can
cause people to question the utilitarian benets of the prod-
uct, which in turn can lead to lower product evaluations.
Mediation analysis: hedonic benets. We conducted a
second moderated mediation analysis to determine whether
an emphasis on perceived hedonic benets accounted for
the increase in evaluations for the moderately incongruent
products that were functionally positioned. Again, study
replication served as the level-two random effect and prod-
uct form and product positioning served as level-one xed
effects. As previously, there was a product formproduct
positioning interaction on product evaluations (B = .66,
SE = .11, p < .001). The second model showed a product
formproduct positioning interaction on perceived hedo-
nic benets (B = .96, SE = .12, p < .001). Finally, when
we controlled for perceived hedonic benets, the interac-
tion became nonsignicant (B = .13, SE = .10, p = .19),
while the relationship between perceived hedonic benets
and product evaluations remained (B = .56, SE = .04, p <
.001). Again, the random study coefcient was not signi-
cant in any of the three models (ps > .10). We again exam-
ined the conditional indirect effects of product position-
ing (Preacher, Rucker, and Hayes 2007; Model 2). As we
predicted, perceived hedonic benets mediated the positive
relationship between product form and product evaluation
when the products were functionally positioned (indirect
B = .83, SE = .10, z = 7.93, p < .001) but not when they
were experientially positioned (z = .14, p = .13). This nd-
ing supports the prediction (H
2b
) that when products are
functionally positioned, moderately incongruent form can
cause people to focus on the hedonic benets of the prod-
uct, which in turn can lead to higher product evaluations.
Alternative explanation: perceived risk. Although the
mediation analyses support the prediction that discrete atti-
tude components underlie the observed asymmetries in
product evaluations, an alternative explanation is that the
manipulation of experiential positioning might have inad-
vertently manipulated perceived risk. Research has shown
that an increase in perceived risk can lead to a decline in
evaluations for moderately incongruent products (Campbell
and Goodstein 2001). Thus, it was important for any mean-
ingful interpretation to account for this effect. Overall, a
mixed linear model revealed that participants perceived the
moderately incongruent products as more risky than the
congruent products (B= 1.09, SE = .13, p < .001). Similarly,
participants perceived the experientially positioned prod-
ucts as more risky than the functionally positioned products
(B = .61, SE = .13, p < .001). The product formproduct
positioning interaction was not signicant (B = .25, SE =
.18, p = .16), nor was the random study coefcient (B =
.61, SE = .50, p = .23). When we added risk to the nal
model in the product evaluation analysis (Model 2), the
interaction between product positioning and product form
on product evaluations remained (B = .86, SE = .31, p <
.001), independent of risk (B = .03, SE = .06, p = .59) and
independent of the interaction between risk and product
form (B = .02, SE = .08, p = .77). Finally, a bootstrapping
analysis conrmed that perceived risk did not mediate the
relationship between product form and product evaluations
regardless of whether the target products were functionally
positioned (z = 1.22, p = .22) or experientially positioned
(z = 1.24, p = .21). Thus, not only were the study replica-
tions consistent, but perceived risk also could not account
for the observed asymmetries in product evaluations.
Confound check: price perception. Finally, we conducted
an analysis of price perception to rule out the potential
confound that people considered the functionally (experi-
entially) positioned moderately incongruent products more
(less) expensive and, thus, more (less) desirable. We con-
ducted a mixed linear model, with study replication as
the level-two random effect and estimated price (normal-
ized), product form, and product positioning as level-one
xed effects. Not surprisingly, participants perceived the
experientially positioned products as more expensive than
the functionally positioned products (B = 9,665.11, SE =
4,531.61, p < .05). The main effect of product form was not
signicant (B = 2,110.95, SE = 4,531.61, p = .64), nor was
the product formproduct positioning interaction on price
(B = 2,757.84, SE = 6,408.66, p = .67). The exaggerated
coefcients are obviously being driven by Study 3 (cars).
When broken down by replication, the only study that did
not show a main effect of product positioning on price was
Study 2, in which price was given to participants and they
were asked to recall it (F = 1.16). Importantly, this did not
change the observed asymmetries in evaluations. When we
added price to the nal model in the product evaluation
meta-analysis (Model 2), the product positioningproduct
form interaction remained signicant (B = .92, SE = .14,
p < .001), independent of price (B = .01, SE = .01, p =
Evaluation of Incongruent Product Form 1017
.92) and independent of the interaction between price and
product form (B = .01, SE = .01, p = .30). Importantly, the
random study coefcient was not signicant (B = .26, SE =
.22, p = .25). Thus, price perception did not account for the
pattern of effects in product evaluations.
Discussion
The primary purpose of the meta analysis was to further
the unique observation that people favor experientially posi-
tioned products with congruent form but favor function-
ally positioned products with moderately incongruent form.
The secondary purpose was to explain why this occurs.
The results from four replications conrmed the observa-
tions in Study 1 using different product categories and var-
ious means of product positioning. We left Study 1 with
the cautionary observation that the manipulation of product
positioning may have inadvertently manipulated perceived
risk. This was of particular concern because prior research
has shown that perceived risk can attenuate the moderate
incongruity effect (Campbell and Goodstein 2001). Evi-
dence from the four replications suggests that this was not
the case.
The meta-analysis results support our theorizing; an
increase in perceived hedonic benets mediated the
increase in product evaluations for moderately incongru-
ent form when the product was functionally positioned,
whereas a decrease in perceived utilitarian benets medi-
ated the decrease in product evaluations for moderately
incongruent form when the product was experientially posi-
tioned. The reversal in product evaluations ts recent nd-
ings that consumers must rst establish a products func-
tionality before engaging in the hedonic dimensions of
consumption (Chitturi, Raghunathan, and Mahajan 2007,
2008). Consider that for most functional goods, consumers
have preconceived notions of functionality and thus indulge
in cosmetic or aesthetic novelty, whereas for experiential
goods, they are not driven by functionality but rather by
higher-order aspirations of pleasure, aesthetics, or esteem.
This is not to suggest that form adjustments are not aes-
thetic but simply to caution that there are situations
particularly when functionality is uncertainwhen they
may cause confusion. Thus, although experiential position-
ing is a valuable tactic to increase hedonic appeal, as are
alterations in product form, combining the two may not be
good practice for marketers.
GENERAL DISCUSSION
This work builds on existing research that suggests that
utilitarianism precedes hedonism until functional expecta-
tions are met (Berry 1994; Bhm and Pster 1996; Chitturi,
Raghunathan, and Mahajan 2007). We are the rst, how-
ever, to explore how this pertains to incongruent product
form. In particular, evidence emerged to suggest that the
moderate incongruity effect might be contingent on how a
product is positioned. Indeed, not all positioning strategies
share the same objective. We nd that when a product is
positioned on functional dimensions, consumers show more
preferential evaluations for moderately incongruent form
than for congruent form. This is consistent with the mod-
erate incongruity effect. Unique to this research, however,
we nd that when a product is positioned on experien-
tial dimensions, consumers prefer congruent to moderately
incongruent form. The effect was robust across four stud-
ies using four different product categories and four distinct
positioning strategies.
Although research has illustrated that moderate incon-
gruity can augment product evaluations (Meyers-Levy and
Tybout 1989) and some subsequent studies have attenu-
ated the effect (Campbell and Goodstein 2001; Peracchio
and Tybout 1996), the literature on incongruity has been
mute on the role of product positioning. This is rather sur-
prising given that positioning strategy signicantly affects
how a product is evaluated (Kalra and Goodstein 1998) and
informs the type of inferences that can be made (Pham and
Muthukrishnan 2002).
Furthermore, although research has questioned the eco-
logical validity of the moderate incongruity effect by show-
ing that perceived risk can cause preferential evaluations
for a congruent option (Campbell and Goodstein 2001), our
research is unique in that the asymmetries resulting from
product positioning could not be attributed to perceived
risk. Instead, we found that when a product is positioned on
functional dimensions, moderately incongruent form causes
consumers to perceive more hedonic benets, whereas
when a product is positioned on experiential dimensions,
moderately incongruent form causes consumers to perceive
less utilitarian benets.
As a whole, not only does this work answer calls for
greater realism (Reibstein, Day, and Wind 2009) by test-
ing form adjustments within the actual advertising context,
but it also answers recent calls to highlight the mecha-
nisms underlying causal inferences through careful valida-
tion over multiple studies and by highlighting the specic
effects of one variable while ruling out others (Bullock,
Green, and Ha 2010, p. 554). Thus, this work contributes
to literature on consumer behavior by illustrating that when
consumers know how a product functions, they will show
preferential evaluations for moderately incongruent form.
The observation that an increase in perceived hedonic
benets mediated this relationship is consistent with the
idea that when functionality is known, people prefer the
hedonically superior option (Chitturi, Raghunathan, and
Mahajan 2007). This work also contributes by showing that
when consumers are unsure of a products functionality,
they will show preferential evaluations for congruent form.
The observation that a decline in perceived utilitarian ben-
ets mediated this relationship is consistent with the idea
that when functionality is in question, people prefer a func-
tionally superior option (Berry 1994).
The ndings from this research also extend understand-
ing of the conditions under which preference reversals for
functional versus hedonic options may occur. Kivetz and
Simonson (2002), as well as Okada (2005), demonstrate
that preference reversals are possible through the manipu-
lation of both the mode of preference elicitation and the
mode of presentation. Chitturi, Raghunathan, and Mahajan
(2007) advance this observation by illustrating that varying
only the mode of preference elicitation is sufcient. Our
research illustrates that there are certain circumstance when
varying only the mode of presentation is sufcient. This
nding is important because it illustrates that the way an
innovative (incongruent) product is presented can lead to
preference reversals.
The ndings of our research suggest some general guide-
lines for the positioning of new products with incongruent
1018 JOURNAL OF MARKETING RESEARCH, DECEMBER 2011
form. Marketers who have a product with innovative form
can use product positioning to help resolve the functional
uncertainty that results from the form adjustment. If an
incongruous product is positioned on its functional bene-
ts, and functional uncertainty is resolved, consumers can
then focus on the hedonic benets of the product. If, how-
ever, a marketer positions an incongruent product on its
experiential benets, consumers may question the products
functionality, and this can go unresolved. Consequently,
consumers may prefer a congruent alternative because they
can infer its functionality from memory. Thus, if incon-
gruity can elicit functional uncertainty, and if function pre-
cedes aesthetics in consumer choice, marketers should take
care to inform consumers of product functionality whenever
form adjustments are used for hedonic purposes. Despite
the popular belief that experiential positioning is a more
sustainable positioning strategy (Pine and Gilmore 1999), it
is important for marketers to realize that this type of posi-
tioning does little to infer functional inference. Thus, any
degree of functional uncertainty may override the objec-
tives of experiential positioning.
Implicit in this research is a cautionary note for scholars
examining the effects of product incongruity. Incongruent
form has different implications for product preference than
incongruent function. Though not explicitly contrasted in
this work, form incongruity can call functionality into
question, whereas functional incongruity does not nec-
essarily elicit functional uncertainty but merely causes
an expectancy violation that may result in skepticism or
confusion. Further research should explore the difference
between the two in greater detail. Furthermore, although
our study exposes variables that are consistent with a more
realistic consumption environment, it is difcult to reli-
ably simulate the true risk involved in having to actu-
ally purchase a good. Thus, although we remained rela-
tively consistent with prior work exploring risk (in that
an actual purchase was not elicited), the question remains
whether asymmetric evaluations translate into asymmet-
ric choice. Additional research is clearly needed in this
area.
REFERENCES
Berry, C.J. (1994), The Idea of Luxury: A Conceptual and Histor-
ical Investigation. New York: Cambridge University Press.
Bhm, Gisela and Hans-Rudiger Pster (1996), Instrumental or
Emotional Evaluations: What Determines Preference? Acta
Psychologica, 93 (September), 13548.
Bullock, John G., Donald P. Green, and Shang E. Ha (2010), Yes,
but What Is the Mechanism? (Dont Expect an Easy Answer),
Journal of Personality and Social Psychology, 98 (4),
55058.
Campbell, Margaret C. and Ronald C. Goodstein (2001), The
Moderating Effect of Perceived Risk on Consumers Evalua-
tions of Product Incongruity: Preference for the Norm, Journal
of Consumer Research, 28 (December), 43949.
Chitturi, Ravindra, Rajagopal Raghunathan, and Vijay Mahajan
(2007), Form versus Function: How the Intensities of Specic
Emotions Evoked in Functional Versus Hedonic Trade-Offs
Mediate Product Preferences, Journal of Marketing Research,
44 (November), 702714.
, , and (2008), Delight by Design: The
Role of Hedonic versus Utilitarian Benets, Journal of Mar-
keting, 72 (May), 4863.
Dhar, Ravi and Klaus Wertenbroch (2000), Consumer Choice
Between Hedonic and Utilitarian Goods, Journal of Marketing
Research, 37 (February), 6071.
Goodstein, Ronald C. (1993), Category-Based Applications and
Extensions in Advertising: Motivating More Extensive Ad Pro-
cessing, Journal of Consumer Research, 20 (June), 8799.
Gregan-Paxton, Jennifer, Steve Hoefer, and Min Zhao (2005),
When Categorization Is Ambiguous: Factors That Facilitate
the Use of a Multiple Category Inference Strategy, Journal of
Consumer Psychology, 15 (2), 12740.
Kalra, Ajay and Ronald C. Goodstein (1998), The Impact of
Advertising Positioning Strategies on Consumer Price Sensitiv-
ity, Journal of Marketing Research, 35 (May), 21024.
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), 15570.
Malt, Barbara C. and Eric C. Johnson (1992), Do Artifact
Concepts Have Cores? Journal of Memory and Language,
31 (April), 195217.
Mandler, George (1982), The Structure of Value: Accounting
for Taste, in Affect and Cognition: The Seventeenth Annual
Carnegie Symposium on Cognition, Margaret S. Clarke and
Susan T. Fiske, eds. Hillsdale, NJ: Lawrence Erlbaum Asso-
ciates, 336.
Matan, Adee and Susan Carey (2001), Developmental
Changes Within the Core of Artifact Categories, Cognition,
78 (January), 126.
Meyers-Levy, Joan and Alice M. Tybout (1989), Schema Con-
gruity as a Basis for Product Evaluation, Journal of Consumer
Research, 16 (June), 3955.
Muller, Dominique, Charles M. Judd, and Vincent Y. Yzerbyt
(2005), When Moderation Is Mediation and Mediation Is
Moderation, Journal of Personality and Social Psychology,
89 (December), 85263.
Noseworthy, Theodore J., June Cotte, and Seung H. Lee (2011),
The Effects of Ad Context and Gender on the Identica-
tion of Visually Incongruent Products, Journal of Consumer
Research, 38 (August), 35875.
and Miranda R. Goode (2011), Contrasting Rule-Based
and Similarity-Based Category Learning: The Effects of Mood
and Prior Knowledge on Ambiguous Categorization, Journal
of Consumer Psychology, 21 (July), 36271.
, Karen Finlay, and Towhidul Islam (2010), From a Com-
modity to an Experience: The Moderating Role of Thematic
Positioning on Congruity-Based Product Judgment, Psychol-
ogy & Marketing, 27 (May), 46586.
Okada, Eric M. (2005), Justication Effects on Consumer Choice
of Hedonic and Utilitarian Goods, Journal of Marketing
Research, 42 (February), 4353.
Peracchio, Laura A. and Alice M. Tybout (1996), The Moderat-
ing Role of Prior Knowledge in Schema-Based Product Evalua-
tion, Journal of Consumer Research, 23 (December), 17792.
Pham, Michael T. and A.V. Muthukrishnan (2002), Search and
Alignment in Judgment Revision: Implications for Brand Posi-
tioning, Journal of Marketing Research, 39 (February), 1830.
Pine, Joseph B. and James H. Gilmore (1999), The Experience
Economy: Work Is Theatre & Every Business a Stage. Boston:
Harvard Business School Press.
Preacher, Kristopher J., Derek D. Rucker, and Andrew F. Hayes
(2007), Addressing Moderated Mediation Hypotheses: Theory,
Methods, and Prescriptions, Multivariate Behavioral Research,
42 (1), 185227.
Evaluation of Incongruent Product Form 1019
Reibstein, David J., George Day, and Jerry Wind (2009), Is
Marketing Academia Losing Its Way? Journal of Marketing,
73 (July), 13.
Stayman, Douglas M., Dana L. Alden, and Karen M. Smith
(1992), Some Effects on Schematic Processing on Consumer
Expectations and Disconrmation Judgments, Journal of Con-
sumer Research, 14 (September), 24055.
Veryzer, Robert W. and Wesley J. Hutchinson (1998), The Inu-
ence of Unity and Prototypicality on Aesthetic Responses
to New Product Designs, Journal of Consumer Research,
24 (March), 37494.
Voss, Kevin E., Eric R. Spangenberg, and Bianca Grohmann
(2003), Measuring the Hedonic and Utilitarian Dimen-
sions of Consumer Attitude, Journal of Marketing Research,
40 (August), 31020.
Copyright of Journal of Marketing Research (JMR) is the property of American Marketing Association and its
content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for individual use.