Global Leadership

Article

Understanding Cultural Differences in Consumers’ Reactions to Foreign-Market Brand Extensions: The Role of Thinking Styles

Helena F. Allman, Kelly Hewett, and Mandeep Kaur

Abstract One popular strategy for brands to grow outside of their home markets is to use line extensions. Although extensive research has examined how factors such as culture influence product acceptance across markets, far less is known regarding how such factors influence parent brand perceptions in response to an extension. This article investigates the role of one aspect of culture, consumers’ thinking styles, in terms of postextension parent brand spillover effects, and considers potential implications for firms’ global brand strategies, including the extension’s direction (upward or downward) and parent brand concept (functional or prestige). Through two studies, one using consumer panels in the United States and India and the other set in the United States, the authors examine consumer-level differences by measuring thinking styles. Results based on both studies indicate that vertical line extension type affects analytic thinkers’ reactions more than those of holistic thinkers, and parent brand image benefits (suffers) when a vertical line extension is for a functional (prestige) brand, regardless of consumers’ thinking style. The authors find interesting differences in the role of thinking styles based on the particular dimension examined.

Keywords cross-cultural, brand/product choice, consumer behavior/cognition

Online supplement: https://doi.org/10.1177/1069031X19836780

Consumer reactions to firms’ branding strategies are critical in

firms’ abilities to successfully leverage equity in their brands

across markets. As such, consumers’ evaluations of branding

practices across cultures have been a central topic in the mar-

keting literature for decades (Gürhan-Canli, Sarıal-Abi, and

Hayran 2018). One topic receiving a great deal of attention is

culture’s impact on product extensions’ acceptance in foreign

markets (Ng 2010). With extensions, firms use positive brand

images to introduce products into new (brand extensions) or

existing (line extensions) categories, with extensions benefiting

from parent brand perceptions (Till and Priluck 2000). How-

ever, extension-related information can also influence parent

brand evaluations (Loken and John 1993), called reciprocal

spillover effects (Balachander and Ghose 2003). As noted by

Gürhan-Canli, Sarıal-Abi, and Hayran (2018), research asses-

sing culture’s role in consumer–brand interactions is needed

given increasing market integration due to technology

advances (e.g., 5G mobile technology) that are argued to

“support a ubiquitous connected world” (Scott 2018). We heed

this call and explore the importance of differences in consu-

mers’ thinking styles for reciprocal spillover effects, consider-

ing firms’ global brand strategies.

Studies examining reciprocal spillover effects have con-

firmed the role of features of the parent brand, the extension,

and target-foreign-market consumers; some explore their inter-

active effects (Gürhan-Canli and Maheswaran 1998; Milberg,

Park, and McCarthy 1997). Monga and John (2010) find that

consumers’ thinking styles influence approaches to processing

brand- and extension-related information. Analytic thinkers,

typically from Western markets, tend to focus on parent and

extension features and try to reconcile them. However, holistic

thinkers in Eastern markets seek alternative ways to evaluate

brands, such as connecting extensions with parent brand repu-

tation. We thus expect information on extension and parent

brand features to influence spillover effects for analytic thin-

kers more than for holistic thinkers.

Helena F. Allman is Assistant Professor of Marketing, Marketing and Economics

Department, University of West Florida, USA (email: [email protected]).

Kelly Hewett is Associate Professor of Marketing, Marketing and Supply

Chain Management Department, University of Tennessee, USA (email:

[email protected]). Mandeep Kaur is ITAS tutor, University of Technology,

Sydney, Australia (email: [email protected]).

Journal of International Marketing 2019, Vol. 27(2) 1-21

ª American Marketing Association 2019 Article reuse guidelines:

sagepub.com/journals-permissions DOI: 10.1177/1069031X19836780

journals.sagepub.com/home/jig

This study also aims to contribute to the international mar-

keting literature by considering key aspects of a brand’s

extension strategy—namely, extension type (upward or

downward) and parent brand concept (BC) (functional or

prestige) in exploring cross-cultural impacts of thinking styles

on spillover effects. Vertical line extension (VLE) reciprocal

spillover effects can be particularly significant (Aaker and

Keller 1990). Upward VLEs can signal brand success, and

downward VLEs can attract a wider range of customers

(Heath, DelVecchio, and McCarthy 2011), especially in for-

eign markets. However, because both VLE types can dilute

brands on the basis of perceptions that they represent a depar-

ture from an established image (Janiszewski and Van Osselaer

2000), it is important to examine the VLE direction as part of

the brand’s overall strategy.

Our study addresses three key gaps in extant brand exten-

sion research in the international marketing domain: tendencies

to emphasize (1) national grouping as opposed to consumer-

level cultural differences; (2) particular features of a parent

brand, an extension, or foreign-market consumers as opposed

to a comprehensive set of global brand strategy factors in addi-

tion to cross-cultural differences in brand-related information

processing; and (3) extension perceptions rather than effects on

parent brand perceptions in foreign markets. In addressing

these gaps, our overarching goal is to offer global marketing

managers guidance in their strategies to introduce brand exten-

sions to foreign markets. We address two research questions:

RQ1: How do consumers with holistic versus analytic cog-

nitive thinking styles differentially weigh information

regarding the upward versus downward nature of a

foreign-market VLE and parent BC in their postextension

brand image evaluations?

RQ2: Given differences in thinking styles and their influence

on postextension brand image evaluations, how should mar-

keters target customers within and across different markets

when introducing VLEs? That is, do consumer thinking

styles vary within country as well as across countries, and

are nationality or thinking style differences more important

in explaining differences in the impact of VLE type and

parent BC on spillover effects?

Leveraging arguments based in schema theory, we propose

that parent brand and VLE fit (based on upward/downward

VLE and parent BC perceptions) will influence reciprocal spil-

lover effects. We further expect thinking styles to influence

extension-related information processing; thinking styles help

determine whether a VLE conveys a fit with the parent brand

and its importance for parent brand image. In answering our

research questions, we offer three key contributions:

1. We address a key gap in the global brand extension lit-

erature (Czellar 2003; Yorkston, Nunes, and Matta 2010) by

studying consumer-level cultural differences as opposed to

national groupings. Because horizontal market segments exist

across countries (Bolton and Myers 2003; Ter Hofstede, Steen-

kamp, and Wedel 1999), and cultures are converging (Agarwal,

Malhotra, and Bolton 2010; Ralston et al. 2008), accounting for

heterogeneity (and homogeneity) within and across cultures

(Broderick, Greenley, and Mueller 2007) is critical.

2. By studying parent brands’ BC, VLE type (downward vs.

upward), and foreign-market consumers (i.e., their thinking

styles), we extend prior research by assessing how an inclusive

set of factors can jointly influence firms’ abilities to leverage

brand equity across markets.

3. We extend brand extension research by focusing on reci-

procal spillover effects and not extension perceptions. Identify-

ing extension impacts on parent brands is vitally important in

evaluating global strategies.

We find that consumers upgrade (downgrade) parent brand

postextension images after an upward (downward) VLE, and

that the impact of VLE type on spillover effects is greater for

analytic versus holistic thinkers. Parent brand image benefits

(suffers) when a VLE is for a functional (prestige) brand,

regardless of consumer thinking style. Tendencies to focus

on a focal object versus relationships between an object and

its surroundings, and a tendency to assume stability versus

constant fluctuation, appear to drive cross-cultural differences

in our results. Next, we review literature linking culture and

brand perceptions. We then describe our hypotheses, methods,

and results. We conclude with key implications and future

research areas.

Relevant Literature

Table 1 summarizes culture and global brand extension

research. We focus on studies examining cultural influence

on brand perceptions as firms expand to foreign markets. We

identify (1) whether a focal brand is existing, fictitious, or an

extension; (2) the study’s dependent variable (e.g., evaluations

of or behavior toward a focal brand or parent brand); (3) con-

ceptual foundations for and operationalization of cultural

aspects studied; and (4) key findings.

Global marketing research at the intersection of national

culture and brand perceptions has largely focused on consumer

perceptions of brands already marketed in a particular country.

Studies have used global brand data taken from secondary

sources (Hsieh, Pan, and Setiono 2004; Talay, Townsend, and

Yeniyurt 2015; Van der Lans, Van Everdingen, and Melnyk

2016); case studies with fictitious (Torelli et al. 2012) or

unnamed brands (Zhang and Nelson 2016); global brands such

as Coca-Cola (Shepherd, Chartrand, and Fitzsimons 2015),

H&M, and Zara (Lee et al. 2014); or automobile brands (Bar-

tikowski and Cleveland 2017; Monga and John 2010). In brand

extension research, studies tend to include an extension without

considering its direction. For example, Buil, Chernatony, and

Hem (2009) and Monga and John (2007) use experimental

settings, manipulating extension fit with a parent BC. Ng

(2010) manipulates the extent to which an extension is

“typical” based on the parent brand without considering exten-

sion direction.

Focal outcome variables differ widely across studies. Most

examine focal brand perceptions in a particular market

2 Journal of International Marketing 27(2)

Table 1. Summary of Relevant Literature.

Citation

Culture

Brand Focus Dependent Variable Key Study Finding(s)

Conceptual Treatment Operationalization

Hsieh, Pan, and Setiono (2004)

Hofstede (PDI, IDV, UAI)

Measured at the individual level

Existing global and local brands

Brand purchase behavior

National culture and sociodemographics moderate effects of product-image appeals on brand purchase.

Monga and John (2007)

Thinking styles Measured at the individual level

Brand extension with global parent

Brand extension evaluation

Eastern (holistic-thinking) consumers evaluate extensions more favorably than Western (analytic-thinking) consumers when they see higher fit with parent.

Buil, Chernatony, and Hem (2009)

Hofstede (PDI, IDV, MAS, UAI, LTO)

Country/ nationality as grouping variable

Line extension with foreign parent

Parent brand equity; attitude toward brand extensions

Brand extensions with high fit decrease negative feedback effects on parent brand equity, and effects differ across countries.

Monga and John (2010)

Thinking styles Measured at the individual level

Existing global brands

Brand extension evaluation

For functional brands, holistic thinkers evaluate distant extensions more favorably than analytic thinkers; for prestige brands, holistic and analytic thinkers respond equally favorably.

Ng (2010) Thinking styles Priming Eastern vs. Western cultural orientations

Failed/successful brand extension with foreign parent

Spillover effects on parent brand

Differential treatment of conflicting information leads to cross-cultural differences in brand dilution effects after failed extensions.

Torelli et al. (2012)

Triandis (vertical/ horizontal individualism/ collectivism)

Measured at the individual level

Fictitious brand Brand evaluation The authors propose and test the structure of abstract BCs as representations of human values and show that this method can be applied across culturally distinct markets.

Lee et al. (2014) N.A. a

Region as grouping variable

Existing global brand

Brand purchase probability; CLV

Promotion management activity affects brand attitude in Korea, but not in Europe.

Shepherd, Chartrand, and Fitzsimons (2015)

Power vs. universalism

Measured at the individual level

Existing global brand

Brand attitude Consumers vary in satisfaction with U.S. society, endorsement of its defining values, and the values they endorse versus reject in brands.

Talay, Townsend, and Yeniyurt (2015)

Hofstede (PDI, IDV, UAI)

Index values at brand-country level

Existing global brands

Brand market share

Global brands perform better than nonglobal counterparts, and culture moderates these effects.

Zhang and Nelson (2016)

Triandis (vertical/ horizontal individualism/ collectivism)

Measured at the individual level

Unnamed brand Brand consciousness, purchase behavior

Vertical individualism influences brand, status consciousness and product choice.

Van der Lans, Van Everdingen, and Melnyk (2016)

Inglehart (traditional/ secular; survival/self- expression)

World Value Survey country- level measures

Existing global and local brands

Brand purchase intentions

Intrinsic benefits (i.e., quality and uniqueness) are most important for purchase intentions, and culture moderates these effects.

Bartikowski and Cleveland (2017)

Ethnic/national identification

Measured at the individual level

Existing global brand

Brand attitude National identity affects preference for brands with local vs. global consumer culture positioning.

This study Thinking styles Measured at the individual level

VLE with foreign parent

Spillover effects on parent brand

The influence of brand strategy characteristics on spillover effects varies based on consumers’ thinking styles; these effects also vary across dimensions of thinking styles.

aNot applicable because no conceptual arguments related to culture; however, nationality is used in the analysis to compare results by country. Notes: PDI ¼ power distance index; IDV ¼ individualism versus collectivism index; UAI ¼ uncertainty avoidance index; MAS ¼ masculinity index; LTO ¼ long- term orientation index; CLV ¼ customer lifetime value.

Allman et al. 3

(Bartikowski and Cleveland 2017; Buil, Chernatony, and Hem

2009; Monga and John 2007, 2010; Shepherd, Chartrand, and

Fitzsimons 2015; Torelli et al. 2012). Others investigate influ-

ences on a certain brand’s behavioral responses, such as pur-

chase intentions (Van der Lans, Van Everdingen, and Melnyk

2016), purchases (Lee et al. 2014; Hsieh, Pan, and Setiono

2004; Zhang and Nelson 2016), or brand market share (Talay,

Townsend, and Yeniyurt 2015). Similar to our study, Ng (2010)

examines parent brand spillover effects; however, she focuses

on reactions to an existing extension’s failure. Because strong

parent brand images motivate extensions, those extensions’

impact on parent brands is arguably an issue of great practical

importance. For example, Balachander and Ghose (2003) high-

light the Contadina canned-foods brand, which benefited from

introducing a line of fresh refrigerated pastas and sauces. Thus,

it is surprising that few studies examine spillover effects in a

global context.

As with much research on national culture, studies in this

domain tend to conceptualize consumers’ differences along

the cultural dimensions proposed by Hofstede (1984) (Buil,

Chernatony, and Hem 2009; Hsieh, Pan, and Setiono 2004;

Talay, Townsend, and Yeniyurt 2015), or individualism–col-

lectivism, as conceptualized by Triandis (1995) (Torelli et al.

2012; Zhang and Nelson 2016). Studies using other culture

frameworks rely on the traditional/secular–survival/self-

expression continuum from Inglehart and Baker (2000) (Van

der Lans, Van Everdingen, and Melnyk 2016), Schwartz’s

(1994, 1999) universal values (Shepherd, Chartrand, and Fitz-

simons 2015), or other sources of difference such as ethnic

identification (Bartikowski and Cleveland 2017). Our

approach differs from others stressing culture’s influence on

brand information processing: for example, Ng (2010) primes

thinking styles rather than measuring them, and although

Monga and John (2010) measure thinking styles, they focus

on extension, not parent perceptions.

Conceptual Framework and Hypotheses

While our primary interest is the influence of thinking styles on

spillover effects, we first describe established relationships

among VLE type and BC and spillover effects. Findings

regarding these elements are typically argued to be driven by

parent–extension “fit” or similarity (Aaker and Keller 1990).

Schema theory suggests that consumers organize brand knowl-

edge in memory-based structures called schemata (Fiske and

Taylor 1991). Brand schema are continuously updated with

relevant information (Weber and Crocker 1983). Schematic

knowledge facilitates information retrieval for making infer-

ences and decisions (Sujan and Bettman 1989). Such fit, from

transferring parent brand perceived quality, influences percep-

tions of extensions (Völckner and Sattler 2006) and parent

brands (Buil, Chernatony, and Hem 2009).

Although early fit research focused on brand-to-product

direction evaluations, more recent applications of schema the-

ory explain how image associations also transfer in the opposite

direction. For instance, Ng (2010) shows how negative

publicity associated with a newly failed product can harm a

parent brand. Similarly, Buil, Chernatony, and Hem (2009)

find that perceived fit between a parent brand and a fictitious

extension influences both perceptions of the extension and the

difference in post- and pre-extension perceptions of the Nike

parent brand.

Research has shown that conceptual fluency, or being pre-

sented with stimuli that bring an evaluated target to mind more

readily, leads to improved attitudes (Lee and Labroo 2004).

The term “processing fluency” refers to the “ease of mental

operations concerned with stimulus meaning and its relation to

semantic knowledge structures” (Reber, Schwartz, and Win-

kielman 2004). Previous studies have concluded that fluent

processing, or the ability to process information with greater

speed and efficiency, leads to positive affective reactions,

which in turn influences cognitive evaluations (Winkielman,

Halberstadt, and Catty 2006). Studies on brand extensions

leverage arguments related to fit, holding that consumer reac-

tions to extensions will be based on the relevance of parent

brand associations in the extension category (Park, Milberg,

and Lawson 1991) and the relevance of BC in judging such fit.

For example, concept consistency, or the extent to which an

extension and parent share abstract meanings, will be greater

for upward extensions of a prestige brand because they lead to

inferences of greater prestige than downward extensions

(Kirmani, Sood, and Bridges 1999). Upward prestige brand

extensions or downward nonprestige brand extensions would

be consistent, whereas upward (downward) extensions for a

nonprestige (prestige) brand would be inconsistent. As an

illustration, scenarios used by Kirmani, Sood, and Bridges

(1999) include a downward extension for BMW (prestige

brand) described as being the first BMW priced below its

current price range; for Acura (nonprestige brand), stimuli

instead stressed reliability.

In accordance with schema theory, congruence between

the parent brand schema and new brand-relevant informa-

tion regarding the extension will play a central role in how

VLE-related information affects the parent brand image,

regardless of consumers’ thinking styles. For example, with

a prestige parent brand such as BMW, an upward VLE

would be viewed as congruent, whereas a downward VLE

would be incongruent. With incongruent stimuli, reductions

in cognitive processing fluency become relevant for such

spillover effects. That is, information regarding the new

VLE would be processed more quickly and more efficiently

the greater the congruence with the existing parent BC, as

with an upward VLE for a prestige brand such as BMW.

Such reductions in processing fluency may be less for hol-

istic than for analytic thinkers because holistic thinkers will

be less concerned with reconciling incongruent attributes.

Dependent Variable: Reciprocal Spillover Effects

Brand extension research has highlighted an extension’s effect

on the parent brand’s overall image, called the “feedback

effect,” “reciprocal spillover effect,” or just “spillover effect”

4 Journal of International Marketing 27(2)

(Czellar 2003). Our focus is on a VLE’s reciprocal spillover

effects on parent brand image. We use the terms “feedback,”

“spillover,” and “reciprocal spillover” interchangeably.

VLE and Parent Brand Image

The fit between information attached to a VLE and existing

brand information determines attitude change toward a parent

brand. In line with schema theory, a lack of fit between VLEs

and parent brands will negatively affect parent brand percep-

tions (Gürhan-Canli and Maheswaran 1998; Loken and John

1993; Milberg, Park, and McCarthy 1997). Higher-quality

VLEs can improve brand evaluations, whereas VLEs

with below-average price/quality can reduce them (Heath, Del-

Vecchio, and McCarthy 2011; Lei, De Ruyter, and Wetzels

2008). Whereas upward VLEs increase variety and signal

brand quality, downward VLEs improve only variety apprai-

sals; positive variety signals may clash with signals of lower

sophistication, offsetting positive impacts (Heath, DelVecchio,

and McCarthy 2011). Because research has verified a link

between VLE type and parent brand image (Lei, De Ruyter

and Wetzels 2008), we offer no formal hypothesis.

BC and Parent Brand Image

According to Park, Milberg, and Lawson (1991, p. 186),

“A function-oriented BC is understood primarily in terms of

brand-unique aspects that are related to product performance,

while a prestige-oriented BC is understood primarily in terms

of consumers’ expression of self-concepts or images.” Accord-

ing to schema theory, quality-related associations are important

for functional brands such as Hyundai, and self-concept-related

associations are important for prestige brands such as BMW

(Kirmani, Sood, and Bridges 1999). For a prestige or functional

parent, VLEs may lead to incongruent signals that influence

parent perceptions, reducing processing fluency. While

expanding product lines can impede prestige band exclusivity

views (Kim and Lavack 1996), the potential dilution from

functional brands expanding upward or downward may be less

severe. Functional brands are less likely to be perceived as

exclusive; VLEs in either direction may be viewed as less

incongruent with the parent BC. As with VLE type, we offer

no formal hypothesis for BC’s main effect.

Interactive Effects Between VLE and BC on Parent Brand Image

In addition to brand- and VLE-related factors’ direct effects,

research supports BC and VLE type’s interactive effects on

spillover. According to schema theory, with a prestige (func-

tional) BC and downward (upward) VLE, signals could be

perceived as incongruent. With an upward VLE, the positive

impact on reciprocal spillover effects will be strengthened for

prestige brands and weakened for functional brands, regardless

of the culture in the extension’s market. In addition, a VLE that

is consistent with the parent brand image will result in greater

processing fluency than an inconsistent VLE, leading to more

positive reactions; an inconsistent VLE would not benefit the

parent in terms of greater fluency and enhanced evaluations. In

a cross-cultural context, Sichtmann and Diamantopoulos

(2013) find that parent and extension fit is positively associated

with the image of the parent’s country of origin (COO), which

is also linked to parent brand perceptions. Moreover, Pina,

Iversen and Martinez (2010) find that spillover effects depend

on extension evaluations, which themselves depend on the

extension’s fit with parent brand image regardless of national

culture. Because this interactive effect has been shown in pre-

vious research, we again offer no formal hypothesis.

Target-Foreign-Market Consumers: Impact of Thinking Styles

Our research questions focus on the impact of thinking styles as

brands are extended cross-culturally. Evaluations vary on the

basis of cognitive thinking styles, which are influenced by

social differences across cultures (Masuda and Nisbett 2001;

Nisbett 2003). In cultures with stronger social networks, people

have more integrative world orientations (Kuhnen, Hannover,

and Shubert 2001); they think holistically about the environ-

ment. Cognitive styles in cultures with weaker social networks

are more context independent; people consider the environ-

ment more analytically (Miyamoto, Nisbett, and Masuda

2006). These cognitive styles are sources of cross-cultural dif-

ferences in judgment and behavior.

As developed by Choi, Koo, and Choi (2007), this construct

has four dimensions: (1) causality, reflecting a focus on the

internal disposition of a focal object and its features versus

relationships and interactions between the object and its sur-

roundings; (2) locus of attention, which represents focusing on

an object versus the field in which it is embedded; (3) percep-

tion of change, which reflects tendencies to take a cyclical view

with constant fluctuations versus a more linear view that

assumes stability; and (4) attitude toward contradictions, which

reflects an assumption that contradictions can be reconciled

versus applying formal logic to choose one proposition. In the

context of VLEs, consumers with analytic thinking styles will

tend to focus more heavily than holistic thinkers on particular

information about the extension (e.g., VLE type) and associ-

ated parent brand attributes. Analytic thinking has also been

termed field independent, or oriented toward focal objects, as

opposed to holistic thinking, which is field dependent, or

focused on the field in which objects are embedded (Lalwani

and Shavitt 2013).

In our context, the field in which the brand and VLE are

embedded could represent broader associations between them,

such as overall brand reputation or competitive dynamics. As

argued by Choi, Koo, and Choi (2007), holistic thinkers see

everything as “intertwined” as opposed to separating object-

related information. Thus, we expect parent BC- and VLE-

related information to affect their evaluations less than analytic

thinkers’ appraisals. Next, we discuss our expectations regard-

ing cross-cultural differences in the effects of VLE type and

Allman et al. 5

parent BC. Building on the work of Monga and John (2010), we

emphasize causality and locus of attention, relevant in their

emphasis on processing object- (parent brand) and feature-

(BC and VLE type) related information as opposed to inter-

connections between an object and its environment.

VLE type. For VLE type, our interest is in judging whether thinking styles moderate the main effect of VLE type (upward

or downward) on spillover effects. We expect analytic and

holistic thinkers to differentially weigh information on the VLE

such that analytic thinkers will leverage it more heavily than

holistic thinkers in forming evaluations. Whereas analytic con-

sumers will attempt to isolate an object from its field and

categorize it with respect to its properties, holistic thinkers will

see the object as intertwined with its context and will focus less

on specific information in evaluating it (Ji, Peng, and Nisbett

2000). As a result, analytic consumers will be more prone to

search for and process information about the object and its

attributes that would help explain it. Because the VLE is an

extension of the parent brand, information about the VLE is

relevant for explaining the parent brand (evaluated object).

Therefore, we expect analytic thinkers to focus on information

regarding VLE type, resulting in stronger effects of this partic-

ular factor on their postextension parent brand image evalua-

tions. Thus, we hypothesize the following:

H1: VLE type has a stronger influence on analytic con-

sumers’ evaluations of postextension parent brand image

than on those of holistic consumers.

Parent BC. In the case of parent BC, consistent with our argu- ments regarding VLE type, we expect analytic and holistic

thinkers to differentially weigh the available information on

parent BC such that analytic thinkers will leverage it more

heavily in forming evaluations. Analytic thinkers tend to

attempt to isolate objects from their contexts and evaluate them

with respect to their properties (Ji, Peng, and Nisbett 2000).

Because the parent BC is a central property that helps position a

brand among target consumers (Park, Jaworski, and MacInnis

1986), it will be particularly relevant in forming parent brand

evaluations. Thus, we expect analytic thinkers to leverage

information regarding the parent BC more heavily in forming

evaluations, resulting in stronger effects of this factor on their

postextension parent brand image evaluations. Drawing on

these arguments, we hypothesize,

H2: Parent BC has a stronger influence on analytic con-

sumers’ evaluations than on holistic consumers’ evalua-

tions of postextension parent brand image.

Interactive effect of VLE type and parent BC. From a schema theory perspective, analytic thinkers will focus more on the

level of congruency among parent and extension product fea-

tures because of their tendency to emphasize relevant informa-

tion regarding a focal object (Choi, Koo, and Choi 2007). Thus,

analytic thinkers will tend to focus more on the level of exten-

sion (in)congruency between the new brand-relevant

information regarding the extension (e.g., an upward [down-

ward] VLE for a functional [prestige] parent informing their

evaluations). Alternatively, because holistic thinkers are more

likely to avoid separating the object from its surrounding envi-

ronment, characteristics of the focal object, such as its con-

gruency between its BC and the direction of the VLE, will be

less influential among these consumers. In line with these argu-

ments, we expect postextension parent brand image evaluations

to be affected more strongly by the interaction of the parent

brand (BC) and extension (VLE type) characteristics for ana-

lytic than for holistic thinkers. We hypothesize,

H3: The interactive effect of VLE type and BC has a

stronger influence on analytic consumers’ evaluations

than on holistic consumers’ evaluations of postextension

parent brand image.

Although we do not formally hypothesize a direction (positive

or negative) in terms of the impact of thinking styles on reci-

procal spillover effects, we expect the positive impact of an

upward VLE and negative effect of a prestige BC on spillover

effects to be greater for analytic as opposed to holistic thinkers.

We expect the damaging effect of a downward (upward) VLE

for a prestige (functional) brand to be even greater for analytic

thinkers as opposed to holistic thinkers.

Impact of different dimensions of thinking styles. Our second research question pertains to differences in specific thinking

style dimensions’ impacts. We view this as an empirical ques-

tion and offer no formal hypothesis. In Study 1 we extend the

work of Monga and John (2010) and examine items represent-

ing causality and locus of attention. In Study 2, we include all

items from Choi, Koo, and Choi (2007) in tackling our second

research question.

Study 1

Study 1 employs a scenario-based survey experiment with

actual automobile brands and hypothetical VLEs. We chose

this context because auto manufacturing operations are regu-

larly located in countries not associated with a brand’s origin.

In addition to reflecting practice, having the focal brand origin

not associated with either of the countries of manufacture

(COMs) may reduce a potential confound with brand origin

and COM. This industry has also been argued to be apt for

studying consumer-based brand equity, price, and other prod-

uct attributes (Fetscherin and Toncar 2010). It has been used to

study relationships between VLEs and brand attitudes, an ante-

cedent to brand image (Dall’Olmo Riley, Pina, and Bravo

2013), and global brand studies have also deemed it a suitably

rich context (Magnusson, Peterson, and Westjohn 2014; Talay,

Townsend, and Yeniyurt 2015).

This study was run among consumer panels recruited by a

web-based survey firm. We selected the United States and

India to increase the likelihood of respondents with analytic

(United States) and holistic (India) thinking styles (Lalwani

and Shavitt 2013; Monga and John 2007). We build on the

6 Journal of International Marketing 27(2)

work of Monga and John (2010), assessing thinking styles with

the items used in their study. We employ a 2 (BC: prestige/

functional) � 2 (VLE type: upward/downward) � 2 (cognitive style: holistic/analytic) between-subjects design, controlling

for brand familiarity. Research suggests that COM perceptions

can also affect brand evaluations (Chao 2001; Insch and

McBride 2004; Magnusson, Westjohn, and Zdravkovic

2011). We control for COM by including products from favor-

able and unfavorable countries. Based on pretests, pairings

include BMW (prestige) and Hyundai (functional) and two

COMs: Germany (favorable) and South Korea (unfavorable).

Pretests

Pretests with 112 undergraduate students at two major U.S.

universities confirmed the brand selection. Brand concept was

assessed using a seven-point poor–excellent scale, “Overall,

how would you rate the Hyundai (BMW) brand on the fol-

lowing dimension: Prestige.” The mean for Hyundai (M ¼ 4.21) was lower (t ¼ 17.03, p < .001) than for BMW (M ¼ 6.46). T-test results indicate that BMW is viewed as a prestige

brand and Hyundai as functional. We assessed category and

country image associations with the COM on a seven-point

(poor–excellent) scale: “Overall, how would you rate prod-

ucts made in Germany (South Korea)?” and “Overall, what is

your evaluation of cars made in Germany (South Korea)?”

(Diamantopoulos, Schlegelmilch, and Palihawadana 2011).

Subjects viewed products made in Germany (M ¼ 6.29) more favorably (t ¼ 16.25, p < .001) than those from South Korea (M ¼ 3.94). Cars made in Germany (M ¼ 5.96) were viewed more favorably (t ¼ 11.25, p < .001) than cars from South Korea (M ¼ 4.02).

We confirmed the brand–country pairs’ suitability before

our study through discussions with four faculty members and

six graduate students at a major Indian university. Although no

formal pretest was done in India, the discussions gave us con-

fidence in the pairings’ suitability. Results of post hoc tests in

the Indian sample also replicate U.S. pretest findings. The

control group data show that prestige ratings for Hyundai

(M ¼ 5.63) were lower (t ¼ �6.45, p < .005) than for BMW (M ¼ 6.46). For COM, Indian respondents rated products made in Germany (M ¼ 6.14) more favorably (t ¼�31.77, p < .005) than those from South Korea (M ¼ 4.30) and rated German cars (M ¼ 6.23) more favorably (t ¼ �11.68, p < .005) than cars from South Korea (M ¼ 5.42).

Scenario Development

Stimuli started with brand information, followed by a hypothe-

tical new VLE description. Scenarios described the brand name

and new VLE name and provided a narrative with key attri-

butes including price and COM. Prices were the same distance

from the brand’s average price for both VLE scenarios (31% price increase/decrease for Hyundai; 38% price increase/ decrease for BMW). The brands’ normal average price was

computed as the mean of the manufacturer’s suggested retail

price of all subcategory models the brand lists on its U.S.

website (MHyundai ¼ $26,110; MBMW ¼ $51,673). For India, prices were based on the grand mean of prices for all Hyundai

and BMW models on their Indian websites. We chose the VLE

name, depiction, and price to be realistic and consistent with

industry and brand standards for the midsize four-door sedan

subcategory. Prices are listed in Table 2. Price increases/

decreases for VLEs were consistent with prior research

(Dall’Olmo Riley, Pina, and Bravo 2013). Differences in per-

centage increase/decrease between Hyundai and BMW (31% vs. 38%) ensured realism. Next, subjects rated parent brands and answered items on thinking styles, brand familiarity, and

demographics. For the scenarios, see Appendix A.

Scales and Measurement

Parent brand image was measured with items adapted from

previous studies. The brand image measure consisted of

four seven-point (poor–excellent) items: innovativeness,

design, prestige, and workmanship, adapted from Roth and

Romeo (1992). We assessed internal consistency reliability

of this composite dependent measure using Cronbach’s

alpha. This four-item Likert scale measure displayed ade-

quate reliability (a ¼ .91 for both the samples). Further- more, dropping any item did not improve reliability; thus,

we deemed the four-item measure to be reliable. We aver-

aged responses to the four statements to arrive at a parent

brand image score.

Consistent with Monga and John (2010), we measured cog-

nitive styles using six items representing two factors (causality

and locus of attention) from the Choi, Koo, and Choi (2007)

scale. (For the items, see Appendix C.) Respondents rated their

agreement on a seven-point scale (1 ¼ “strongly disagree,” and 7 ¼ “strongly agree”). Analytic (holistic) thinking styles resulted in lower (higher) means. The coefficient alpha is .75,

which is adequate because the construct measures a concept

with great breadth (Choi, Koo, and Choi 2007). We asked a

qualifying question on brand familiarity on a seven-point scale

(1 ¼ “not at all familiar,” and 7 ¼ “very familiar”). Only responses of three or greater were included.

Table 2. Sample Breakdown by Stimuli: Studies 1 and 2.

BC VLE Type COM

Study 1 Samples

Study 2

India U.S. U.S.

Hyundai (functional)

Downward (9.90 lakh/$15,990)

South Korea 50 50 48 Germany 53 53 40

Upward (22.30 lakh/$36,190)

South Korea 52 52 44 Germany 52 52 44

BMW (prestige)

Downward (19.10 lakh/$30,990)

South Korea 52 53 50 Germany 54 53 44

Upward (47.10 lakh/$76,210)

South Korea 53 52 46 Germany 54 55 41

Allman et al. 7

Data Collection

After the pretest, the study included (1) a main survey with

respondents randomly assigned to one of eight product sce-

narios and (2) a control group survey that gauged initial

images of functional and prestige brands. A total of 420

respondents from the United States and 420 from India parti-

cipated in the main survey, while 110 U.S. respondents and

267 from India were randomly assigned to functional or pres-

tige brand conditions as part of our control group. All were

randomly assigned to control or main groups. Appendix B

provides respondent characteristics.

Analysis and Results

Our first step was to compute a distance brand image score to

assess spillover effects. This figure represents the distance (dif-

ference and direction) between main and control group respon-

dents’ composite parent brand image scores (main group

individual brand image score � control group’s mean brand image score). Positive (negative) scores show improvement

(decrease) in parent brand image. This score was the dependent

variable in our generalized linear model (GLM) analysis of

variance (ANOVA).

We first assess measurement equivalency of scales across

U.S. and Indian samples using procedures endorsed by Steen-

kamp and Baumgartner (1998). First, to assess configural

invariance, we perform exploratory factor analysis on the

U.S. and Indian responses separately and then on the pooled

sample. Items load similarly in the two groups. We then run a

confirmatory factor analysis with loadings first constrained to

be equal across samples and then estimated freely. A chi-square

difference test in which the constrained and unconstrained

models are compared results in a nonsignificant chi-square

(w2diff ¼ 5.57, d.f.diff ¼ 3). A similar test to assess equivalence of factor covariance results in significant difference values

(w2diff ¼ 20.2, d.f.diff ¼ 6). Although the chi-square for the second test is significant, because it is sensitive to sample size,

we compare other statistics. The difference in Tucker–Lewis

index for the models for invariance of loadings is .03 and

difference in root mean square error of approximation is

.002; for models testing for invariance of factor covariations,

the difference in Tucker–Lewis index is .001 and difference in

root mean square error of approximation is .001. This indicates

that, practically, little differentiates model fit, suggesting

metric equivalency. Thus, we combine U.S. and Indian samples

for the rest of our analyses.

To address our second research question, we assess differ-

ences in thinking styles across samples. As expected, Ameri-

cans have lower mean ratings (t ¼ �10.416, p < .001) than Indians (higher ratings indicate holistic thinkers). Comparing

country groups rather than those based on a median split on

thinking styles, we found a greater proportion of analytic thin-

kers among U.S. (80%) than Indian (40%) respondents (w2 ¼ 60.342, p < .001). We examined results using the causality and locus-of-attention dimensions individually and find results for

causality to be consistent with those with the full scale, with a

significant causality � VLE type interaction (F-statistic ¼ 5.70, p < .017), a nonsignificant two-way interaction with BC, and a nonsignificant three-way BC � VLE type � caus- ality interaction. For locus of attention, the interaction with

VLE is marginally significant (F-statistic ¼ 2.65, p < .10). Other coefficients are consistent with those for causality.

GLM ANOVA Results

Next, we conduct a GLM ANOVA with distance brand image

as the dependent variable and VLE type, BC, and thinking style

as independent variables. We included the COM of the VLE,

familiarity with the brands, and nationality as controls. The

results (Table 3) show that both VLE type and BC main effects

on parent brand image are significant. A significant two-way

interaction between VLE type and consumer thinking style is

also evident. A three-way interaction among the factors (VLE

type, BC, and thinking style) is not significant.

Mean comparisons show that consumers upgrade their

parent brand postextension image after an upward VLE and

downgrade them after a downward VLE (MupwardVLE ¼ .273 > MdownwardVLE ¼�.033). Consumers also upgrade their parent brand postextension image of functional brands and

downgrade them for prestige brands (Mfunctional ¼ .497 > Mprestige ¼�.257). The main effect for holistic versus analytic thinking styles is also significant (Mholistic ¼ .430 > Manalytic ¼ �.190). The BC � VLE interaction is nonsignificant (F-statistic ¼ .123, p < .727).

For our hypotheses regarding consumers’ thinking styles,

we find a significant interaction between thinking styles and

upward versus downward extension (F-statistic ¼ 6.184, p < .014), in support of H1. H2 is not supported, as the interaction between thinking styles and BC is nonsignificant

(F-statistic ¼ .235, p < .628). To explore the nature of the significant interaction between thinking styles and VLE type,

we compare means on spillover effects for both upward and

downward VLEs and analytic and holistic thinking styles

Table 3. Study 1 Postextension Brand Image Change: GLM ANOVA Results.

Hypothesis F-Statistic p-Value

BC 35.348 .000 Vertical line extension type (VLE) 5.908 .016 Thinking styles (HA) 17.486 .000 BC � VLE .123 .727 VLE � HA H1 6.184 .014 BC � HA H2 .235 .628 BC � VLE � HA H3 .001 .978 Controls

COM 2.600 .108 Nationality .578 .448 Familiarity BMW .280 .597 Familiarity Hyundai 9.304 .003

8 Journal of International Marketing 27(2)

in Figure 1. Finally, a nonsignificant three-way interaction

leaves H3 unsupported (F-statistic ¼ .001, p < .978). To further assess the interaction between VLE type and

thinking styles, t-tests compared mean postextension parent

brand image evaluations after an upward versus a downward

VLE for analytic thinkers and holistic thinkers. For analytic

thinkers, the negative impact of an upward VLE is relatively

lesser than a downward VLE (t ¼ 2.517, p < .013), whereas for holistic thinkers there is no significant difference in the impact

on the postextension brand image based on an upward versus

downward VLE (t ¼�.246, p < .806). These results are con- sistent with H1. The VLE type influences parent brand spillover

effects more for analytic thinkers than holistic thinkers.

Regarding our central concern, consumers’ thinking styles,

results of post hoc analyses reveal similar patterns as previ-

ously. Analytic thinkers upgrade postextension parent brand

image evaluations with an upward extension and lower their

assessments with downward extensions. In contrast, holistic

thinkers appear to find some merit in all cases in terms of their

parent brand views, improving postextension brand image

assessments after an extension irrespective of the VLE type.

We discuss these findings next.

Study 1 Discussion

Study 1 results suggest that for analytic thinkers, introducing an

upward versus downward VLE may offer a greater payoff in

terms of reciprocal spillover effects. We find no evidence of

interactive effects of thinking styles and BC or the three-way

interaction between these variables and VLE type on postex-

tension brand perceptions. Regarding our second research

question, we find that relying on nationality would have led

to a misclassification of a sizable number of respondents

(20% of U.S. and 40% of Indian samples). Furthermore, these results suggest that global marketing managers should be wary

of segmenting consumers on the basis of country; within-

country differences in thinking styles may prove to be impor-

tant in designing strategies for introducing VLEs and managing

potential spillover effects on parent brands. The results also

suggest that VLE direction matters regardless of BC. Spillover

effects after an upward VLE are higher than those after a down-

ward VLE. Next, we discuss Study 2.

Study 2

In Study 2, we sought to confirm our finding that culture differs

within a country as well as across countries. As such, we aim to

replicate our results from Study 1 in a single country: the

United States. We also utilize the full thinking styles scale to

assess whether other dimensions might also differ within coun-

tries and moderate the impact of BC and VLE on spillover

effects. Study 1 followed the approach of Monga and John

(2010), including items for only the causality and locus-of-

attention dimensions. However, conceptual arguments for the

impact of other dimensions are also consistent with our expec-

tations. First, arguments based on schema theory are consistent

with the notion of attitude toward contradictions. Holistic and

analytic thinkers are likely to respond differently to contradic-

tory propositions, as with incongruent stimuli such as an

upward VLE for a functional brand. With incongruent stimuli,

analytic thinkers are more inclined to reject one and favor the

other in forming evaluations. Holistic thinkers follow the prin-

ciple of the “Middle Way”; they are inclined to find merit in

multiple propositions (Peng and Nisbett 1999). Some VLE

scenarios entail contradictions—that is, downward (upward)

VLEs for prestige (functional) brands. In such cases, analytic

thinkers will more closely scrutinize seemingly conflicting

inputs and parse them into independent elements; they will

follow rules to conclude congruency. Holistic thinkers will

look for “harmony,” embracing paradox (Chen 2002).

It is also possible that assumptions of constant fluctuations

versus a more linear view that assumes stability (i.e., percep-

tions of change) could influence approaches to processing

VLE-related information when evaluating parent brands.

According to Choi, Koo, and Choi (2007), because of their

more linear perspective, analytic thinkers are more likely to

produce rules related to internal properties of evaluated object.

Thus, they may be more confident in utilizing relevant infor-

mation in forming evaluations. Therefore, we aimed to verify

whether dimensions of other thinking styles might explain spil-

lover effects and offer insight regarding unsupported hypoth-

eses. We also used an alternative experimental design gauging

parent brand perceptions both before and after scenario expo-

sure to offer further evidence for our findings.

Scenarios and Measures

Study 2 utilized a scenario-based pretest–posttest survey

experiment with the same brands and hypothetical VLEs as

.121

−.501

.426 .434

−1.0

−.8

−.6

−.4

−.2

0

.2

.4

.6

.8

1.0

Upward Downward

Analytic Holistic

Figure 1. Study 1: Change in parent brand image for upward versus downward VLE based on thinking styles.

Allman et al. 9

in Study 1. Descriptions included each of eight hypothetical

VLE key variables and one control variable of interest: BC,

VLE, and COM. We revised the scenarios to ensure that both

upward and downward VLE versions present equal numbers of

essential functional features (e.g., backup camera, blind-spot

technology) to preclude these features’ influence on postexten-

sion responses. Appendix A presents all scenarios. For compar-

ability, we kept the average brand prices and new VLE prices

identical to those in Study 1.

Because one of our goals was to understand whether par-

ticular dimensions of thinking styles might be driving Study 1

results, we assessed this variable by using all 24 items from

the Choi, Koo, and Choi (2007) scale. Thus, the survey

included all four holistic–analytic (HA) dimensions. Appen-

dix C lists the items used in each study. Six items were

reverse-coded. For each respondent we calculated the mean

on each dimension and then used a median split to create a

dichotomous variable for each. We used the median of each

participant’s grand mean to create a dichotomous variable

classifying them as holistic or analytic.

We utilized Qualtrics, collecting data electronically from a

representative U.S. consumer panel recruited by an online sur-

vey distribution firm. Respondents were paid a nominal fee for

participating. Appendix B summarizes the sample. We ran-

domly assigned participants into a prestige (BMW) or func-

tional (Hyundai) BC flow and then randomly assigned them

into one of four scenarios within each brand. After being pre-

screened (for country of permanent residence and country of

birth), respondents rated brands on items from Roth and

Romeo’s (1992) scale. To avoid demand effects, we included

a filler task requiring them to respond to an open-ended tech-

nology question and a series of questions from the technology

acceptance model scale (Ashraf, Thongpapanl, and Auh 2014),

a task also unrelated to our study. We eliminated those who did

not answer both comprehensively. The survey then presented a

randomly assigned VLE scenario and the four brand image

questions to assess the change in brand image after the VLE.

Finally, we asked questions on COM perceptions and demo-

graphics. We calculated spillover brand image as the difference

between the mean of the items after versus before the scenario.

Cronbach’s alpha in each of the ten measurement points (two

pre- and eight postscenario measurements) indicates adequate

reliability, ranging from .848 for prescenario BMW to .936 for

post-Hyundai–upward VLE–South Korea COM scenario.

Data Analyses and Results

We tested our hypotheses in a 2 (BC: prestige vs. functional) � 2 (VLE: upward vs. downward) � 2 (cognitive style: analytic vs. holistic) between-subjects design. We used spillover effects

as our dependent variable in all GLM ANOVA models. We ran

the GLM ANOVA with the full HA scale dichotomous variable

first and then again with the individual dimensions of the scale.

Each included COM and brand familiarity as covariates, con-

sistent with Study 1. The full-factorial ANOVA with full HA

scale with dichotomous classification, VLE, and BC as factors

yielded the results in Table 4. The results show a main effect of

BC and VLE on spillover; a significant two-way interactive

effect of VLE and HA (F-statistic ¼ 5.242, p < .05), in support of H1; and a nonsignificant two-way interactive effect of BC

and HA (F-statistic ¼ 1.692, p < .194), failing to support H2. Finally, a nonsignificant three-way interaction for BC, VLE,

and HA (F-statistic ¼ .004, p < .951) again fails to support H3. We also ran our analyses using thinking styles computed using

the exact items used in Study 1 and found consistent results.

We provide plots of the interaction effects in Figure 2.

Although we see some similarities in the patterns of results

as in Study 1, there are some differences. For Study 2, as with

Study 1, the impact of downward VLE on parent brand image is

relatively greater (and more negative) than an upward VLE (t ¼ 6.46, p < .001) for analytic thinkers. However, for holistic thinkers, whereas we observed no significant difference in the

impact on the postextension brand image based on VLE type in

Study 1, we now observe a significant difference, with the

impact of downward VLE on parent brand image relatively

greater (and more negative) than an upward VLE (t ¼ 2.07, p < .040). The interaction still reveals support for H1, with a greater impact of information regarding VLE type on

parent brand spillover effects for analytic thinkers than hol-

istic thinkers.

Table 4. Study 2 GLM ANOVA Results Postextension Brand Image Change.

Hypothesis

F-Statistic (p-Value)

HA: All Dimensions Causality Locus of Attention Attitude Toward Contradictions

Perceptions of Change

BC 36.348 (.000) 38.836 (.000) 34.220 (.000) 34.653 (.000) 37.975 (.000) VLE type (VLE) 34.477 (.000) 36.088 (.000) 32.704 (.000) 32.982 (.000) 34.970 (.000) Thinking styles (HA) 2.714 (.100) .001 (.973) 4.101 (.044) 1.041 (.308) .225 (.636) BC � VLE .951 (.330) 1.319 (.251) .744 (.389) .957 (.329) 1.255 (.263) VLE � HA H1 5.242 (.023) 4.708 (.031) .002 (.963) 4.354 (.038) .897 (.344) BC � HA H2 1.692 (.194) 2.803 (.095) .578 (.448) .058 (.809) 4.489 (.035) BC � VLE � HA H3 .004 (.951) .343 (.559) 1.777 (.183) 1.250 (.264) .835 (.361) COM 4.103 (.044) 5.117 (.024) 5.093 (.025) 4.128 (.043) 5.131 (.024) Familiarity .005 (.945) .001 (.969) .204 (.652) .074 (.786) .001 (.981)

10 Journal of International Marketing 27(2)

Table 4 also includes results based on each of the four HA

dimensions separately. We find interesting differences. First,

for the interaction of each dimension and VLE (H1) we find

significant spillover effects from the causality and attitude-

toward-contradictions dimensions, but not the others; this sug-

gests that our Study 1 results may have been driven at least in

part by causality. For BC � HA (H2), in contrast to our Study 1 findings based on items representing causality and locus-of-

attention dimensions, we find significant spillover effects for

the interaction of causality and perception-of-change dimen-

sions individually with BC. Finally, for H3—the three-way

interaction between thinking styles, BC, and VLE type—the

coefficients are nonsignificant for all dimensions, consistent

with Study 1 results. Examining the plots of the significant

interaction effects from individual HA dimensions (Figure 3,

Panels A–C), we see similar patterns of results for the causality

and attitude-toward-contradictions dimensions. In both cases,

the impact of downward VLE on parent brand image is rela-

tively greater (and more negative) than an upward VLE for

both analytic and holistic consumers (causality: tanalytic ¼ 5.50, p < .001; tholistic ¼ 2.77, p < .001; locus of attention: tanalytic ¼ 4.38, p < .001; tholistic ¼ 3.62, p < .001). Regarding the interaction with BC and causality (Figure 3, Panel C), the

impact of an extension for a parent with a prestige BC on parent

brand image versus a parent with a functional BC is greater for

analytic than holistic consumers (tanalytic ¼ 5.47, p < .001; tholistic ¼ 3.14, p < .002).

Study 2 Discussion

Our results from Study 2 largely corroborate those of Study 1 in

terms of supporting H1 and failing to support H2 and H3.

Regarding the effect of VLE type (H1), we find significant

spillover effects from the causality and attitude-toward-

contradictions dimensions, but not others. For BC (H2), we find

significant spillover effects for the interaction of causality and

perception-of-change dimensions individually with BC but not

the full scale or other dimensions. In teasing out the effect of

causality and incorporating the perception-of-change dimen-

sion, we detect cross-cultural differences’ moderating effect

on the relationship between parent BC and spillover. This may

at least in part explain our lack of support for H2 in both studies.

That is, it seems that consumers’ tendency to take a more

linear view that assumes stability versus a cyclical view with

constant fluctuations enhances the importance of parent BC

for spillover effects. The lack of findings for the overall HA

scale and the other dimensions for the relationship between

BC and postextension parent evaluations may also explain our

lack of support for H3. That is, the influence of parent BC on

spillover effects, coupled with the more limited effects of only

one dimension of thinking styles, may have reduced our abil-

ity to detect the three-way interaction between BC, VLE type,

and thinking styles.

General Discussion

We aimed to better understand the role of consumers’ thinking

styles in a firm’s global brand expansion strategy. To this end,

we examined how consumers’ cognitive style influences parent

brand image when parent BC and VLE type vary. We connect

that evidence to our research questions and then explore impor-

tant theoretical and managerial implications.

RQ1 asked how consumers with holistic versus analytic

cognitive thinking styles differentially weigh information

regarding the upward versus downward nature of a foreign-

market VLE and parent BC in their postextension brand image

evaluations. We find differences in the influence of brand strat-

egy characteristics based on consumers’ thinking styles.

Although introducing an upward VLE—or VLEs for parent

brands with a functional BC in general—leads to more positive

results in terms of postextension parent brand image, the detri-

mental impact of not following these approaches seems to be

more severe among consumers with more analytic thinking

styles. That is, information pertaining to an extension’s strategy

affects analytic thinkers’ postextension brand evaluations more

than holistic thinkers’ evaluations. Analytic thinkers appear to

focus more heavily on VLE characteristics—namely, the

extension’s direction—in evaluating the parent brand.

RQ2 asked how, based on differences in thinking styles and

their influence on postextension brand image evaluations, mar-

keters should target customers within and across markets when

introducing VLEs. That is, do consumer thinking styles vary

within country as well as across countries, and are nationality

or thinking style differences more important in explaining dif-

ferences in the impact of VLE type and parent BC on spillover

effects? Although, overall, thinking styles for Indian and U.S.

respondents were in line with expectations (analytical for

United States and holistic for India), we find that, in terms of

the influence of thinking styles, results differ when consumers

are classified on the basis of nationality as opposed to using

.592

−.120

.247

−.064

−.2

−.1

0

.1

.2

.3

.4

.5

.6

.7

Upward Downward

Analytic Holistic

Figure 2. Study 2: Change in parent brand image for upward versus downward VLE based on thinking styles.

Allman et al. 11

individual measures of thinking styles as an aspect of culture.

This result underscores a challenge for marketers in developing

brand strategies that account for heterogeneity in consumer

thinking styles both across and within markets. In addition, our

finding that culture’s influence varies on the basis of the par-

ticular thinking style dimensions we examined highlights inter-

esting opportunities for managers for segmenting consumers

both across and within markets. We find that causality (i.e.,

differences in the tendency to concentrate on a focal object vs.

relationships between an object and its surroundings) appears

to drive variation in the impact of both VLE type and BC on

postextension parent evaluations. However, while attitude

toward contradictions also moderates the VLE type’s influence

on spillover, perceptions of change (i.e., a tendency to assume

stability vs. constant fluctuation) matters for BC’s impact.

Theoretical Implications

Our results present interesting implications for theory as well as

the study of branding and factors influencing the success of

A: Change in Parent Brand Image Based on VLE Type: Causality Dimension

.526

−.189

.332

−.001

−.3

−.2

−.1

0

.1

.2

.3

.4

.5

.6

Upward Downward

Analytic Holistic

B: Change in Parent Brand Image Based on VLE Type: Attitude Toward ContradictionsDimension

.542

−.150

.266

−.054

−.2

−.1

0

.1

.2

.3

.4

.5

.6

Upward Downward

Analytic Holistic

C: Change in Parent Brand Image Based on BC: Causality Dimension

−.177

.514

Prestige −.033

.367

−.3

−.2

−.1

0

.1

.2

.3

.4

.5

.6

Functional

Analytic Holistic

Figure 3. Study 2: Comparison of results using individual dimensions of thinking styles.

12 Journal of International Marketing 27(2)

firms’ global brand strategies. We explore these implications in

the following subsections.

Culture and thinking styles. Our findings regarding the influence of consumers’ thinking styles, in conjunction with findings

regarding potential misclassification of consumers when

using nationality as a proxy for thinking style, highlight the

potential risk of assuming that cultural influences are uniform

within a country. Whereas some researchers suppose that cul-

tures are on a path to convergence as a result of industrializa-

tion and modernization, others contend that cultural values

are independent of changing economic conditions, so conver-

gence is unlikely (Inglehart and Baker 2000); therefore,

accounting for cultural differences remains a critical issue for

marketers. In addition, research recognizes that consumers

can be multicultural, with values and practices from more

than one culture (Briley, Morris, and Simonson 2005), rein-

forcing the need to consider cultural differences at the indi-

vidual level. Our finding that thinking styles can vary within

national borders, at least for the two countries in our sample,

is consistent with the notion of heterogeneity in culture among

consumers in a given market.

Our finding that the effects of thinking styles depend on the

particular dimension being examined calls attention to the

complex nature of culture and its potential influences. We

observe a larger drop in parent brand image after a downward

versus upward VLE for those more likely to focus on the object

(brand) and its features (VLE type) (reflecting the causality

dimension). We also observe a larger drop for prestige versus

functional BC for consumers with more of an object-focus

perspective and those who use a linear perspective and assume

stability (reflecting the attitude-toward-contradictions dimen-

sion; Choi, Koo, and Choi 2007). These findings support the

notion that an extension represents a departure from a parent

brand’s concept, which may enhance the impact of a parent BC

on postextension brand perceptions more for those who tend to

follow a linear approach to information processing. In addition,

the direction of the VLE, which is related to the brand itself as

opposed to relationships with other objects in its field, has a

stronger influence in terms of spillover effects for consumers

with a tendency to focus on an object and its features.

Our results may also offer further insight into cultural dif-

ferences in approaches to processing information and the dif-

ferent roles of the individual dimensions of thinking styles. As

opposed to treating thinking styles as a continuum from holistic

to analytic, or using a subset of scale items, as previous studies

have done (e.g., Badrinarayanan et al. 2012; Monga and John

2010), given the different effects of the individual dimensions

demonstrated we urge researchers to consider all HA dimen-

sions and investigate their effects. Just as Triandis’s (1995)

observation that the widely used individualism–collectivism

aspect of culture can further be differentiated based on a hor-

izontal–vertical distinction, the influences of tendencies along

the HA continuum may not fully capture complexities in the

influence of thinking styles.

Schema theory. We offer implications regarding the value of schema theory (Mandler 1982) for developing expectations

for reciprocal spillover effects from VLE strategies. Con-

sistent with schema theory, our results support the notion

that highly incongruent combinations of incoming infor-

mation can be detrimental for parent brand image. How-

ever, thinking styles may limit how well schema theory

predicts reciprocal spillover effects. Drawing on additional

analyses comparing mean spillover effects from COM, we

find that holistic thinkers may not always respond in a

manner consistent with schema theory. When faced with

the incongruent combination of a downward VLE produced

in a favorable COM, holistic thinkers do not downgrade their

parent brand image evaluations (Mhol, downVLE, favCOM ¼ .684 vs. Mhol, upVLE, favCOM ¼ .538; n.s.; for charts of these mean comparisons, see Web Appendix A). Indeed, spil-

lover effects from a downward VLE and upward VLE

produced in an unfavorable COM are virtually identical

(Mhol, downVLE, unfavCOM ¼ .266 vs. Mhol, upVLE, unfavCOM ¼ .278; n.s.). By embracing both pieces of information (Peng and

Nisbett 1999), holistic thinkers react less to incongruence

between brand strategy and market conditions.

Given linkages between the HA and self-construal frame-

works (Kuhnen, Hannover, and Schubert 2001), our results

also suggest other links between cultural differences based on

self-construal and thinking styles. Research suggests that con-

sumers with an independent self-construal tend to adopt an

analytic thinking style, and those with an interdependent self-

construal adopt holistic ones (Lalwani and Shavitt 2013).

Moreover, people with an independent self-construal rely on

affective feelings in making decisions, whereas those with an

interdependent self-construal rely on cognitive reasoning

(Hong and Chang 2015). Building on research on affect and

risk taking (Loewenstein et al. 2001), our results suggest that

holistic thinkers, due to their reliance on affective feelings,

may also be more risk averse than those who rely on reason-

ing. Although we are unable to offer direct evidence of such

risk aversion, we encourage future research to investigate this

issue further.

The study of COM influences. Drawing on research on the role of COO (Papadopoulos et al. 2011), we conducted post hoc anal-

yses to determine whether our results hold based on COM. We

reran the GLM ANOVA for a favorable COM (Germany) and

unfavorable COM (South Korea). Consistent with prior

research (Chao 2001), results (See Web Appendices A, B, and

C.) reveal detrimental impacts of an unfavorable COM for

parent brand image, particularly for VLEs for prestige parents

(MunfavCOM, prestige ¼ �.345 < MfavCOM, prestige ¼ �.158) and for downward VLEs (MunfavCOM, down ¼ �.152 < MfavCOM, down ¼ .149). These results highlight potential limits in the COO’s brand evaluation impact. Consistent with other

research in the automotive context (Pappu, Quester, and Cook-

sey 2007), country of brand origin can influence brand evalua-

tions; however, these results hold only under certain conditions

related to VLE type or BC.

Allman et al. 13

The study of brand strategy. We also offer implications for the study of brand strategy. Research suggests that offering prod-

ucts within a parent brand’s product class or category can

reduce risk with lower launch costs, enabling parent brand

images to be leveraged in the same category (Loken and

Amaral 2010). Yet our results highlight risks of this decision.

We reveal detrimental effects of introducing a prestige brand

VLE regardless of market characteristics or VLE strategy. We

also contribute to research on conditions under which brand

performance may vary across markets on the basis of differences

in buyer information use in decision making (Grewal et al. 2013;

Ng 2010). We show that parent brand dilution effects demon-

strated previously (Milberg, Park, and McCarthy 1997) may

depend on other factors. Consistent with Sood and Keller

(2012), whose results appear to contradict this congruency

expectation, our results suggest boundary conditions in parent–

extension congruency’s influence on spillover effects.

Managerial Implications

We offer direction to managers as to how VLE type and BC can

enhance positive or reduce negative impacts in foreign mar-

kets. When launching brand extensions, managers must care-

fully weigh variables such as promotional expenditures,

packaging, and labeling (Keller 2003; Keller and Aaker

1992). Decisions regarding market selection and segmenting

and targeting consumers can also influence the overall success

of a firm’s global brand strategy.

Influence of culture on consumers’ approaches to processing information. In addition, our results have implications for how firms approach cross-cultural communications. With greater

understanding of consumers’ information processing

approaches, firms can aim to shape reactions to extensions

based on the strategic use of communications, similar to the

suggestion of Zhu and Myers-Levy (2009) in their study of the

impact of product displays on context effects from consumers’

modes of cognition. Communications could be designed based

on tendencies consistent with the causality and attitude-toward-

contradictions dimensions in particular. For example, because

information regarding parent BC is more influential in terms of

postextension parent evaluations for consumers with more of

an object-focus perspective, extension-related messages could

stress the parent BC. Furthermore, because parent BC is also

influential for those who use a linear perspective and assume

stability, these messages could emphasize the parent brand’s

traditions, linking that notion to expectations of consistency in

both the parent and extension delivering on its BC. In line with

findings that VLE-related information is more influential for

those likely to focus on the object (brand) and its features (VLE

type), messages focused explicitly on the extensions’ features

may be particularly effective.

Segmenting based on culture versus nationality. Our findings regarding thinking style’s heterogeneity within countries sug-

gest that firms should consider nationality-based segments with

caution. We reran our Study 1 analysis using nationality instead

of the HA measure. We find a significant interaction between

BC and nationality, and mean comparison based on nationality

reveals differences relative to those based on the HA scale. We

find significant main effects for BC (F-statistic ¼ 86.234, p < .001), VLE type (F-statistic ¼ 27.868, p < .001), and nation- ality (F-statistic ¼ 9.719, p < .002) and interaction effects for BC � nationality (F-statistic ¼ 18.645, p < .001) and VLE type � nationality (F-statistic ¼ 8.169, p < .004). We find a nonsignificant interaction for BC � VLE type (F-statistic ¼ .204, p < .651), and a nonsignificant three-way interaction (F-statistic ¼ 1.346, p < .246).

It appears that while holistic thinkers’ postextension brand

perceptions tend to be higher than those of analytic thinkers

for both prestige and functional brands, the comparison for

U.S. and Indian respondents does not follow the same pattern

(Figure 4). U.S. and Indian respondents’ postextension pres-

tige brand perceptions do not vary, but U.S. respondents

report higher spillover effects than Indian respondents for

functional brands. For VLE type, results differ on the basis

of whether culture is captured through nationality or thinking

styles; holistic thinkers give higher evaluations than analytic

thinkers for both VLE types, yet Indian respondents give

lower evaluations than U.S. respondents for functional brands

(no difference for prestige brands).

These results, coupled with the misclassification of a sub-

stantial number of respondents if we use nationality as a proxy

for thinking styles, calls into question assumptions that think-

ing styles differ based on nationality. Indeed, Hispanic Amer-

icans, Asian Americans, and consumers from certain states are

more likely to think holistically compared with other groups in

the United States (Monga and John 2010). We note supportive

anecdotal evidence regarding risks in assuming cultural homo-

geneity within countries. Analysts attribute Walmart’s troubles

in China in part to its lack of appreciation for heterogeneity in

Chinese consumers’ preferences across cities (Salomon 2016).

Thus, our results should serve as a warning about the pitfalls of

assuming homogeneity within markets.

Potential for priming thinking styles. Finally, in addition to insights regarding how to effectively target consumers on the basis of

psychographic characteristics (Samuel 2016), considering our

findings in light of studies such as those of Monga and John

(2010), which demonstrate the use of situational primes to

increase accessibility of thinking styles, managers may be able

to influence brand evaluations based on extensions. Specifi-

cally, studies such as these demonstrate how exposing consu-

mers to cues can cause them to shift between different thinking

styles. Thus, cues embedded in the environment or in commu-

nications may be an effective tool for shaping consumer reac-

tions to extensions and, ultimately, parent brand perceptions.

Limitations and Future Research Directions

Despite this study’s comprehensiveness, it has some limitations

in scope and methodology. First, we conducted our study in

14 Journal of International Marketing 27(2)

only two countries, so caution should be used in drawing con-

clusions about how consumers’ thinking styles affect results.

Future research exploring these relationships across a wider

range of countries could enable researchers to tease out other

country differences (e.g., stage of economic development) that

may influence these findings. We also measure country image

at a single point; however, it may change with time (Jaffe and

Nebenzahl 2001). Thus, an interesting avenue for research

would be longitudinal studies to detect shifting country images

and changes in the relationships uncovered here.

Next, we focus on the automobile industry, and we urge

caution in attempting to generalize our results to other sectors.

Because country image’s impact can vary on the basis of cate-

gory (Elliott and Cameron 1994; Pappu, Quester, and Cooksey

2007), results may differ if tested in other categories. We urge

extensions of our results to other sectors, particularly those that

differ from durable goods (e.g., consumer packaged goods).

Our focus on the automotive category was based on the indus-

try’s common usage in COO research (Han and Qualls 1985;

Häubl 1996; Pappu, Quester, and Cooksey 2006, 2007). Relo-

cating production to countries not associated with a brand’s

origin is a trend, as indicated by many original equipment

manufacturers, joint ventures (Gao, Hensley, and Zielke

2014), and assembly lines operating outside brands’ home mar-

kets (Isidore 2014). While not our focus, the relocation decision

is an interesting context in which to explore how COB–COM

incongruency affects relationships among variables in our

studies.

We focus specifically on consumers’ thinking styles. Other

research emphasizes differences based on values (Hofstede

1984), knowledge structures (Briley, Morris and Simonson

2000), behaviors, or information processing speed (Hall and

Hall 1990). Thus, other cultural factors could influence our

findings. Given our finding that thinking styles may not align

with nationality, future research might consider these con-

cepts at the individual level to further expand understanding

of culture’s influence. We also concede that consumers in

India and the United States may differ on other dimensions

not accounted for here. Future research might control for

other differences in exploring culture’s influence on market-

ing concepts.

Finally, despite our theorizing regarding the importance of

fit between the existing parent brand schema and new VLE-

related information for parent brand image perceptions, we

acknowledge that we do not empirically measure perceived fit.

As in previous studies leveraging similar theoretical arguments

(Kirmani, Sood, and Bridges 1999), we postulate that congru-

ence occurs when an upward (downward) VLE is paired with a

prestige (functional) brand, and our pretest verified that respon-

dents indeed view the respective brands as high on functional

or prestige. Future studies might directly measure perceptions

of fit and assess their influence.

A: Change in Parent Image Based on VLE Type

.52

−.00

−.07

−.50

.43 .43

−1.0

−.8

−.6

−.4

−.2

0

.2

.4

.6

.8

1.0

Upward Downward

U.S. India Analytic Holistic

.12

.09

B: Change in Parent Brand Image Based on BC

−1.0

−.8

−.6

−.4

−.2

0

.2

.4

.6

.8

1.0

Prestige Functional

U.S. India Analytic Holistic

.78

.70

.17 .22

.08

−.60

−.15

−.18

Figure 4. Study 1: Comparison of results using respondent nationality.

Allman et al. 15

Appendix A. Summary of Descriptive Items (Keywords) from the VLE Scenarios for the Two Brands: Hyundai Versus BMW (U.S. Questionnairea).

BMW Hyundai

On average, BMW cars normally sell for $51,600. BMW has started to make and sell a new midsize four door sedan as described below. Please read the new car description carefully before answering the brand evaluation questions.

On average, Hyundai cars normally sell for $26,100. Hyundai has started to make and sell a new midsize four door sedan as described below. Please read the new car description carefully before answering the brand evaluation questions.

VLE Type VLE Type

Downward Upward Downward Upward

New product name: BMW 228e

New product name: BMW 688i

New product name: Hyundai ASTER

New product name: Hyundai EXELCIUS

Study New product description: Study New product description: Study New product description: Study New product description: 1, 2 Well-made economical

midsize sedan 1 Exclusive driving

experience 1, 2 Well-made economical

midsize sedan 1, 2 Meticulously crafted,

world-class, premium midsize sedan

1, 2 Fuel efficient engine delivers comfortable drive in the city or on a highway

1, 2 Meticulously crafted, world-class, premium midsize sedan

1, 2 Fuel efficient engine delivers a comfortable drive

1, 2 4L Gasoline Direct Turbo Injection engine

1, 2 Four speed automatic transmission

1, 2 4L Gasoline Direct Turbo Injection engine

1, 2 Four speed automatic transmission

1, 2 Dual exhaust and transmission traction control

1, 2 Front/rear disc brakes 1, 2 Dual exhaust & transmission traction control

1, 2 Front and rear disc brakes 1, 2 Exclusive driving experience

1, 2 Height adjustable steering column

1, 2 More total interior volume than competitors

1, 2 Height adjustable steering column

1, 2 More total interior volume than competitors

1, 2 Adjustable driver seat with lumbar support

1, 2 Sleek interior design 1, 2 Adjustable driver seat with lumbar support

1, 2 Sleek interior design

1, 2 Heated front seats 1, 2 Hand sewed soft leather seats

1, 2 Heated front seats 1, 2 Hand sewed soft leather seats

1, 2 Steering wheel mounted audio & cruise controls

1, 2 Choice of couture colors 1, 2 Steering wheel mounted audio and cruise controls

1, 2 Choice of couture colors

1, 2 One touch auto-up driver’s window and side mirrors with turn-signal indicators

1, 2 1,2 Fluid lines and sophisticated style

1, 2 One touch auto-up driver’s window and side mirrors with turn-signal indicators

1, 2 Fluid lines and sophisticated style

2 Fabric seats in choice of grey or beige colors

1, 2 16-inch alloy wheels 2 Standard design 1, 2 16-inch alloy wheels

2 Standard design 1, 2 New panoramic sunroof 2 Standard wheels 1, 2 New panoramic sunroof 2 Standard wheels 1, 2 Backup camera 2 Fabric seats in your choice of

grey or beige colors 1, 2 Backup camera

2 Optional sunroof 1, 2 New Blind Spot Technology

2 Optional sunroof 1, 2 New Blind Spot Technology

2 Modern headlights 1, 2 New Driver Selectable Steering Mode (DSSM): normal, control, sport

2 Modern headlights Side mirrors with visual turn- signal indicators

1, 2 New Driver Selectable Steering Mode (DSSM): normal, control, sport

2 Side mirrors with visual turn- signal indicators

1, 2 Luxurious heated steering wheel

2 Back up camera with visual signal warning

1, 2 Luxurious heated steering wheel

2 Back up camera with visual signal warning

1, 2 Heated and cooled seats 2 New Blind Spot Technology with visual signal warning

1, 2 Heated and cooled seats

2 New Blind Spot Technology with visual signal warning

1, 2 High intensity headlights 2 USB port 1, 2 High intensity headlights

2 USB port 1, 2 Dual USB ports 2 Four speakers 1, 2 Dual USB ports and state of the art audio system

(continued)

16 Journal of International Marketing 27(2)

Appendix A. (continued)

VLE Type VLE Type

Downward Upward Downward Upward

2 Four speakers 1, 2 State of the art audio system with 8 speakers & sub- woofer

2 Touch command tablet 1, 2 Eight speakers and a sub- woofer

2 Touch command tablet 2 One touch auto-up driver’s control of all four windows

2 One touch auto-up driver’s control of all four windows

2 Side mirrors with turn- signal indicators

2 Side mirrors with turn- signal indicators

2 Touch & voice command tablet with gesture recognition

2 Touch & voice command tablet with gesture recognition

Comfortable Value Drive Price: $30,990 Made in Germany/South Korea

Exhilarating Luxury Drive Price: $76,210 Made in Germany/South

Korea

Comfortable Value Drive Price: $15,990 Made in Germany/South Korea

Exhilarating Luxury Drive Price: $36,190 Made in Germany/South

Korea

aThe Indian questionnaire in Study 1 contained the same eight scenarios with identical information except the average brand prices and the prices of the new VLEs, both of which were listed in the local currency (lakh). See Table 2 for all price points.

Appendix B. Sample Characteristics for Both Studies.

Study Location N Sex Age (Years)

1 United States 420 Male ¼ 210 Female ¼ 210

Range ¼ 19–81 Mean ¼ 50 Median ¼ 51 Mode ¼ 55

10th percentile ¼ 32 90th percentile ¼ 66

1 India 420 Male ¼ 210 Female ¼ 210

Range ¼ 18–75 Mean ¼ 45 Median ¼ 46 Mode ¼ 54

10th percentile ¼ 29 90th percentile ¼ 60

2 United States 357 Male ¼ 217 Female ¼ 140

Range ¼ 20–71 Mean ¼ 36 Median ¼33 Mode ¼ 29

10th percentile ¼ 25 90th percentile ¼ 52

Appendix C. Scale Items for HA Used in Both Studies.

Factor/Dimension Study Factor/Dimension Study

Causality Attitude Toward Contradictions

� Everything in the universe is somehow related to each other.

1, 2 � We should avoid going to extremes and embrace everyone’s opinions.

2

� Any phenomenon has numerous numbers of causes, although some of the causes are not known.

1, 2 � Choosing a middle ground in an argument should be avoided. (r) 2

� Everything in the world is interconnected in a causal relationship.

1, 2 � It is desirable to be in harmony, rather than in discord, with others of different opinions than one’s own.

2

(continued)

Allman et al. 17

Associate Editor

Bulent Menguc served as associate editor for this article.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to

the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for

the research, authorship, and/or publication of this article: This project

was supported by a grant from the University of West Florida through

the Office of Research and Sponsored Programs and by the research

support funds from the College of Business at the University of West

Florida.

References

Aaker, David A., and Kevin Lane Keller (1990), “Consumer Evalua-

tions of Brand Extensions,” Journal of Marketing, 54 (1), 27–41.

Agarwal, James, Naresh K. Malhotra, and Ruth N. Bolton (2010), “A

Cross-National and Cross-Cultural Approach to Global Market

Segmentation: An Application Using Consumers’ Perceived Ser-

vice Quality,” Journal of International Marketing, 18 (3), 18–40.

Ashraf, Abdul R., Narongsak (Tek) Thongpapanl, and Seigyoung Auh

(2014), “The Application of the Technology Acceptance Model

Under Different Cultural Contexts: The Case of Online Shopping

Adoption,” Journal of International Marketing, 22 (3), 68–93.

Badrinarayanan, Vishag, Enrique Becerra, Chung-Hyun Kim, and

Sreedhar Madhavaram (2012), “Transference and Congruence

Effects on Purchase Intentions in Online Stores of Multi-Channel

Retailers: Initial Evidence from the U.S. and South Korea,” Jour-

nal of the Academy of Marketing Science, 40 (4), 539–57.

Balachander, Subramanian, and Sanjoy Ghose (2003), “Reciprocal

Spillover Effects: A Strategic Benefit of Brand Extensions,” Jour-

nal of Marketing, 67 (1), 4–13.

Bartikowski, Boris, and Mark Cleveland (2017), “‘Seeing Is Being’:

Consumer Culture and the Positioning of Premium Cars in China,”

Journal of Business Research, 77 (August), 195–202.

Bolton, Ruth N., and Matthew B. Myers (2003), “Price-Based Global

Market Segmentation for Services,” Journal of Marketing, 67 (3),

108–28.

Briley, Donnel A., Michael Morris, and Itamar Simonson (2000),

Reasons as Carriers of Culture: Dynamic vs. Dispositional Models

of Cultural Influence on Decision Making. Stanford, CA: Stanford

Business School.

Briley, Donnel A., Michael Morris, and Itamar Simonson (2005),

“Cultural Chameleons: Biculturals, Conformity Motives, and Deci-

sion Making,” Journal of Consumer Psychology, 15 (4), 351–62.

Broderick, Amanda J., Gordon E, Greenley, and Rene Dentiste

Mueller (2007), “The Behavioural Homogeneity Evaluation

Framework: Multi-Level Evaluations of Consumer Involvement

in International Segmentation,” Journal of International Business

Studies, 38 (5), 746–63.

Buil, Isabel, Leslie de Chernatony, and Leif E Hem (2009), “Brand

Extension Strategies: Perceived Fit, Brand Type, and Culture

Appendix C. (continued)

Factor/Dimension Study Factor/Dimension Study

� Nothing is unrelated. 2 � It is more desirable to take the middle ground than go to extremes.

2

� Even a small change in any element of the universe can lead to significant alterations in other elements.

2 � It is more important to find a point of compromise than to debate who is right/wrong, when one’s opinions conflict with other’s opinions.

2

� Any phenomenon entails a numerous number of consequences, although some of them may not be known.

2 � When disagreement exists among people, they should search for ways to compromise.

2

Locus of Attention Perception of Change

� It is more important to pay attention to the whole context rather than the details.

1, 2 � Every phenomenon in the world moves in predictable directions. (r)

2

� It is more important to pay attention to the whole than its parts.

1, 2 � A person who is currently living a successful life will continue to stay successful. (r)

2

� The whole rather than its parts, should be considered to understand a phenomenon.

1, 2 � An individual who is currently honest will stay honest in the future. (r)

2

� The whole is greater than the sum of its parts. 2 � If an event is moving toward a certain direction, it will continue to move toward that direction. (r)

2

� It is not possible to understand the parts without considering the whole picture.

2 � Current situations can change at any time. 2

� We should consider the situation a person is faced with, as well as his/her personality, to understand one’s behavior.

2 � Future events are predictable based on present situations. (r) 2

Notes: The scores of the reverse items were reversed for analysis. (r) ¼ Reverse-coded item.

18 Journal of International Marketing 27(2)

Influences,” European Journal of Marketing, 43 (11/12),

1300–1324.

Chao, Paul (2001), “The Moderating Effects of Country of Assembly,

Country of Parts, and Country of Design on Hybrid Product

Evaluations,” Journal of Advertising, 30 (4), 67–81.

Chen, Ming-Jer (2002), “Transcending Paradox: The Chinese ‘Middle

Way’ Perspective,” Asia Pacific Journal of Management, 19 (2/3),

179–99.

Choi, Incheol, Minkyung Koo, and Jong An Choi (2007), “Individual

Differences in Analytic Versus Holistic Thinking,” Personality

and Social Psychology Bulletin, 33 (5), 691–705.

Czellar, Sandor (2003), “Consumer Attitude Toward Brand Exten-

sions: An Integrative Model and Research Propositions,” Interna-

tional Journal of Research in Marketing, 20 (1), 97–115.

Dall’Olmo Riley, Francesca, José M. Pina, and Rafael Bravo (2013),

“Downscale Extensions: Consumer Evaluation and Feedback

Effects,” Journal of Business Research, 66 (2), 196–206.

Diamantopoulos, Adamantios, Bodo Schlegelmilch, and Dayananda

Palihawadana (2011), “The Relationship Between Country-of-

Origin Image and Brand Image as Drivers of Purchase Intentions,”

International Marketing Review, 28 (5), 508–24.

Elliott, Gregory R., and Ross C. Cameron (1994), “Consumer Percep-

tion of Product Quality and the Country-of-Origin Effect,” Journal

of International Marketing, 2 (2), 49–62.

Fetscherin, Marc, and Mark Toncar (2010), “The Effects of the Coun-

try of Brand and the Country of Manufacturing of Automobiles: An

Experimental Study of Consumers’ Brand Personality Perceptions.”

International Marketing Review, 27 (2), 164–78.

Fiske, Susan T., and Shelley E. Taylor (1991), Social Cognition, 2nd

ed. New York: McGraw-Hill.

Gao, Paul, Russell Hensley, and Andres Zielke (2014), “A Road Map

to the Future for the Auto Industry,” McKinsey Quarterly (Octo-

ber), http://www.mckinsey.com/industries/automotive-and-assem

bly/our-insights/a-road-map-to-the-future-for-the-auto-industry.

Grewal, Rajdeep, Alok Kumar, Girish Mallapragada, and Amit Saini

(2013), “Marketing Channels in Foreign Markets: Control

Mechanisms and the Moderating Role of Multinational Corpora-

tion Headquarters–Subsidiary Relationship,” Journal of Marketing

Research, 50 (3), 378–98.

Gürhan-Canli, Zeynep, and Durairaj Maheswaran (1998), “The

Effects of Extensions on Brand Name Dilution and Enhancement,”

Journal of Marketing Research, 35 (4), 464–73.

Gürhan-Canli, Zeynep, Gülen Sarial-Abi, and Ceren Hayran (2018),

“Consumers and Brands Across the Globe: Research Synthesis and

New Directions,” Journal of International Marketing, 26 (1),

96–117.

Hall, Edward T., and Mildred Reed Hall (1990), Understanding Cul-

tural Differences. Yarmouth, UK: Intercultural Press.

Han, C. Min, and William J. Qualls (1985), “Country of Origin Effects

and Its Impact upon Consumers’ Perception of Quality,” in His-

toric Perspective in Consumer Research: National and Interna-

tional Perspectives, Chin Tiong. Tan, and Jagdish Sheth, eds.

New York: Association for Consumer Research.

Häubl, Gerald (1996), “A Cross-National Investigation of the Effects

of Country of Origin and Brand Name on the Evaluation of a New

Car,” International Marketing Review, 13 (5), 76–97.

Heath, Timothy B., Devon DelVecchio, and Michael S. McCarthy

(2011), “The Asymmetric Effects of Extending Brands to Lower

and Higher Quality,” Journal of Marketing, 75 (4), 3–20.

Hofstede, G. (1984), Culture’s Consequences: International Differ-

ences in Work-related Values. Newbury Park, CA: SAGE

Publications.

Hong, Jiewen, and Hannah H. Chang, (2015), “‘I’ Follow My Heart

and ‘We’ Rely on Reasons: The Impact of Self-Construal on Reli-

ance on Feelings Versus Reasons in Decision Making,” Journal of

Consumer Research, 41 (6), 1392–1411.

Hsieh, Ming-Huei, Shan-Ling Pan, and Rudy Setiono (2004),

“Product-, Corporate-, and Country-Image Dimensions and Pur-

chase Behavior: A Multicountry Analysis,” Journal of the Acad-

emy of Marketing Science, 32 (3), 251–70.

Inglehart, Robert, and Wayne E. Baker (2000), “Modernization, Cul-

tural Change and the Persistence of Traditional Values,” American

Sociological Review, 65 (1), 19–51.

Insch, Gary S., and J. Brad McBride (2004), “The Impact of Country-

of-Origin Cues on Consumer Perceptions of Product Quality: A

Binational Test of the Decomposed Country-of-Origin Construct,”

Journal of Business Research, 57 (3), 256–65.

Isidore, Chris (2014), “South Carolina Plant to Become BMW’s

Largest,” CNNMoney, (March 28), http://money.cnn.com/2014/

03/28/news/companies/bmw-plant/.

Jaffe, Eugene D., and Israel D. Nebenzahl (2001), National Image and

Competitive Advantage. Copenhagen: Copenhagen Business

School Press.

Janiszewski, Chris, and Stijn M.J. Van Osselaer (2000), “A Connec-

tionist Model of Brand-Quality Associations,” Journal of Market-

ing Research, 37 (3), 331–50.

Ji, Li-Jun, Kaiping Peng, and Richard E. Nisbett (2000), “Culture,

Control, and Perception of Relationships in the Environment,”

Journal of Personality and Social Psychology, 78 (May), 943–55.

Keller, Kevin Lane (2003), “Brand Synthesis: The Multidimension-

ality of Brand Knowledge,” Journal of Consumer Research, 29 (4),

595–600.

Keller, Kevin Lane, and David A. Aaker (1992), “The Effects of

Sequential Introduction of Brand Extensions,” Journal of Market-

ing Research, 29 (1), 35–60.

Kim, Chung K., and Anne M. Lavack (1996), “Vertical Brand Exten-

sions: Current Research and Managerial Implications,” Journal of

Product & Brand Management, 5 (6), 24–37.

Kirmani, Amna, Sanjay Sood, and Sheri Bridges (1999), “The Own-

ership Effect in Consumer Responses to Brand Line Stretches,”

Journal of Marketing, 63 (1), 88–101.

Kuhnen, 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 Personality and

Social Psychology, 80 (3), 397–409.

Lalwani, Ashok K., and Sharon Shavitt (2013), “You Get What You

Pay For? Self-Construal Influences Price-Quality Judgments,”

Journal of Consumer Research, 40 (2), 255–67.

Lee, Angela Y., and Aparna A. Labroo (2004), “The Effect of Con-

ceptual and Perceptual Fluency on Brand Evaluation,” Journal of

Marketing Research, 41 (2), 151–65.

Allman et al. 19

Lee, Chang Han, Eunju Ko, Henrikki Tikkanen, Michel Cao Tuan

Phan, Gaetano Aiello, and Raffaele Donvito, et al. (2014),

“Marketing Mix and Customer Equity of SPA Brands: Cross-

Cultural Perspectives,” Journal of Business Research, 67 (10),

2155–63.

Lei, Jing, de Ruyter Ko, and Martin Wetzels (2008), “Consumer

Responses to Vertical Service Line Extensions,” Journal of Retail-

ing, 84 (3), 268–80.

Loewenstein, George F., Elke U. Weber, Christopher K. Hsee, and

Ned Welch (2001), “Risk as Feelings,” Psychological Bulletin, 127

(2), 267–86.

Loken, Barbara, and Nelson Amaral (2010), “Brand Dilution: The

Impact of the User of Counterfeits on Original Brand Perception,”

in Advances in Consumer Research, Vol. 37, Margaret C.

Campbell, Jeff Inman and Rik Pieters, eds. Duluth, MN: Associa-

tion for Consumer Research, 859–60.

Loken, Barbara, and Deborah Roedder John (1993), “Diluting Brand

Beliefs: When Do Brand Extensions Have a Negative Impact?”

Journal of Marketing, 57 (3), 71–84.

Magnusson, Peter, Robert Peterson, and Stanford A. Westjohn (2014),

“The Influence of National Cultural Values on the Use of Rewards

Alignment to Improve Sales Collaboration,” International Market-

ing Review, 31 (1), 30–50.

Magnusson, Peter, Stanford A. Westjohn, and Srdan Zdravkovic

(2011), “‘What? I Thought Samsung Was Japanese’: Accurate or

Not, Perceived Country of Origin Matters,” International Market-

ing Review, 28 (5), 454–72.

Mandler, George (1982), The Structure of Value: Accounting for

Taste. Center for Human Information Processing Report, Issue

101. San Diego: University of California.

Masuda, Takahiko, and Richard E. Nisbett (2001), “Attending Holi-

stically Versus Analytically: Comparing the Context Sensitivity of

Japanese and Americans,” Journal of Personality and Social

Psychology, 81 (5), 922–34.

Milberg, Sandra J., C. Whan Park, and Michael S. McCarthy (1997),

“Managing Negative Feedback Effects Associated With Brand

Extensions: The Impact of Alternative Branding Strategies,” Jour-

nal of Consumer Psychology, 6 (2), 119–40.

Miyamoto, Yuri, Richard E. Nisbett, and Takahiko Masuda (2006),

“Culture and the Physical Environment,” Psychological Science

(0956-7976), 17 (2), 113–19.

Monga, Alokparna Basu, and Deborah Roedder John (2007), “Cultural

Differences in Brand Extension Evaluation: The Influence of

Analytic Versus Holistic Thinking,” Journal of Consumer

Research, 33 (4), 529–36.

Monga, Alokparna Basu, and Deborah Roedder John (2010), “What

Makes Brands Elastic? The Influence of Brand Concept and Styles

of Thinking on Brand Extension Evaluation,” Journal of Market-

ing, 74 (3), 80–92.

Ng, Sharon (2010), “Cultural Orientation and Brand Dilution: Impact

of Motivation Level and Extension Typicality,” Journal of

Marketing Research, 47 (1), 186–98.

Nisbett, Richard E. (2003), The Geography of Thought: How Asians

and Westerners Think Differently, & Why. New York: The Free

Press.

Papadopoulos, Nicolas, Alia el Banna, Steven A. Murphy, and José I.

Rojas-Mendez (2011), “Place Brands and Brand-Place Associa-

tions: The Role of ‘Place’ in International Marketing,” in Hand-

book of Research in International Marketing, 2nd ed., Subhash C.

Jain and David A. Griffith, eds. Cheltenham, UK: Edward Elgar

Publishing, 88–116.

Pappu, Ravi, Pascale G. Quester, and Ray W. Cooksey (2006),

“Consumer-Based Brand Equity and Country-of-Origin

Relationships,” European Journal of Marketing, 40 (5/6),

696–717.

Pappu, Ravi, Pascale G. Quester, and Ray W. Cooksey (2007),

“Country Image and Consumer-Based Brand Equity: Relationships

and Implications for International Marketing,” Journal of Interna-

tional Business Studies, 38 (5), 726–45.

Park, C. Whan, Bernard J. Jaworski, and Deborah J. Maclnnis (1986),

“Strategic Brand Concept-Image Management,” Journal of Mar-

keting, 50 (4), 135–45.

Park, C. Whan, Sandra Milberg, and Robert Lawson (1991),

“Evaluation of Brand Extensions: The Role of Product Feature

Similarity and Brand Concept Consistency,” Journal of Consumer

Research, 18 (2), 185–93.

Peng, Kaiping, and Richard E. Nisbett (1999), “Culture, Dialectics,

and Reasoning About Contradiction,” American Psychologist, 54

(9), 741–54.

Pina, José M., Nina M. Iversen, and Eva Martinez (2010), “Feedback

Effects of Brand Extensions on the Brand Image of Global Brands:

A Comparison Between Spain and Norway,” Journal of Marketing

Management, 26 (9/10), 943–66.

Ralston, David A., David H. Holt, Robert H. Terpstra, and Yu Kai-

Cheng (2008), “The Impact of National Culture and Economic

Ideology on Managerial Work Values: A Study of the United

States, Russia, Japan, and China,” Journal of International Busi-

ness Studies, 39 (1), 8–26.

Reber, Rolf, Norbert Schwarz, and Piotr Winkielman (2004),

“Processing Fluency and Aesthetic Pleasure: Is Beauty in the Per-

ceiver’s Processing Experience?” Personality & Social Psychol-

ogy Review, 8 (4), 364–82.

Roth, Martin S., and Jean B. Romeo (1992), “Matching Product Cate-

gory and Country Image Perceptions: A Framework for Managing

Country-of-Origin Effects,” Journal of International Business

Studies, 23 (3), 477–97.

Salomon, Robert (2016), “Here’s Why Walmart Stumbled on The

Road to China,” Fortune (February 21), http://fortune.com/2016/

02/21/why-walmart-stumbled-on-road-to-china/.

Samuel, Alexandra (2016), “Psychographics Are Just as Important for

Marketers as Demographics,” Harvard Business Review (March

11), https://hbr.org/2016/03/psychographics-are-just-as-impor

tant-for-marketers-as-demographics.

Schwartz, Shalom H. (1994), “Are There Universal Aspects in the

Structure and Contents of Human Values?” Journal of Social

Issues, 50 (4), 19–45.

Schwartz, Shalom H. (1999), “A Theory of Cultural Values and Some

Implications for Work,” Applied Psychology: An International

Review, 48 (1), 23–47.

Scott, Mike (2018), “5G Technology Set to Fuel Technological

Change and Transform Global Economy,” Forbes (October 29),

20 Journal of International Marketing 27(2)

https://www.forbes.com/sites/mikescott/2018/10/29/5g-technol

ogy-set-to-fuel-technological-change-and-transform-global-econ

omy/#71d4e4362f51.

Shepherd, Steven, Tanya L. Chartrand, and Gavan J. Fitzsimons

(2015), “When Brands Reflect Our Ideal World: The Values and

Brand Preferences of Consumers Who Support versus Reject

Society’s Dominant Ideology,” Journal of Consumer Research,

42 (1), 76–92.

Sichtmann, Christina, and Adamantios Diamontopoulos (2013),

“The Impact of Perceived Brand Globalness, Brand Origin

Image, and Brand Origin–Extension Fit on Brand Extension

Success,” Journal of the Academy of Marketing Science, 41 (5),

567–85.

Sood, Sanjay, and Kevin Lane Keller (2012), “The Effects of

Brand Name Structure on Brand Extension Evaluations and

Parent Brand Dilution,” Journal of Marketing Research, 49

(3), 373–82.

Steenkamp, Jan-Benedict E.M., and Hans Baumgartner (1998),

“Assessing Measurement Invariance in Cross-National Consumer

Research,” Journal of Consumer Research, 25 (1), 78–90.

Sujan, Mita, and James R. Bettman (1989), “The Effects of Brand

Positioning Strategies on Consumers’ Brand and Category Percep-

tions: Some Insights from Schema Research,” Journal of Market-

ing Research, 26 (4), 454–67.

Talay, M. Berk, Janell D. Townsend, and Sengun Yeniyurt (2015),

“Global Brand Architecture Position and Market-Based Perfor-

mance: The Moderating Role of Culture,” Journal of International

Marketing, 23 (2), 55–72.

Ter Hofstede, Frenkel, Jan-Benedict E.M. Steenkamp, and Michel

Wedel (1999), “International Market Segmentation Based on

Consumer-Product Relations,” Journal of Marketing Research,

36 (1), 1–17.

Till, Brian D., and Randi Lynn Priluck (2000), “Stimulus General-

ization in Classical Conditioning: An Initial Investigation and

Extension,” Psychology and Marketing, 17 (1), 55–64.

Torelli, Carlos J., Ayşegül Özsomer, Sergio W. Carvalho, Hean Tat

Keh, and Natalia Maehle (2012), “Brand Concepts as Representa-

tions of Human Values: Do Cultural Congruity and Compatibility

Between Values Matter?” Journal of Marketing, 76 (4), 92–108.

Triandis, Harry C. (1995), Individualism & Collectivism. Boulder,

CO: Westview Press.

Van der Lans, Ralf, Yvonne van Everdingen, and Valentyn Melnyk

(2016), “What to Stress, to Whom and Where? A Cross-Country

Investigation of the Effects of Perceived Brand Benefits on Buying

Intentions,” International Journal of Research in Marketing, 33

(4), 924–43.

Völckner, Franziska, and Henrik Sattler (2006), “Drivers of Brand

Extension Success,” Journal of Marketing, 70 (2), 18–34.

Weber, Renee, and Jennifer Crocker (1983), “Cognitive Processes in

the Revision of Stereotypic Beliefs,” Journal of Personality and

Social Psychology, 45 (5), 961–77.

Winkielman, Piotr, Jasmin Halberstadt, Tedra Fazendeiro, and Steve

Catty (2006), “Prototypes Are Attractive Because They Are Easy

on the Mind,” Psychological Science, 17 (9), 799–806.

Yorkston, Eric A., Joseph C. Nunes, and Shashi Matta (2010), “The

Malleable Brand: The Role of Implicit Theories in Evaluating

Brand Extensions,” Journal of Marketing, 74 (1), 80–93.

Zhang, Jing, and Michelle R. Nelson (2016), “The Effects of Vertical

Individualism on Status Consumer Orientations and Behaviors,”

Psychology and Marketing, 33 (5), 318–30.

Zhu, Rui (Juliet), and Joan Meyers-Levy (2009) “The Influence of

Self-View on Context Effects: How Display Fixtures Can

Affect Product Evaluations,” Journal of Marketing Research,

46 (1), 37–45.

Allman et al. 21

Copyright of Journal of International Marketing 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.