Cultural Orientation and Brand Dilution Impact of Motivation Level and Extension Typicality

Cultural Orientation and Brand Dilution: Impact of Motivation Level and Extension
Typicality
Author(s): Sharon Ng
Source: Journal of Marketing Research, Vol. 47, No. 1 (Feb., 2010), pp. 186-198
Published by: American Marketing Association
Stable URL: http://www.jstor.org/stable/20618964
Accessed: 21-07-2016 18:55 UTC
REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/20618964?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
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 support@jstor.org.
American Marketing Association is collaborating with JSTOR to digitize, preserve and extend access to
Journal of Marketing Research
This content downloaded from 193.226.34.227 on Thu, 21 Jul 2016 18:55:30 UTC
All use subject to http://about.jstor.org/terms
SHARON NG*
This research examines cross-cultural differences in brand dilution
effects and the moderating role of motivation and extension typicality.
Drawing from recent findings that indicate that culture affects the way peo
ple treat conflicting information, this research predicts that Easterners and
Westerners react differently to failures by a brand extension. In contrast
to previous findings that have suggested that failure in a typical extension
leads to less brand dilution for Westerners when they are highly motivated
(than when they are less motivated), this study argues that Easterners
exhibit greater brand dilution when they are less motivated (than when
they are highly motivated). The opposite pattern of results should emerge
when the extension is atypical. Three studies provide support for these
predictions and the underlying processes.
Keywords: brand dilution, culture, self-construal, motivation, extension
typicality
Cultural Orientation and Brand Dilution:

Impact of Motivation Level and
Extension Typicality
The introduction of brand extensions or new products
using an established brand name is one strategy by which
firms leverage a brand's goodwill (Aaker and Keller 1990;
Boush and Loken 1991). This strategy enables firms to
introduce new products at much lower risks and costs.
The proliferation of brand extensions in the marketplace
testifies to the value firms place on this strategy. How
ever, using a brand extension strategy is not without draw
backs. Although successful brand extensions help enhance
the equity of a brand, the converse is true as well: Should
an extended product fail, the negative affect or associations
it generates may filter back to the parent brand, thereby
"diluting" its equity (Loken and John 1993; Milberg, Park,
and McCarthy 1997). Thus, firms need to be cautious when
making brand extension decisions and understand how var
ious factors, such as culture, may affect consumers' reac
tions to extension success and failure.
the effectiveness of branding strategies. The number of

U.S. companies that have expanded into foreign markets
has increased exponentially, and the successful penetra
tion of overseas markets requires a good understanding
of the impact of culture on people's purchasing behavior.
A fundamental issue that marketing managers face is how
Both practitioners and researchers acknowledge the

importance of understanding the impact of culture on
resolve any inconsistency. They are more likely to focus on

just one piece of the information to the exclusion of another
*Sharon Ng is Associate Professor of Marketing, Nanyang Business

School, Nanyang Technological University (e-mail: angsl@ntu.edu.sg).
The author acknowledges financial support from Nanyang Business
School and thanks the two anonymous JMR reviewers for their helpful
comments. The author also thanks Akshay Rao and Michael J. Houston
for their insightful comments on a previous draft of this article. Special
thanks to Barbara Loken for her advice and mentorship during the early
part of this research. Ziv Carmon served as associate editor for this article.
consumers in different cultures react to various branding
strategies. In this research, I examine one aspect of this
issue by investigating how culture affects people's reactions

to brand extension failures.
Recent findings in cross-cultural literature show that peo
ple across cultures do not process information similarly.

Specifically, Easterners (e.g., Chinese people) are more tol
erant of inconsistency. When faced with two pieces of con
flicting information, they prefer to seek a middle position
by taking both pieces of information into account. In con
trast, Westerners (e.g., Americans) feel a greater need to
piece in such situations (Nisbett et al. 2001). This find

ing suggests a difference in the way Easterners and West
erners evaluate brand extension failure information. Why?
Most brands that are extended in the marketplace possess

at least a moderately positive brand image, so when an
extension fails, the negative information conflicts with the
brand's positive image. If Easterners and Westerners differ
in the way they process conflicting information, they also
should differ in their perceptions of brand dilution effects.
This claim is consistent with previous findings that the
? 2010, American Marketing Association Journal of Marketing Research
ISSN: 0022-2437 (print), 1547-7193 (electronic) 186 Vol XLyn (Febmary 201Q)> lg6_19g
Cultural Orientation and Brand Dilution 187

impact of an extension failure on brand equity depends on
other factors, such as the consistency of the extended prod
uct with the parent brand, motivation levels, information
accessibility and diagnosticity, and firms' branding strate
gies (Ahluwalia and Gurhan-Canli 2000; Gurhan-Canli and

Maheswaran 1998; John, Loken, and Joiner 1998; Loken
and John 1993; Milberg, Park, and McCarthy 1997). Build
ing on these findings, I propose that culture is another
important variable to consider.
Thus, the objective of this research is to address two basic
issues that are critical to understanding how firms should
brand their products in foreign markets. First, does culture
affect how consumers react to brand failure information?
I examine this issue in the context of bicultural consumers.
Specifically, I assess how these bicultural people respond

when different cultural orientations or processing styles
temporarily become more accessible. Second, do motiva
tion level and extension typicality moderate the effect? By
showing how the effect of culture varies according to peo
ple's motivation level and type of extended products, I high
light the complexity associated with understanding when
and how culture affects consumers' judgments. Answers to
these questions also can help marketing managers make bet
ter branding decisions in foreign markets.
The rest of this article proceeds as follows: First, I briefly
discuss relevant literature that provides the basis for the

hypotheses. Second, I report on the three studies conducted
to test these hypotheses. Third, I conclude with a discussion
of the implications of the findings and avenues for further
research.
Loken, and Joiner (1998) find that flagship products are rel
atively immune to negative feedback from extended prod
ucts. Milberg, Park, and McCarthy (1997) also show that

firms strategically try to mitigate any negative feedback
effects from product extensions through subbranding.
A second group of studies focuses on identifying vari

ables that moderate brand dilution effects. For example,
Loken and John (1993) examine how perceptions of "gen
tleness" and "quality" for Johnson & Johnson may depend
on the typicality of the brand extension (i.e., extent to which
the extension appears inconsistent with the parent brand).
They find that dilution is less likely when extension typ

icality is salient and consumers perceive the extension as
atypical of the family brand. Ahluwalia and Gurhan-Canli
(2000) find that enhancement and dilution effects are great
est when extension information is highly accessible (i.e., top
of the mind).
Thus, substantial research has investigated the factors

that affect brand dilution. However, most existing research
pertains to Western cultures. Do extension failures have

similar effects in Eastern cultures? Existing research is
silent on this issue. Drawing from recent findings in cross
cultural literature, I propose that people in different cultures
react differently to extension failures, depending on their
motivation level and the typicality of the extended product.
Impact of Motivation and Extension Typicality

Existing literature suggests that a key factor that affects
brand dilution effects is motivation level. Motivation affects
a person's willingness to process information (Eagly and
THEORETICAL FRAMEWORK
Extant Literature on Brand Feedback Effects
To renew consumers' interest in their products and keep
pace with changing consumer tastes, firms must introduce
new products in the marketplace. However, the manner in
which new products should be introduced remains a critical
question. Should new products represent a separate brand

name or the same family brand? Each strategy has pros
and cons. Introducing new products under the family brand
name helps reduce promotion costs and encourages accep

tance of new products, especially if consumers are already
familiar with the brand and transfer their existing brand
beliefs to the new products (Keller 2003). However, infor

mation does not flow in only one direction. Should the
Chaiken 1993; Petty and Wegener 1999). Dual process
theories such as the elaboration likelihood model and the
heuristic-systematic model postulate that motivation may
affect the quality and quantity of information processed

(Petty and Wegener 1999). When motivation is low, people
engage in less resource-demanding processes. They pro
cess less information and scrutinize the same information
less carefully (Gurhan-Canli and Maheswaran 1998; Petty

and Wegener 1999). They rely only on a subset of infor
mation to make their judgments. However, when motiva
tion increases, people engage? in relatively more extensive
and effortful processing. They are more likely to scrutinize
all information presented and show increased consideration
of diagnostic information to arrive at a reasoned attitude

(Aaker and Sengupta 2000; Gurhan-Canli and Maheswaran
1998; Petty, Cacioppo, and Schumann 1983; Petty and
new product fail, negative beliefs generated from the exten

sion failure may filter back to the parent brand, thereby
Wegener 1999).
diluting its brand equity (John, Loken, and Joiner 1998;

Loken and John 1993). Because firms must protect their
less effortful processing and rely only on a subset of infor

mation when they experience low motivation, there is a lack
of agreement about what information people focus on when
they receive conflicting information. One stream of litera
ture (e.g., consistency bias, negativity effects, predecisional
distortion) argues that when motivation is low, people dis
count any inconsistent information and rely on their prior
attitude or beliefs in their judgment. For example, consis
tency bias research argues that with low motivation, people
are more likely to base their evaluations on prior brand atti
tudes and discount any inconsistent information. Research
brand equity, understanding when and to what extent brand
dilution occurs is important both managerially and the

oretically. As testimony to the importance of this issue,
many studies have attempted to address these questions
(e.g., Ahluwalia and Gurhan-Canli 2000; Gurhan-Canli and
Maheswaran 1998; John, Loken, and Joiner 1998; Loken
and John 1993; Milberg, Park, and McCarthy 1997).
Prior studies consist of two categories. One group of
studies focuses primarily on identifying situations in which

firms are insulated from dilution effects. For example, John,
Although general consensus states that people engage in
in predecisional distortion also shows that people tend to
188 JOURNAL OF MARKETING RESEARCH, FEBRUARY 2010

distort new information to make it consistent with their
existing preferences (Russo, Meloy, and Medvec 1998).

According to this perspective, in the event of an exten
sion failure, greater dilution occurs when motivation is high
than when motivation is low because, when motivation is

low, consumers rely more on their prior attitude toward
the brand (which is generally positive, because brands that
possess positive equity are more likely to engage in brand
extensions). However, when motivation increases, the con
sumer pays greater attention to all information presented
and gives greater weight to information that is more diag
nostic. Prior research on the negativity effect has shown that
people consider negative information (e.g., extension fail

ure) more diagnostic than positive information (Ahluwalia
2002), so greater dilution may occur with high motivation.
However, another group of studies argues for the reverse
effects. A person may be less likely to process all the infor
mation presented when he or she has low motivation, but in
that case, the person also chooses to rely on the information
he or she perceives as more diagnostic (Chaiken, Liberman,
and Eagly 1989). For example, Maheswaran and Chaiken

(1991) find that with low motivation, when the valence
of attribute information is inconsistent with that of the
heuristic cues, people tend to rely on the more diagnostic
attribute information to make their judgments. Aaker and
Sengupta (2000) also find that when there is an incongruity
between the source (e.g., endorser) and attribute informa
tion, the more diagnostic attribute information influences
consumers' judgments more.
According to this research stream, consumers should pay
greater attention to the extension failure information (which
is more diagnostic) when they have low motivation. How

ever, when people invest more effort to process all the
information available to them with high motivation, the
inclusion of other pieces of information (including positive
information, such as prior brand attitude) in the evaluation
might mitigate the negative effect of the extension failure.
In this situation, greater dilution is expected in the low

motivation condition.
This perspective is not to suggest that people do not

focus on diagnostic information when they have high moti
vation, nor that positive and negative information con

tribute equally to people's attitudes. Rather, a person who
takes into account both positive and negative information
should exhibit more favorable attitudes than another per

son who only takes into account negative information. With
high motivation, the more diagnostic negative information
should exert a greater impact, but the mere consideration of
the positive information also should exert some influence
on attitudes.
Thus, the two perspectives make different predictions

about the impact of motivation on brand dilution. Although
they appear contradictory, they may be equally valid,

albeit in different conditions. Studies by Gurhan-Canli and
Maheswaran (1998) show that depending on the typical

ity of the extension, higher motivation leads to greater or
lesser dilution. Specifically, the failure of an extension in
a product category that appears typical or similar to the

brand's existing products leads to greater dilution in the
low motivation condition than in the high motivation con
dition. Failure in a typical extension is a piece of diag

nostic information that people cannot ignore. Because they
consider only a subset of information when they experi

ence low motivation, people likely focus on this negative
information. However, when motivation increases, includ
ing other information reduces the impact of the extension
failure information. However, when an extension is atypi
cal, failure in the extension is not diagnostic of the brand's
quality, and in the low motivation condition, consumers
subtype this extension, such that it has minimal influence
on the overall brand schema. In the high motivation con
dition, consumers consider all pieces of information, and
extension failure information might exert some influence
on perceptions, even if it is not highly diagnostic informa
tion. In such cases, greater dilution should occur in the high
motivation condition than in the low motivation condition.
Thus, the impact of motivation on brand dilution effects

depends on the perceived typicality of the extension.
Hla: For a typical extension, when Westerners perceive an
extended product's performance on an attribute as below
expectations, they evaluate the parent brand lower on that
attribute when their motivation is low than when their
motivation is high.
Hlb: For an atypical extension, when Westerners perceive an
attribute when their motivation is high than when their

motivation is low.
Brand Dilution Effects in Eastern Culture

Despite some differences, most prior studies concur that
in the face of conflicting information, people focus only on
selected information when they have low motivation. When
their motivation is high, people take into account more

information and pay greater attention to information that is
more diagnostic. However, is such information processing

universal? The answer is yes and no. Research shows that
similar to Westerners, Easterners engage in more effortful
processing of information when their motivation is high.
Easterners scrutinize the information provided in greater

detail and pay greater attention to information that is
more diagnostic (Aaker and Maheswaran 1997; Aaker and
Sengupta 2000). However, unlike Westerners, when their
motivation is low, Easterners do not necessarily focus only
on selected information in making a judgment. Research in
marketing and psychology shows that even with low moti
vation, Easterners have a greater tolerance for ambiguity

that leads them to incorporate information from conflicting
perspectives in their judgments.1
For example, Aaker and Sengupta (2000) show that when
motivation is low, Westerners focus on selected pieces

of information, whereas Easterners consider the multi
ple pieces of information provided. In their studies, they
also show that in low motivation conditions, when partici
pants from an Eastern origin receive conflicting information
about an attribute and an endorser, they are more likely

]An exception is Aaker and Maheswaran (1997), who find that in low
motivation conditions, Easterners behave like Westerners and focus only
on the diagnostic information. However, Aaker and Sengupta (2000) argue
that this finding applies only to the situation in which one of the cues is
highly diagnostic.

to take into account both pieces of information, whereas
motivation should lead to less favorable attitudes toward
Westerners focus primarily on the attribute. They argue that

this distinction is because Easterners are more tolerant of
and the positive information in low motivation conditions).
the parent brand, versus considering both the negative
ambiguity and less compelled to resolve the incongruity.

This argument is consistent with prior research in cross
cultural literature that shows that Easterners deal with con
However, when the extension is atypical, more elaborate
tradictions in a dialectic or compromise manner (Peng

and Nisbett 1999). They process information holistically,
parent brand. Thus, they focus more on their prior attitude
attending to and assigning some causality to context (Choi
and Nisbett 2000). To Easterners, the world is constantly

changing, and it is important to understand the interrela
tion among events (Nisbett et al. 2001). An ever-changing
world also implies constant contradiction. This empha

sis on change and interrelation among events leads to a
more dialectic way of thinking, which "involves reconcil
ing, transcending, or even accepting apparent contradic
tions" and the search for the "Middle Way" (Nisbett et al.
2001, p. 294; see also Peng and Nisbett 1999). Unlike

Westerners, who are chronically more likely to reject one of
the propositions and focus on the piece of information they
consider more plausible or diagnostic, Easterners believe

that they can find truth in each perspective and should seek
a balance between extreme views.
In support of this argument, in psychology literature,
processing with high motivation should lead Easterners to
realize that the extension failure is not diagnostic of the
and perceive less dilution when their motivation is high

than when their motivation is low (i.e., focusing on the
more diagnostic positive information with high motivation
should lead Easterners to adopt more favorable attitudes

toward the parent brand, versus considering both negative
and positive information in low motivation conditions).2
H2a: For a typical extension, when Easterners perceive an
attribute when their motivation is high than when their

motivation is low.
H2b: For an atypical extension, when Easterners perceive an

attribute when their motivation is low than when their
motivation is high.
Peng and Nisbett (1999) reveal that participants who read

about two contradictory studies demonstrated more moder
ate attitudes than those who read only one of them. This
finding suggests that participants who read about the two
contradictory studies took into account both perspectives,
as reflected in their attitude. Evidence for greater toler
ance of inconsistency in Eastern cultures also emerges from
other domains. For example, Bagozzi, Wong, and Yi (1999)
show that unlike Americans, who are more likely to experi
ence only one form of emotions, both negative and positive
emotions can coexist for Chinese people. Cousins (1989)
also shows that Easterners consider both dispositional traits
and contextual cues when making a judgment; Westerners
take into account only the dispositional traits. Thus, sig
nificant evidence shows that unlike Westerners, who zoom
in on a particular aspect, Easterners are chronically more
likely to take information from varied perspectives into
account.
According to these findings, because extension failure

by a positively viewed brand may be construed as two
pieces of contradictory information (i.e., the positive image
of the brand versus the negative information about the new
product), Easterners should react differently to an extension
failure than Westerners. Specifically, Easterners chronically
take into account more pieces of information, so they are
more likely to incorporate both negative extension failure
information and positive brand attitude in their judgments
when they have low motivation. However, when motivation
increases, closer scrutiny should reveal that some informa
tion is more diagnostic than others, with greater weight

placed on this information (Aaker and Sengupta 2000).
Whether this weight leads to greater or lesser degrees of
dilution depends on the typicality of the extension. When
the extension is typical, more elaborate processing leads

Easterners to focus on the more diagnostic extension failure
information, prompting greater dilution when motivation is
high compared with when motivation is low (i.e., focus

ing on the more diagnostic negative information with high
Brand Enhancement Effects
Both Hj and H2 indicate cross-cultural differences in

brand dilution effects. What happens when an extension is
successful? In the event of extension success, the positive
information is consistent with the positive brand informa
tion. Because there should be no difference in the way
Easterners and Westerners evaluate congruent information,
a similar pattern of results is expected in both cultures.

Consistent with previous research that has shown that
greater elaboration leads to more polarized and stronger
attitudes, greater brand enhancement is also expected when
people are highly motivated than when they are not as moti
vated, across cultures.
H3: Both Easterners and Westerners exhibit a greater degree
of brand enhancement when their motivation is high than
when their motivation is low.
To test these hypotheses, I conduct three studies. Study 1

examines the impact of culture on the feedback effect of
negative information on a typical extension by making

either Eastern or Western cultural orientations more acces
sible through priming. Study 2 builds on the findings in

Study 1 by priming participants' processing styles more
directly and examining the moderating role of extension
typicality. Study 3 further examines what happens in the
event of brand extension success.
2It is argued that only people who take into account both existing
(positive) brand information and (negative) extension failure information
exhibit a different attitude from those who consider only one piece of
information. It is not argued that existing brand information and extension
failure information contribute equally to people's attitudes. The way the

two pieces of information combine may differ across contexts and people;
thus, no predictions are made about the relative impact of each piece of
information.
STUDY I
Method
Design. The objective of Study 1 is to examine whether
people across cultures exhibit different brand dilution
effects. In particular, it focuses on how extension failure
on a key parent brand attribute affects consumers' subse
quent parent brand beliefs about that attribute.3 To test the
hypotheses, a 2 (cultural prime: U.S. culture versus Chinese
culture) x 2 (motivation: low versus high) between-subjects
design was used. Typicality was held constant across con

ditions, and all participants reviewed information about a
moderately typical extension (Hla and H2a). All the factors

were manipulated, and participants were randomly assigned
to each condition.
Culture was manipulated by increasing the accessibil

ity of specific cultural orientations (i.e., Eastern culture
or Western culture) to participants through priming. This
methodology is consistent with recent findings that multi
ple cultural orientations exist within a single person (e.g.,
bicultural people; see Hong et al. 2000; Lau-Gesk 2003),

such that in any given context, one orientation may dom
inate the other. However, a less dominant cultural orien
tation may become temporarily more accessible through a

variety of priming techniques. In contrast to cross-national
designs traditionally used in prior literature, this approach
enables researchers to control for a host of extraneous vari
ables that may lead to potential confounding (e.g., Aaker

and Lee 2000; Chen, Ng, and Rao 2005; Hong et al. 2000).
Consistent with this paradigm, for the following studies,
I recruited bicultural respondents and made their Eastern
or Western cultural orientation temporarily more accessible
through visual primes.
Participants. One hundred five Chinese Singaporean stu

dents from a Singaporean university participated in this
study for partial fulfillment of course credit. Most previous
research has focused on the difference between Eastern cul
ture (epitomized by Chinese culture) and Western culture

(epitomized by U.S. culture), so for these studies, partici
pants viewed either a U.S. prime or a Chinese prime. Singa
porean students are a good sample for such studies because
most Singaporeans retain some knowledge of their ethnic
culture (e.g., Chinese) and yet are also exposed to Western
influences (e.g., products, films). Thus, they embody both
traditional Eastern culture and Western culture, to which
they have been exposed since birth. Moreover, because
English is the first language in the country, no problems
are associated with having to translate the questionnaire.
Approximately 26 students were in each cell, and a control
group of participants who were not exposed to the exten
sion information provided the baseline condition as a check
for dilution.
3In general, prior literature has operationalized brand dilution as either
dilution in the general affect toward the brand (e.g., Keller and Aaker
1992) or dilution in a person's perception of specific attribute beliefs (e.g.,
Loken and John 1993). The second operationalization is adopted because,

when a company uses an existing brand to introduce a new product, it
usually hopes to transfer some prevailing, positive association from the
parent brand to the new product. These associations define the brand and
often are the brand's key features that distinguish it from competitors
(Loken and John 1993).
Procedure and stimuli. Participants completed the online

study in front of a computer. As primes of each cultural
orientation, participants viewed collages comprising either
U.S. cultural icons or Chinese cultural icons. Hong and
colleagues (2000) have applied a similar priming tech
nique successfully in marketing studies (Chen, Ng, and Rao
2005). On the cover page of the study, participants read

that they would be asked to complete two unrelated studies.
The first study aimed to assess their knowledge of impor

tant icons, so they were to examine the icons carefully.
On the next page, participants saw either a U.S. collage
or a Chinese collage. The U.S. collage contained pictures
of U.S. cultural icons (e.g., Superman, Statute of Liberty,
U.S. flag, bald eagle). The Chinese collage included pic

tures of Chinese cultural icons (e.g., Confucius, Great Wall
of China, calligraphy, Chinese opera) (Chen, Ng, and Rao

2005). After viewing the collage, participants recalled the
icons they had just seen, which helped reinforce the prime.
Next, the respondents proceeded to the second study.4
The first page of the second study included the moti

vation manipulation. Participants read that Sony would be
introducing a new type of personal digital assistant (PDA)
that allows users to play games (with controls similar to
Game Boy) in either Singapore (high motivation condition)
or Asia (low motivation condition). Sony is a familiar brand
in Singapore (7 = "highly familiar," M = 6.5), and a pretest
showed that, in general, people are favorable toward it (7 =
"very favorable," M = 6.6). Because for this study extension
typicality remained constant, the extension product needed
to be perceived as moderately typical of Sony's existing
products. Prior research also has shown that culture or pro
cessing styles may affect people's perceptions of extension
similarity (Monga and John 2005; Ng and Houston 2005).
Easterners or people who process information holistically
can find relationships among products better and therefore
view products as more similar than Westerners or people
who process information analytically. Therefore, the experi
mental product must be one that both groups view similarly
in terms of typicality. A PDA fits this bill quite well. First,
Sony currently has products in the PDA category, and the
gaming capabilities described in relation to the extended

product are similar to the game consoles currently sold by
Sony. A pretest shows that PDAs with gaming capabilities
were typical of Sony's products, according to participants

primed with both Chinese and U.S. collages (7 = "typical,"
MUSi = 5.64, MChinese = 5.35). Second, because the influence
of positive brand information relative to negative extension
information is the key variable of interest, a product for
which the brand name is an important attribute was also
needed. Pretests show that brand name was important in
PDA purchases (7 = "important," M = 5.8).
The instructions indicated that the experimenters were
interested in participants' opinion of a new product. To
aid in their evaluation, they would receive a recent Con
sumer Reports review of the new product. The review
provided information about product performance on four
4 A suspicion check conducted at the end of the experiment asked partic
ipants to write down what they believed were the objectives of each study.
Two independent judges coded these responses with regard to whether the
participants linked the two studies together. Almost all the participants
believed the story and did not link the two studies.

the motivation manipulation was successful. Only the main
attributes: design, battery life, audio quality, and price. The

four attributes were chosen because, to test for the dilution
effect of motivation was significant (F(l, 101) = 21.66,
of specific beliefs, the attributes must be associated strongly
p < .01). Participants in the high motivation condition (M =
with Sony and important for consumers' PDA purchases.

From actual reviews of PDAs published in magazines, in
4.44, SE = .16) were more motivated than those in the low

motivation condition (M = 3.42, SE= .16). No other main
newspapers, and on the Web, ten PDA attributes were con

sidered important. In a pretest, a group of 20 students rated
the extent to which they associated Sony strongly with the
attributes, as well as the importance of the attributes for
PDA purchases. The attribute they associated most strongly
effects and interaction effects were significant (p > .1).
more in line with Eastern (Western) culture, they should
with Sony and considered important for PDAs was audio

quality, which served as the focal attribute. For the nega
Ng, and Rao 2005). A full-factorial ANOVA performed
tive evaluation of the new product on the focal attribute,
on the politicians listed showed a significant effect of cul
the review rated the new PDA poorly on audio quality.

A separate pretest with 30 participants also showed that
the Chinese prime condition listed significantly more East
this information was moderately inconsistent with Sony's

image (7 = "inconsistent," M = 5.3; SD = 1.23). In addition,
to mask the focal attribute, the review contained ratings
of three other attributes, namely, design, battery life, and
price. To make the review sound believable, the new PDA

received negative ratings in terms of design but moderately
positive ratings on battery life and price. If the new product
were rated inconsistently on all four attributes, the scenario
would become too unbelievable; it is rare for a new product
to receive poor ratings on all dimensions. After reading the
review, participants rated Sony and the new PDA on the

four attributes, the perceived typicality of the new PDA,
and the consistency of the information provided, relative to
Sony's image.
To assess whether the negative information had any
impact on actual product choice, participants indicated
which brand of MP3 player (Sony or Creative Zen) they
would like to receive, should their names be drawn in a
lucky drawing promoted as part of the experiment. An MP3
player provides the choice option because it is another prod
uct for which audio quality is of paramount importance.

If the negative information provided affects people's per
ceptions of Sony's audio quality, they should be less will

ing to choose the Sony MP3 player. A Creative Zen MP3
player is the alternative MP3 player because it is popular
in Singapore, and most people have a moderately favorable
evaluation of the device (7 = "favorable," M = 5.92). The

pretest showed no significant difference in attitudes between
the two MP2 players (MSony = 5.66; F(l, 24) = .69, p > .1);
the participants also viewed the two brands as equivalent in

audio quality (MCreative = 5.69, MSony = 5.75; F(l, 24) = .37,
p>.\).
Finally, participants indicated any thoughts they had
about the review and completed some manipulation checks

and demographic measures. Participants in the control con
Two measures checked whether the priming manipu

lation was successful. First, participants named the first
three politicians that came to mind; if participants thought
be likely to list more Eastern (Western) politicians (Chen,
tural priming (F(l, 101) = 28.27, p < .01). Participants in

ern politicians (M= 1.78, SE = .10) than those in the U.S.
prime condition (M = 1.22, SE = .10). Participants in the
U.S. prime condition also listed significantly more Western
politicians (M = 2.12, SE = .11) than those in the Chinese

prime condition (M = .88, SE= .11). No other main effects
or interaction effects were significant (p > A).
Second, participants indicated their attitude toward a list
of four values frequently associated with Chinese culture

(e.g., taking care of aged parents, greeting teachers) and
four values frequently associated with the Western culture
(e.g., being unique, independence).5 Both scales demon
strated good psychometric properties (Chinese scale: a =
.78; Western scale: a = .83). Participants in the Chinese
prime condition agreed more with the Chinese cultural val
ues (M = 5.07, SE = .10) than with Western cultural values
(M = 4.31, SE=.10; F(l, 101) = 30.32,p<.01). In contrast,
those in the U.S. prime condition agreed more with the list
of Western cultural values (M = 5.62, SE = .17) than with
the Chinese cultural values (M = 4.56, SE = .17; F(l, 101) =
19.84, p < .01). Thus, the results from both manipulation
checks converged, in support of the cultural manipulation.
Third, the analysis of participants' ratings of the con
sistency of the information provided with Sony's image
showed that the audio quality information appeared mod
erately inconsistent with Sony's image (7 = "inconsistent,"
M = 5.3, SD = 1.66). Comparisons across the four condi
tions revealed no significant differences in the perceptions
of consistency (p > .1). There also was no difference in the
perceptions of extension fit between the new PDA and Sony
across conditions (p > .1).
Analysis. An ANOVA with cultural prime and motiva
tion as the between-subjects variables centered on the focal
attribute of audio quality. The main effects of cultural
prime and motivation were not significant (F(l, 101) =
.80, p > .1; F(l, 101) = .03, p > .1, respectively). How
dition were not exposed to the manipulations; they only
ever, consistent with expectations, a significant two-way
indicated their evaluations of Sony on the four attributes

and their overall attitude toward Sony.
interaction emerged between cultural prime and motivation
(F(l, 101) = 9.83,/?< .01) (see Figure l).6 Contrasts showed

that participants in the Chinese prime condition rated Sony
much lower in audio quality when their motivation was

high (M = 3.71, SE = .19) than when their motivation was
Results
Manipulation checks. For the motivation manipulation
check, participants responded to a two-item, seven-point
scale about how interested and involved they were when
completing the questionnaire (with 7 = "highly involved,"
"interested"). A full-factorial analysis of variance (ANOVA)
performed on the mean of the two-item scale shows that
5These cultural values reflect those typically used to assess self

construal (i.e., independent versus interdependent self) and cultural orien
tation (i.e., individualism versus collectivism).
6 Similar analyses of the three filler attributes did not yield any signifi
cant results {p> .1).

Figure 1
IMPACT OF CULTURAL PRIMES AND MOTIVATION ON BRAND
EVALUATION
focused primarily on the information in the review. The

two judges agreed 94% of the time and resolved any dif
5
H Low motivation
4.5
ferences through discussion. Analysis of the thought data

showed a significant difference across the four conditions
H High motivation
(X2 = 7.46, p < .01). More participants in the U.S. prime
>? 4
low motivation condition indicated that they focused pri
CO
3
O 3.5
marily on the information from the review (n = 21) than
those in the U.S. prime-high motivation condition (n = 9).
' 3B
<
CO
participants' thought data for the extent to which they tried

to reconcile the information in the review with their prior
knowledge about Sony. Participants fell into two categories:
those who tried to reconcile the information and those who
Participants in the U.S. prime-high motivation condition

instead showed a greater desire to reconcile the information
in the review with their existing brand beliefs. The reverse
was true in the Chinese prime conditions: More participants
in the Chinese prime-high motivation condition indicated
that they focused primarily on the information from the
2.5
(/) 2
*? 1.5
CO
review (n = 17) than those (n = 9) in the low motivation
ce 1
condition. Thus, the pattern of results supported the study
predictions.
5H
U.S. Prime
Chinese Prime
low (M = 4.32, SE = .18; F(l, 103) = 5.44, p < .05). Par
ticipants in the U.S. prime condition revealed lower rat

ings of Sony's audio quality when their motivation was low
(M = 3.55, SE = .19) than when their motivation was high

(M = 4.14, SE = .19; F(l, 103) = 4.35, p < .05), consistent
with this study's predictions.
Although the two-way interaction was noteworthy, it

provided no insight about whether dilution occurred. The
manipulation might have increased the attitudes of all par
ticipants, with a greater increase in one condition than in
another. A more stringent test for dilution effects would
determine whether people in the experimental conditions
rated Sony lower in audio quality than those in the con
trol condition, in which the participants did not see any
priming or motivation manipulation. The comparison of
participants' beliefs about Sony's audio quality in all four
experimental conditions with those in the control group
(M = 4.73, SE = .25) showed significant dilution in the
Chinese prime-high motivation condition (MAcontrol group =
1.02, SE = .32; p < .05, confidence intervals [CIs] = .24

and 1.80) and in the U.S. prime-low motivation condi
tion (MAcontrol group = 1.18, SE = .32; p < .01, CIs = .40 and
1.96). However, the attitudes of those in the Chinese prime
low motivation and U.S. prime-high motivation conditions
were not significantly lower than those of the control group
mean (both p > .1). Thus, dilution occurred only in the Chi
nese prime-high motivation and U.S. prime-low motivation
conditions.7
Process evidence. To investigate whether the dilution

occurred as a result of the differential treatment of con
tradictory information, two independent judges coded the
7 An ANOVA of the overall brand evaluations assessed whether dilution
at the attribute level filters down to affect participants' evaluations of the
brand. No significant main effect or interaction effect was found (p > .1).
Impact on choice behavior. Finally, to examine whether

the dilution affected actual purchase behavior, participants'
choice of MP3 player was analyzed. A chi-square test of the
number of people who picked the Sony MP3 player versus
the Creative Zen MP3 player showed a significant differ

ence across conditions (x2 = 7.84, p < .01). More partici
pants in the U.S. prime-high motivation condition picked
the Sony MP3 player (n = 20) than those in the U.S. prime
low motivation condition (n = 11). However, more par

ticipants in the Chinese prime-low motivation condition
picked the Sony MP3 player (n = 22) than those in the
Chinese prime-high motivation condition (n = 9). There
fore, dilution filtered down to affect participants' subse
quent product choices.
Discussion
The results from this study provide support for the
proposition that Easterners and Westerners differ in the
way they process extension failure information and that this
effect is moderated by their level of motivation. Specifi

cally, for Western-primed participants, dilution is greatest
when motivation is low, but for Eastern-primed participants,
dilution is greatest when motivation is high. The findings
from the thought data also offer evidence with regard to
the process explanation; that is, the differential way East
erners and Westerners treat conflicting information leads
to cross-cultural differences in brand dilution effects. This
result is intriguing, especially for Eastern-primed partici
pants, because it contrasts with the general assumption that
people take all pieces of information into account only

when their motivation is high.
Although the Study 1 findings support Hla and Hlb, this
study suffers from several shortcomings. First, the investi
gation successfully primed the participants' cultural orien
tation, but it remains unclear which dimension of culture
drives these effects. The participants could differ on a mul
titude of cultural dimensions. Because the difference in pro
cessing styles is argued to affect the way people react to the
conflicting information, Study 2 addresses this limitation
by manipulating participants' processing styles directly.

Second, there is a potential alternative explanation for the
Study 1 findings. Easterners exhibit a different pattern of
results, perhaps because they deal with conflicting informa
tion differently. However, it also is possible that Easterners
and Westerners deal with conflicting information similarly,
but a difference in the perceived diagnosticity of the infor
I, me), whereas those in the holistic condition circled pro

nouns that represented an interdependent self (e.g., we, us).
Because self-views are related to processing styles, by acti
vating different self-views, this manipulation should be able
to activate the related processing styles (K?hnen, Hannover,
parent brand may be more diagnostic, such that they rely on

parent brand equity only when they have low motivation.
When their motivation is higher, they take into account all
information and therefore exhibit greater dilution. Although
the thought data in Study 1 appear to support the initial
product that Nokia recently introduced. Nokia is the focal

brand in this study because a pretest showed that, in gen
eral, people have positive attitudes toward this brand (M =
5.96). It also is strongly associated with cell phones, and its
camera function induces positive perceptions of the excel
lent pictures it produces. To identify appropriate extension
products for this study, a pretest preceded the actual experi
ment. To examine how holistic and analytic processors view
failure in typical and atypical extensions, this study needed
products for which there is no difference in the typicality
perceptions across both groups of respondents.
mation drives the results. For Easterners, the role of the
explanation, more evidence could rule out the alternative

explanation. Therefore, in Study 2, extension typicality is
manipulated.
How can manipulation of typicality determine the accu
racy of the two explanations? The two process explana
tions make slightly different predictions in contexts that
include extension typicality. If Easterners consider more

information when their motivation is low, when the exten
sion is typical, they should exhibit greater dilution when

their motivation is high (H2a). When the extension is atyp
ical, greater dilution should occur when motivation is low
than when motivation is high (H2b). The alternative expla
nation makes the opposite prediction: If Easterners, similar
to Westerners, take into account only diagnostic informa
tion when they experience low motivation, and they differ
only in the type of information they perceive to be more
diagnostic (e.g., brand name), when they have informa
and Schubert 2001; Monga and John 2007).

Next, participants read a consumer report about a new
Of the 40 participants recruited for the pretest, half

received the holistic processing style manipulation, and the
other half considered the analytic processing style manip
ulation. After the manipulation, participants indicated the
extent to which they believed a list of 15 products were

similar to Nokia's existing products and that the prod
ucts fit Nokia's brand image. The pretest revealed prod
ucts perceived as both typical and atypical by both groups
of participants. An Internet phone that allows people to

engage in video conferencing online is the atypical exten
sion (7 = "highly typical," holistic processor = 3.97, analytic
tion that is highly diagnostic (e.g., failure of a very typi

cal extension), they should behave like Westerners and rely
processor = 4.26; no significant difference across condi
should exhibit greater dilution in low motivation (than in
as an image stabilizer, is the typical extension (holistic

processor = 6.43, analytic processor = 6.12; no significant
on this information as well. In this situation, Easterners

high motivation) conditions. In contrast, when the exten
sion is atypical, they should exhibit greater dilution in high
motivation (than in low motivation) conditions.
STUDY 2
Method
Study 2 aimed to fulfill two objectives: to examine
whether consumers behave differently when the extension
is typical (versus atypical) and to replicate the findings

from Study 1 with a direct manipulation of the participants'
processing styles.
Participants and design. One hundred nineteen students
from a Singaporean university received $5 each to partic

ipate in this study, which involved a 2 (processing style:
analytic versus holistic) x 2 (motivation: low versus high) x
2 (extension typicality: typical versus atypical) between

subjects design. The definition of typicality referred to
product category similarity. All factors were manipulated,
and participants were randomly assigned to each condition.
Seven conditions had 15 participants, and one condition

had 14 participants.
Procedure and stimuli. To activate their different pro
cessing styles, participants received a short description of
a trip to the city; they were told to circle all the pronouns
in the text (K?hnen, Hannover, and Schubert 2001; Monga
and John 2007). Participants in the analytic condition cir
cled pronouns that represented the independent self (e.g.,
tions). A cell phone with the latest camera functions, such
difference across conditions).
The product review offered participants information
about the extension on four attributes: exterior, user friend
liness, software, and camera picture quality. Picture qual

ity was the focal attribute; the other three attributes were
filler attributes. The product received negative ratings on
picture quality and user friendliness and neutral ratings for
the other two attributes. After reading the review, partic
ipants rated Nokia and the extended product on the list

of attributes, as well as the perceived typicality of the
new product. Finally, they completed a list of manipulation
checks and demographic items before being debriefed.
Results
Manipulation checks. To confirm the success of the pro
cessing style manipulation, participants responded to a list
of six questions that measured the extent to which they

believe things in life are related (e.g., "Everything in the
universe is somehow related to each other," "The whole is
greater than the sum of its parts"). This test is consistent

with the argument that holistic processors tend to pay more
attention to the field and perceive interrelationships among
things (Nisbett et al. 2001). The scale demonstrated good

reliability (a = .75), and the ANOVA of the mean of the
scale showed that the manipulation was successful. Partici
pants in the holistic processing condition were more likely
to believe that everything in the universe is highly related

(M = 5.58, SE = .11) than those in the analytic processing
condition (M = 4.98, SE = .11;F(1, 111)= 13.94, p< .01).
The manipulation checks also showed that the motiva
tion manipulation was successful. Participants in the low
Figure 2
EFFECT OF EXTENSION TYPICALITY, MOTIVATION, AND

PROCESSING STYLES ON BRAND EVALUATIONS
motivation condition indicated significantly lower levels of
A: For a Typical Extension
motivation (M = 4.32, SE = .14) than those in the high

motivation condition (M = 4.94, SE= .15; F(l, 111) = 9.35,
p < .01). Finally, to check the typicality manipulation, par

ticipants' ratings of the typicality of the two extension prod
ucts were analyzed. The results show that the manipulation
was successful (F(l, 111) = 27.39, p < .01): Those in the

low typicality condition rated the extended product signif
icantly lower in terms of typicality than those in the high
typicality condition (Mlow = 4.19, SE = .16; Mhigh = 6.22,

SE = .16). There was no difference in the typicality percep
tions for the two processing style conditions.
Attribute evaluation. Similar to previous studies, a
2x2x2 ANOVA was run. The results show a significant

main effect of processing style (F(l, 111) = 4.42, p < .05)
5.5
m Low motivation
75
5
3
High motivation
O
S? 4.5 i
3
*-*
CL
CO
M
Z 3.5
O)
c
(0
ir
and a three-way interaction among processing style, moti
vation, and product typicality (F(l, 111) = 23.21, p < .01)

(see Figure 2). No other effects were significant. Further
2.5
way interactions in both the high typicality (F(l, 111) =

18.44, p < .01) and the low typicality (F(l, 111) = 5.03,
p < .05) conditions were significant. Specifically, when
typicality was high, participants in the analytic process
ing condition rated Nokia's picture quality significantly
less favorably when their motivation was low (M = 3.36,
SE = .26) than when their motivation was high (M = 4.50,
SE = .26; F(l, 111) = 11.06, p < .01). In contrast, partici
pants in the holistic processing condition rated Nokia sig
nificantly less favorably when their motivation was high
(M = 4.11, SE = .26) than when their motivation was low
(M = 5.00, SE = .26; F(l, 111) = 6.62, p < .05). The reverse
findings emerge when typicality was low. That is, in the
low typicality condition, participants in the analytic pro
cessing condition rated Nokia's picture quality less favor
ably when their motivation was high (M = 3.96, SE = .26)
than when their motivation was low (M = 4.69, SE = .26;
F(l, 111) = 4.42, p < .05). Participants in the holistic condi
tion exhibited no significant differences in their ratings of
Nokia's picture quality (Mlow = 4.53, SE = .26; Mhigh = 4.91,

SE = .26; p > . 1). However, the means are directionally con
sistent with the predictions. Overall, the pattern of results
is consistent with expectations.
Discussion
The findings from Study 2 support the hypotheses that
the typicality of an extension affects the degree of dilution
exhibited by Easterners and Westerners. The findings also
provide evidence contrary to the alternative explanation that
the results obtained in Study 1 were a function of Eastern
ers' reliance on parent brand equity in the low motivation
condition.
STUDY 3
Method
Analytic Holistic
Processing Style
probes into the three-way interaction revealed that the two
B: For an Atypical Extension
5.5
H Low motivation
& 5
75
High motivation
O
3
4.5
CO
To
o
"5 3-5
O)
c
CC 3
2.5
Analytic Holistic
Processing Style
information. Participants consider information about either

an extension success (congruent, positive attribute informa
tion) or an extension failure (incongruent, negative attribute
information) by Dell. In contrast to the extension fail

ure, information about a successful extension by a positive
brand should be consistent with parent brand beliefs, such
that no differences are expected in brand beliefs across

priming conditions (H3).
Participants and design. One hundred seventeen stu
dents from a Singaporean university participated in this
study for partial fulfillment of course credit. A 2 (cultural
prime: U.S. versus Chinese) x 2 (motivation: low versus

high) x 2 (brand extension: success versus failure) between
The objective of Study 3 is to examine what happens

when Easterners and Westerners receive nonconflicting
subjects design was run, with all factors manipulated and

participants randomly assigned to each condition. Three

conditions contained 14 participants, and five conditions
each had 15 participants. In addition, another 15 students
served as the control group. Extension typicality remained
constant across all the conditions.
Stimuli. In this study, Dell is the focal brand. The pretests
showed that, in general, people have favorable attitudes

toward Dell (M = 5.2). The product extension pertains to
digital cameras. Dell is not currently in the digital cam
era market, but this extension is believable because many
computer firms market digital cameras. Moreover, digi
The 30 participants in the pretest viewed the information

in the brand dilution condition as more incongruent with
Dell's image (M = 5.0, SE = .23) than the information in the
brand enhancement condition (M = 2.8, SE = .25; F(l, 29) =

3.56, p < .05). After reading the review, participants rated
Dell and the new camera on each attribute, the perceived

typicality of digital cameras, and the congruity of the infor
mation provided with Dell's image. Finally, they provided

their thoughts about the review and their responses to the
manipulation checks, as well as demographic information.
tal cameras increasingly represent technology or computer
related products. Both Chinese-primed and U.S.-primed

pretest participants perceived digital cameras as similar to
Dell's existing products (7 = "very similar," Mus = 5.7,

MChinese = 5.9).
Procedure. The procedure for this study is similar to

that described for Studies 1 and 2. Participants viewed the
priming collage before the motivation and branding manip
ulations. The study manipulation indicated that Dell was

planning to introduce a new product, a digital camera, to
the market. The focal attribute was picture quality, which
is an important attribute for digital cameras and relevant
to aspects of Dell's brand image. Three other attributes

(customization, user friendliness, and price) masked the
attribute of interest. Of the four attributes, two attributes
(picture quality and user friendliness) were manipulated to

create an enhancement or dilution condition.8 The other two
attributes were neutral, that is, neither positive nor nega
tive. To manipulate brand enhancement and brand dilution,
participants were told that the product performed either bet
ter or worse than other models in the market on the two
attributes. For example, in the brand enhancement condi

tion, the text indicated the following:
Dell Digito delivers excellent picture quality, richer
in color saturation, and clearer than other models
in the market. It combines the power of a 3x opti
cal zoom lens, a 3.2 megapixel CCD, and advanced

photo enhancing technique that makes the pictures
sharper and more vibrant relative to those taken by

other models.
The participants in the brand dilution condition instead

read the following:
Dell Digito delivers moderate picture quality. The
color saturation is not as rich and the picture is
not as clear as other models in the market. Even
though it is equipped with a 3x optical zoom lens, a
3.2 megapixel CCD and advanced photo enhancing
technique, the pictures are not as sharp and vibrant

as those taken by other models.
8The user-friendliness variable in the brand enhancement condition indi
cated, "Although Dell Digito comes with a range of automatic preset

shooting modes, it is not as user-friendly as the other models in the mar
ket. The complicated user interface means you would have to spend much
time learning how to use the camera. Not good for novice users or anyone
looking for a point-and-shoot camera with extra features." In the brand
dilution condition, it read, "Dell Digito also comes with a range of auto
Results
Manipulation and confounding check. The manipulation
checks showed that the motivation manipulation was suc
cessful. A full-factorial analysis indicated that only the
main effect of motivation was significant (Mlow = 3.08,
SE = .27; Mhigh = 4.13, SE = .28; F(l, 109) = 7.35, p < .01).
No other effects were significant (p > .1). An ANOVA

also showed that the priming manipulation was successful
(F(l, 109) = 24.08, p < .01). Participants in the U.S. prime

condition listed significantly more Western politicians (M =
2.28, SE = .14) than Eastern politicians (M = .72, SE = .14).

Those in the Chinese prime condition listed significantly
more Eastern politicians (M = 1.88, SE = .14) than West
ern politicians (M = 1.12, SE = .14). No other effects were
significant (p > .1). Tests of participants' ratings of whether
the extended product fits the brand's existing product cat
egories also indicated no significant difference across con
ditions (p> .1).

Results. As expected, a 2x2x2 ANOVA revealed a
significant main effect of brand extension (F(l, 108) =

23.95, p < .001). Participants in the enhancement condition
(M = 4.72, SE = .14) rated the parent brand significantly

higher on picture quality than those in the dilution condi
tion (M = 3.72, SE = .15). The results also showed a signif

icant three-way interaction for picture quality (F(l, 109) =
5.78, p < .05) (see Figure 3).9 All other main effects and
interaction effects were not significant (p > .1). To decipher
the three-way interaction, the interaction effects of brand
enhancement and brand dilution conditions were examined
separately.
Specifically, the two-way interaction between culture and
motivation was not significant in the brand enhancement
condition (F(l, 109) = .81, /?>.l). In both priming con
ditions, evaluations of Dell's picture quality were higher

in the high motivation condition (Mus = 4.81, SE = .27;
^Chinese = 5.04, SE = .26) than in the low motivation con
dition (Mu s = 4.64, SE = .26; MChinese = 4.40, SE = .26).
Compared with the control group (M = 4.31, SE = .20),
both groups of participants rated the parent brand signif
icantly higher in picture quality in the high motivation

condition, which indicates an enhancement effect (U.S.:
MAcontroi group = -51, SE = .31; one-tailed p < .05, CIs = -.12
and 1.13; Chinese: MAcontrol group = .74, SE = .31; p < .05,
CIs = .13 and 1.36). The enhancement effects were not sig
nificant in the low motivation conditions (both p > .1).
matic preset shooting modes that makes it more user-friendly than other
models in the market. The uncomplicated user interface means you won't
spend much time learning how to use the camera. Perfect for novice users
or anyone looking for a point-and-shoot camera with extra features."
9The corresponding interactions for the control attributes?customiza

tion, user friendliness, and price?were not significant (p > A).

from the brand dilution condition did not occur in the
Figure 3
ASYMMETRIC EFFECT OF BRAND ENHANCEMENT AND

BRAND DILUTION
A: When Extension Succeeds (Brand Enhancement)
6
IH Low motivation
5.5 H
incongruent information.
5n
+
JiJ
?
O 4.5
?)
CO
try in brand dilution and enhancement effects across cul

tures. In addition, the lack of cross-cultural differences in
the brand enhancement (or information-congruent) condi
tion supports the argument that the mechanism driving the
cross-cultural difference in brand dilution effects is dif

ferences in the ways Easterners and Westerners evaluate
H High motivation
CO
3
brand enhancement condition (i.e., when information is

congruent), which signals a certain degree of asymme
?.
-i
4H
GENERAL DISCUSSION
A brand extension is an effective way for firms to lever
age their brand's equity (Keller 2003). Extant literature has
documented various benefits of engaging in brand exten
sions (e.g., Aaker and Keller 1990; Boush and Loken
1991), but there are both pros and cons of extending a
3.5
U.S. Prime
Chinese Prime
B: When Extension Fails (Brand Dilution)
High motivation
(0
3
J?
4H
o 3.5 H
O)
-E
CC
2.5
U.S. Prime
potential dilution of a brand's equity, should an extension

fail (Loken and John 1993). This study examines whether
culture influences brand dilution effects using investigations
with bicultural respondents, which document systematic
El Low motivation
> 4.5 H
brand to new categories. The peril of brand extension is the
Chinese Prime
Replicating the results from Studies 1 and 2, the two

way interaction between culture and motivation was sig
nificant in the brand dilution condition (F(l, 109) = 5.70,
p < .05). The planned contrasts showed that in the U.S.
prime condition, dilution was more extreme with low moti
vation (M = 3.51, SE = .27) than with high motivation

(M = 4.12, SE = .28; F(l, 112) = 3.05, p < .05). However,
in the Chinese prime condition, dilution was more extreme
with high motivation (M = 3.19, SE = .28) than with low

motivation (M = 4.07, SE = .28; F(l, 112) = 6.27, p < .05).
Consistent with the findings from Study 1, significant dilu
differences in the way people with a more Eastern orienta

tion, compared with those with a more Western orientation,
evaluate brand extension failure information. Specifically,
Study 1 shows that when a typical extension fails, Eastern
ers exhibit significantly greater dilution effects when their
motivation is high than when their motivation is low. This
effect may emerge because Easterners who are low in moti
vation are chronically more likely to integrate the nega

tive information into their existing brand schema. However,
when motivation is high, they scrutinize the information

in greater detail and place greater weight on the negative
information, which is more diagnostic. However, for West
erners, a failure in a typical extension causes those who are
low in motivation to focus only on the diagnostic negative
information, whereas those with high motivation pay atten
tion to all the information presented. This difference leads
to greater brand dilution in low motivation conditions.
Building on the results of Study 1, Study 2 replicates

the brand dilution results and shows a reverse pattern when
the extension is atypical. Study 3 further shows that in the
event of a brand extension success, both Easterners and

Westerners behave similarly. Altogether, the findings from
these three studies indicate that extension failure has differ
ential effects across cultures. Therefore, they offer signif
icant theoretical and managerial contributions to branding
and cross-cultural literature.
tion occurred only in the U.S. prime-low motivation con

dition (MAcontrol group = .79, SE = .35; p < .05, CIs = .10 and
Theoretical Implications
(MAcontrol group = 1.11, SE = .36; p < .01, CIs = .40 and 1.83).
The other two comparisons were not significant (both with
p > .1), in support of the hypotheses.
as extensions of prior research, the findings from the
1.49) and the Chinese prime-high motivation condition
Discussion
Theoretically, this study contributes on three fronts. First,
three studies provide further evidence that Easterners treat

conflicting information differently than Westerners, which
has important implications in a marketing context. Prior
The results from this study replicate the brand dilution
findings in Studies 1 and 2. The cross-over interaction
cross-cultural psychology literature provides some evidence

that Easterners treat conflicting information differently than
Westerners (e.g., Bagozzi, Wong, and Yi 1999; Peng and

Nisbett 1999), but it is unclear if such processing differ
ences affect how consumers process marketing informa
introductions, other branding strategies (e.g., subbranding,

individual branding) may be more beneficial.
tion. The only exception is research by Aaker and Sengupta

(2000), who examine this issue in a marketing context.
Second, examining cross-cultural differences in the way
regard to the brand enhancement results. Though not dis
people process conflicting information is a new addition

to brand dilution literature. Traditionally, brand dilution
literature has examined this issue from a categorization
perspective, with the assumption that new products that
may be grouped in the same category as existing products
would prompt favorable views but that the parent brand also
becomes more vulnerable to extension failure. This study's
findings bring a new perspective to this issue by showing
that independent of categorization, the way a person pro
cesses incongruent information is important. A person who
is chronically more likely to take into account all pieces
of information exhibits a different level of brand dilution
than one who focuses only on diagnostic information. By

manipulating extension typicality, the studies further show
that differences in cue diagnosticity across cultures can
not explain this effect. However, it is important to qual
ify that the relative strength of each cue (e.g., reputation,
attribute) is not manipulated, so further research should
examine what happens if consumers from different cultures
see cues of varying diagnosticity.
Third, this research builds on previous findings in brand
dilution literature by identifying an important variable that
moderates the extent to which brand dilution effects occur.
Significant research on the topic of brand dilution identifies
several notable factors that moderate brand dilution effects
(e.g., Ahluwalia and Gurhan-Canli 2000; Loken and John

1993). Implicitly, the same brand dilution processes and
findings should be applicable to other cultures. Yet recent
findings in cross-cultural literature show that this assump
tion is not necessarily true. Cross-cultural differences in

information processing occur in various domains, and it
seems naive to assume that the same brand dilution process
would be universal.
This recommendation is particularly significant with
cussed, the findings from Study 3 exhibit an asymme
try in the brand dilution and brand enhancement effects

across cultures. Unlike the results for brand dilution effects,
Easterners and Westerners demonstrate similar patterns for
brand enhancement. For Easterners, the brand dilution and
brand enhancement effects also differ in their magnitude.
Further analysis of the results in Study 3 reveals that though
both brand enhancement and brand dilution effects are sig
nificant among highly motivated Easterners, the difference
between the mean in the brand enhancement condition and
that for the control group is only .56, whereas the differ
ence between the mean in the brand dilution condition and
that for the control group is 1.18. This discrepancy seems
to suggest that in Eastern cultures, for high-involvement
products, introducing a new product under the family brand
will lead to some enhancement in brand equity if the prod
uct succeeds, but it will lead to relatively greater dilution
of brand equity if the new product fails. This observa

tion argues against brand extensions for high-involvement
products.
However, these recommendations only take into account
the impact of extension success and failure on the par

ent brand. This research does not examine the impact of
brand name on the way Eastern consumers evaluate brand
extensions. In making brand extension decisions, compa

nies should weigh the benefits of a new product introduc
tion with an existing family brand against the cost of the
potential brand dilution in the event of an extension failure.
To the extent that the benefits of introducing new products
under existing brands outweigh the potential risks of brand
dilution, it would be beneficial to engage in brand exten

sion. Further investigations could clarify when the bene
fits of brand extension would outweigh the risks of brand
dilution.
Other marketing implications. The evidence that East

erners and Westerners process conflicting information dif
Managerial Implications
Implications for brand management. These results have
important implications for brand managers operating in a
global world. As companies increasingly rely on foreign

markets for growth, managers need a better understand
ing of whether the branding strategies they use at home
can transfer to other cultures. Marketing strategies must
be localized to sell in foreign markets, but the more dif

ficult question is which marketing strategies should be
modified and how. This issue transcends industries. The
results herein provide some insights into these issues,

including the important implication that firms selling high
involvement products in Eastern markets (e.g., China)

should adopt a different branding strategy than those selling
low-involvement products. For example, among Easterners,
the brand dilution effect is stronger when the extension

is typical and the motivation level is high, so firms sell
ing high-involvement products should be judicious when
engaging in close brand extensions. They also should rec
ognize that if an extended product fails, the potential nega
tive feedback effect may be fairly strong. For risky product
ferently also has important implications for marketing

communications strategies. Though not examined in this
research, such processing differences may affect how con
sumers treat negative information about a positive brand.
For example, in the event of a public relations crisis (e.g.,
quality control failure by a favored brand leads to con

sumer harm), Easterners and Westerners should treat such
information differently. If the issue is severe, Easterners
who are highly motivated (e.g., if the issue is personally

relevant) may exhibit a more negative attitude toward the
company than those who are less motivated. In contrast,

Westerners who are highly motivated may exhibit a less
negative attitude toward the company than those who are
less motivated.
Limitations and Further Research

Several limitations of these studies suggest avenues for
further research. First, all the participants in the studies
came from a bicultural society. The use of a bicultural

sample allows for the random assignment of participants
and reduces the potential of confounds across cultures.

Although it was assumed that through priming partici
Hong, Yingyi, Michael W. Morris, Chi-Yue Chiu, and Ver?nica

Benet-Mart?nez (2000), "Multicultural Minds: A Dynamic Con
structivist Approach to Culture and Cognition," American Psy
chologist, 55 (7), 709-720.
pants tend to think in patterns that are more aligned with

one culture than another, the findings might not generalize
to monocultural societies. Additional research should test
whether people in monocultural societies behave similarly.
John, Deborah R., Barbara Loken, and Christopher Joiner (1998),

"The Negative Impact of Extensions: Can Flagship Products Be
Diluted?" Journal of Marketing, 62 (January), 19-32.
Second, the focus was primarily on brands with positive

evaluations. Most brands that engage in brand extensions
enjoy positive perceptions, but it would be worthwhile to
Keller, Kevin L. (2003), Strategic Brand Management: Building,

Measuring, and Managing Brand Equity, 2nd ed. Upper Saddle
River, NJ: Prentice Hall.
examine how the results might change if the parent brand
were less than positively evaluated.
REFERENCES
-and David Aaker (1992), "The Effect of Sequential Intro
Aaker, David A. and Kevin L. Keller (1990), "Consumer Evalua

tions of Brand Extensions," Journal of Marketing, 54 (January),
27-41.
Aaker, Jennifer L. and Angela Y. Lee (2001), " 'F Seek Pleasures
and 'We' Avoid Pains: The Role of Self-Regulatory Goals in
Information Processing and Persuasion," Journal of Consumer
Research, 28 (June), 33-49.
-and Durairaj Maheswaran (1997), "The Effect of Cultural

Orientation on Persuasion," Journal of Consumer Research, 24
(December), 315-28.
- and Jaideep Sengupta (2000), "Additivity Versus Atten
uation: The Role of Culture in the Resolution of Information

Incongruity," Journal of Consumer Psychology, 9 (2), 67-82.
Ahluwalia, Rohini (2002), "How Prevalent Is the Negativity Effect

in Consumer Environments?" Journal of Consumer Research,
29 (September), 270-79.
- and Zeynep Gurhan-Canli (2000), "The Effects of

Extensions on the Family Brand Name: An Accessibility
Diagnosticity Perspective," Journal of Consumer Research, 27

(December), 371-81.
Bagozzi, Richard, Nancy Wong, and Youjae Yi (1999), "The Role

of Culture and Gender in the Relationship Between Positive
and Negative Affect," Cognition and Emotion, 13 (6), 641-72.
Boush, David M. and Barbara Loken (1991), "A Process-Tracing
Study of Brand Extension Evaluation," Journal of Marketing
Research, 28 (February), 16-28.
Chaiken, Shelly, Akiva Liberman, and Alice H. Eagly (1989),

"Heuristic and Systematic Information Processing Within and
Beyond the Persuasion Context," in Unintended Thought, J.S.
Uleman and J.A. Bargh, eds. New York: Guilford Press,
212-52.
Chen, Haipeng, Sharon Ng, and Akshay Rao (2005), "Cultural

Difference in Consumer Impatience," Journal of Marketing
Research, 42 (August), 291-301.
Choi, Incheal and R.W Nisbett (2000), "Cultural Psychology of
Surprise: Holistic Theories and Recognition of Contradiction,"
Journal of Personality and Social Psychology, 79 (6), 890-905.
Cousins, S.D. (1989), "Culture and Self-Perception in Japan and
the United States," Journal of Personality and Social Psychol
ogy, 56 (1), 124-31.
-
Eagly, Alice H. and Shelly Chaiken (1993), The Psychology of

Attitudes. Orlando: Harcourt Brace Jovanovich.
Gurhan-Canli, Zeynep and Durairaj Maheswaran (1998), "The

Effects of Extensions on Brand Name Dilution and Enhance
ment," Journal of Marketing Research, 35 (November),
464-73.
duction of Brand Extensions," Journal of Marketing Research,
29 (February), 35-50.
K?hnen, Ulrich, Bettina Hannover, and Benjamin Schubert
(2001), "The Semantic-Procedural Interface Model of the Self:
The Role of Self-Knowledge for Context-Dependent Versus
Context-Independent Modes of Thinking," Journal of Person
ality and Social Psychology, 80 (3), 397-409.
Lau-Gesk, Loraine G. (2003), "Activating Culture Through Per

suasion Appeals: An Examination of the Bicultural Consumer,"
Journal of Consumer Psychology, 13 (3), 301-315.
Loken, Barbara and Deborah Roedder John (1993), "Diluting

Brand Beliefs: When Do Brand Extensions Have a Negative
Impact?" Journal of Marketing, 57 (July), 71-84.
Maheswaran, Durairaj and Shelly Chaiken (1991), "Promoting

Systematic Processing in Low-Motivation Settings: Effect of
Incongruent Information on Processing and Judgment," Journal
of Personality and Social Psychology, 61 (1), 13-25.
Milberg, Sandra J., C. Whan Park, and Michael S. McCarthy

(1997), "Managing Negative Feedback Effects Associated with
Brand Extensions: The Impact of Alternative Branding Strate
gies," Journal of Consumer Psychology, 6 (2), 119-^40.
Monga, Alokparna Basu and Deborah R. John (2004), "Con

sumer Response to Brand Extensions: Does Culture Matter?"
in Advances in Consumer Research, Vol. 31, Barbara Kahn
and Frances Luce, eds. Provo, UT: Association for Consumer
Research, 216-22.
Ng, Sharon and Michael J. Houston (2005), "Exemplars or

Beliefs? The Impact of Self-View on the Nature and Rela
tive Influence of Brand Associations," Journal of Consumer
Research, 32 (4), 519-29.
Nisbett, Richard E., Kaiping Peng, Incheol Choi, and Ara Noren
zayan (2001), "Culture and Systems of Thought: Holistic vs.
Analytical Cognition," Psychological Review, 108 (2), 291-310.
Peng, Kaiping and Richard E. Nisbett (1999), "Culture, Dialectics,
and Reasoning About Contradiction," American Psychologist,
54 (9), 741-54.
Petty, Richard, John T. Cacioppo, and David W. Schumann
(1983), "Central and Peripheral Routes to Advertising Effec
tiveness: The Moderating Role of Involvement," Journal of
Consumer Research, 10 (September), 135-46.
and Duane T. Wegener (1999), "The Elaboration Likeli
hood Model: Current Status and Controversies," in Dual Pro

cess Theories in Social Psychology, Shelly Chaiken and Yaacov
Trope, eds. New York: Guilford Press, 73-96.
Russo, J. Edward, Margaret G. Meloy, and Victoria Husted Med

vec (1998), "Predecisional Distortion of Product Information,"
Journal of Marketing Research, 35 (November), 438-52.

Cultural Orientation and Brand Dilution Impact of Motivation Level and Extension Typicality

Transféré par

Informations du document

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Cultural Orientation and Brand Dilution Impact of Motivation Level and Extension Typicality

Transféré par

Droits d'auteur :

Formats disponibles

Cultural Orientation and Brand Dilution: Impact of Motivation Level and Extension

Cultural Orientation and Brand Dilution:

the effectiveness of branding strategies. The number of

Both practitioners and researchers acknowledge the

resolve any inconsistency. They are more likely to focus on

*Sharon Ng is Associate Professor of Marketing, Nanyang Business

consumers in different cultures react to various branding

strategies. In this research, I examine one aspect of this

issue by investigating how culture affects people's reactions

ple across cultures do not process information similarly.

trast, Westerners (e.g., Americans) feel a greater need to

piece in such situations (Nisbett et al. 2001). This find

Most brands that are extended in the marketplace possess

This claim is consistent with previous findings that the

? 2010, American Marketing Association Journal of Marketing Research

Cultural Orientation and Brand Dilution 187

gies (Ahluwalia and Gurhan-Canli 2000; Gurhan-Canli and

Specifically, I assess how these bicultural people respond

discuss relevant literature that provides the basis for the

ucts. Milberg, Park, and McCarthy (1997) also show that

A second group of studies focuses on identifying vari

They find that dilution is less likely when extension typ

Thus, substantial research has investigated the factors

pertains to Western cultures. Do extension failures have

Impact of Motivation and Extension Typicality

a person's willingness to process information (Eagly and

question. Should new products represent a separate brand

name helps reduce promotion costs and encourages accep

beliefs to the new products (Keller 2003). However, infor

Chaiken 1993; Petty and Wegener 1999). Dual process

theories such as the elaboration likelihood model and the

heuristic-systematic model postulate that motivation may

affect the quality and quantity of information processed

engage in less resource-demanding processes. They pro

cess less information and scrutinize the same information

less carefully (Gurhan-Canli and Maheswaran 1998; Petty

of diagnostic information to arrive at a reasoned attitude

new product fail, negative beliefs generated from the exten

diluting its brand equity (John, Loken, and Joiner 1998;

less effortful processing and rely only on a subset of infor

brand equity, understanding when and to what extent brand

dilution occurs is important both managerially and the

Prior studies consist of two categories. One group of

studies focuses primarily on identifying situations in which

Although general consensus states that people engage in

in predecisional distortion also shows that people tend to

188 JOURNAL OF MARKETING RESEARCH, FEBRUARY 2010

existing preferences (Russo, Meloy, and Medvec 1998).

than when motivation is low because, when motivation is

people consider negative information (e.g., extension fail

and Eagly 1989). For example, Maheswaran and Chaiken

is more diagnostic) when they have low motivation. How

In this situation, greater dilution is expected in the low

This perspective is not to suggest that people do not

vation, nor that positive and negative information con

should exhibit more favorable attitudes than another per

Thus, the two perspectives make different predictions

they appear contradictory, they may be equally valid,

Maheswaran (1998) show that depending on the typical

a product category that appears typical or similar to the

dition. Failure in a typical extension is a piece of diag

consider only a subset of information when they experi

Thus, the impact of motivation on brand dilution effects

attribute when their motivation is low than when their