© 2016, American Marketing Association

Journal of International Marketing

PrePrint, Unedited
All rights reserved. Cannot be reprinted without the express
permission of the American Marketing Association.

Does Being Perceived as Global Pay Off?

An Analysis of Leading Foreign and Domestic MNCs in India, Japan, and the United
States

Bernhard Swoboda*
Prof. Dr. Prof. h.c.
Trier University
Universitätsring 15
54296 Trier
Germany
Phone: 0049-651-2013050
E-Mail: b.swoboda@uni-trier.de

Johannes Hirschmann
Ph.D. student
Trier University
Universitätsring 15
54296 Trier
Germany
Phone: 0049-651-2012607
E-Mail: j.hirschmann@uni-trier.de

* corresponding author
 1

Does Being Perceived as Global Pay Off?

An Analysis of Leading Foreign and Domestic MNCs in India, Japan, and the United States

ABSTRACT

Multinational product and corporate brands are known to have the advantage over local competi-

tors of being global. However, while foreign multinational corporations (MNCs) increasingly

compete against domestic MNCs, particularly in economically strong countries, little is known

about whether foreign MNCs can use their advantage of being global and how the perceptions of

foreign vs. domestic MNCs drive consumer behavior. To examine this issue, an accessibility-di-

agnosticity theory-based framework is proposed and tested using consumer data from important

countries in which leading foreign and domestic MNCs compete. The results show that perceived

brand globalness (PBG) enhances consumers’ intentional loyalty toward MNCs indirectly by af-

fecting the functional and psychological value offered. The value-creation routes change accord-

ing to firms’ origins, i.e., foreign MNCs translate globalness through functional and psychological

value, while domestic MNCs translate globalness through psychological value. However, the ben-

efits of being global and the value-creation routes for foreign (vs. domestic) MNCs differ between

countries. Finally, consumers’ stronger ethnocentrism weakens—but does not completely erase—

the effects of PBG in all countries and additionally interacts with MNCs’ origins.

Keywords: corporate brands; multinational corporations; emerging giants; perceived brand glob-

alness; offered value; foreignness; ethnocentrism

 2

Leading MNCs, such as The Coca-Cola Company, communicate their globalness to differentiate

themselves from competitors (Cioletti 2013). Globalness is perceived by consumers and stored in

their minds, and it is thus defined as PBG, i.e., the extent to which a product or a corporation is

viewed as a global player (e.g., Steenkamp, Batra, and Alden 2003). In this study, MNCs’ corpo-

rate brands are analyzed. PBG acts as a general impression that provides consumers with access

to MNCs’ offered value (e.g., Aaker 2004), thus allowing them to evaluate an MNC without nec-

essarily having had prior personal experience with it. However, especially in economically

strong and attractive countries, foreign MNCs compete against domestic MNCs, e.g., emerging

giants in emerging economies (e.g., Kumar and Steenkamp 2013), and consumers’ ethnocentric

tendencies particularly affect MNCs (e.g., Sharma 2011). For example, the CEO of the Chinese

Haier Electronics Group Co. realized the importance of PBG and promoted a strong global cor-

porate brand (Balmer 2012). Hence, questions about whether and how foreign MNCs and do-

mestic MNCs can benefit from PBG in such environments arise.

Scholars often study global brands by focusing on consumers’ attitudes or global con-

sumer orientation and—less often but notably—PBG (see Table 1). Scholars of the latter hypothe-

size the direct effects of PBG (e.g., Steenkamp, Batra, and Alden 2003; Winit et al. 2014), show-

ing mostly affective responses, such as brand esteem, global myth and prestige (e.g., Johansson

and Ronkainen 2005), as well as quality and, thus functional responses as pathways that translate

PBG into purchase intentions or preferences (e.g., Akram, Merunka, and Akram 2011; Holt,

Quelch, and Taylor 2004; Xie, Batra, and Peng 2015).

However, the empirical results are restricted to product brands and limit our understanding

of the effects of corporate brands. The latter were analyzed by Swoboda, Pennemann, and Taube

(2012), who study PBG’s effects on patronage behavior for retail chains/corporations, and by

Alden et al. (2013), who study the antecedents of global brand attitudes. In contrast, corporate

brands are important because, in contrast to single products, they are a CEO’s task (e.g., Lafley

2009; Mascarenhas 2009) and are of strategic importance to product brands (in various branding
 3

strategies, Berens, van Riel, and van Bruggen 2005). CEOs may rely on corporate brand value

when deciding on market entry, product introductions, or value transfers to products (even if a

product perspective addresses direct competition in greater detail). Moreover, only three studies

address PBG pathways to success by jointly comparing foreign global vs. domestic-/local-only

brands (Swoboda, Pennemann, and Taube 2012 in China and for products Zhou, Yang, and Hui

2010 in China; Winit et al. 2014 in Thailand), and two studies compare foreign global vs. domestic

global product brands (Riefler 2012 in Austria; Winit et al. 2014).1

As those studies are restricted to one, often emerging, country, we conclude that scant re-

search exists on corporate brands in different environments, and the ways in which MNCs’ value

pathway for PBG can be translated into customer behavior remain unclear. Finally, our

knowledge of the boundaries of corporate brands’ effects is limited. Because PBG is distinct from

foreignness and because origin information evokes functional and psychological value to varying

degrees (e.g., Zhou, Yang, and Hui 2010), MNCs’ foreign vs. domestic origins may constitute an

important boundary condition, particularly in a competitive environment for foreign and domestic

MNCs. Because consumer ethnocentrism affects the PBG effects of product brands (for

lower/higher ethnocentric groups on the PBG-value links, Akram, Merunka, and Akram 2011,

and PBG-purchase likelihood, Steenkamp, Batra, and Alden 2003) and scholars find inconclusive

effects of ethnocentrism on brand attitudes and call for clarification (e.g., Guo 2013, among 40

studies), how more (vs. less) ethnocentric consumers react to the PBG of those MNCs’ corporate

brands is unclear.

Table 1

We aim to advance the literature by analyzing the underlying pathways through which MNCs trans-

form PBG into consumer intentional loyalty in different countries (as consumers in emerging vs.

developed countries are likely to respond differently, e.g., Coulter, Price, and Feick 2003). We

also ask whether and how MNCs’ foreign vs. domestic origins change the total effect of PBG and the

value pathways by which PBG translates into loyalty and how consumer ethnocentrism interacts
 4

with PBG in different countries.

We offer valuable contributions to the extant literature by extending the knowledge on

MNCs’ PBG effects, which is important because a strong corporate brand and similar corporate

reputation are internationally advantageous for MNCs (e.g., when attracting local customers or em-

ployees, Swoboda, Puchert, and Morschett 2015). First, an MNC’s corporate brand needs to have

and give its own value proposition because it is otherwise vulnerable, and it works best when

providing functional and emotional value (e.g., Aaker 2004; Balmer 2010, 2012). Therefore, we

focus on PBG’s pathways to consumer loyalty through functional and psychological value. As a

result, we advance the findings of Swoboda, Pennemann, and Taube (2012), who have extended our

knowledge on these pathways for foreign vs. local-only retail brands in emerging countries and have

called for further analysis of MNCs (also Özsomer 2012).

Second, by treating the PBG of foreign vs. domestic manufacturing MNCs as an independ-

ent predictor that explains PBG-loyalty pathways, we account for the competition between MNCs

(i.e., their global corporate brands, which are subsequently used synonymously) with different

origins. Consumer ethnocentrism may be an important boundary condition of MNCs, i.e., for

CEOs, the total and indirect effects of the PBG of an entire MNC brand within lower (vs. higher)

ethnocentric consumer groups should be of paramount interest. We provide new insights into

such comparative pathways as well as insights into the moderating effect of ethnocentrism as a

continuous variable (e.g., Guo 2013) and the three-way interaction because MNCs’ foreign ori-

gins may strengthen PBG effects, which may be further weakened by ethnocentrism.

Third, we equally develop the applicability of accessibility-diagnosticity theoretical reason-

ing (Feldman and Lynch 1988) in global brand research (e.g., Zhou, Yang, and Hui 2010;

Swoboda, Pennemann, and Taube 2012) to our main and moderation hypotheses (PBG studies of-

ten combine or do not apply theories; see Table 1). Analyses in India, Japan, and the U.S. and alter-

native tests in China, Italy, and in all countries with differing MNC combinations provide further
 5

rigor, stability and generalizability to the results. We chose these countries because they are eco-

nomically strong (belonging to the top ten markets in terms of GDP, United Nations 2014) and

because they are likely to host competing domestic and foreign MNCs.2 From a traditional, Western

MNC perspective, these countries differ with respect to economic strength and culture (Alden et al.

2013). Practically speaking, foreign MNCs may wish to consider the role of PBG in local compe-

tition with domestic MNCs, especially in these important countries.

HYPOTHESIS DEVELOPMENT

Framework and theory

To address our research aims, we propose a conceptual framework (see Figure 1) in which the PBG

of MNCs determines consumer loyalty through functional and psychological value (referring to and

extending frameworks by Steenkamp, Batra, and Alden 2003; Swoboda, Pennemann, and Taube

2012). We focus on intentional loyalty toward an MNC (consumers’ intention and readiness to

buy offers and to establish a good relationship with a firm, as it is a core predictor of consumer

spending; Morgeson III et al. 2011; Oliver 1999) and on offered functional value (quality/value

for money) and psychological value (emotional/social) for three reasons: they seem to be the most

universal among the value concepts; they comprise more cognitive and more affective behavioral

elements; and they are diagnostic in that they determine consumer purchase behavior, i.e., the

PBG of MNCs operates through both types of value (e.g., Alden et al. 2013; Swoboda, Penne-

mann, and Taube 2012). Moreover, we argue that the PBG-loyalty relationship is moderated by

MNCs’ objective origins, as consumers’ responses to PBG may differ according to MNCs’ for-

eign vs. domestic origins, and by ethnocentrism, as consumers’ beliefs about the appropriateness

and, indeed, the morality of purchasing foreign-made offers may alter MNCs’ PBG pathways

(e.g., Shimp and Sharma 1987).

Figure 1
 6

To understand how PBG affects consumers’ loyalty and how this mechanism is affected by bound-

ary conditions, we use Feldman and Lynch’s (1988) accessibility-diagnosticity framework (for

other theories, see Table 1). This theory provides an integrated basis by suggesting that the likeli-

hood that an individual uses information about an object for decision making about that object de-

pends on the information’s accessibility and diagnosticity. Accessibility refers to the ease of retriev-

ing specific information from memory, i.e., the activation potential of available knowledge (Higgins

1996), while diagnosticity refers to the extent to which the inferences based on this information are

adequate to make a decision, i.e., the “degree to which the use of each type of information allows

consumers to accomplish their objectives in the particular decision task at hand” (Lynch,

Marmorstein, and Weigold 1988, p. 171).

In our context, the probability that the PBG associations of an MNC can be used as in-

formation to evaluate consumers’ intentional loyalty is a function of PBG’s accessibility and di-

agnosticity. PBG, i.e., the extent to which a corporation is viewed as global (e.g., Steenkamp,

Batra, and Alden 2003), is likely to be a salient attribute in consumers’ memory due to most

global corporate brands’ high availability in international markets and their resulting high visi-

bility. PBG thus represents easily accessible information (e.g., Swoboda, Pennemann, and

Taube 2012). This information becomes diagnostic, i.e., relevant for consumers’ loyalty behav-

ior, when it influences their perceptions of offered value because consumers base their loyalty

behaviors on the value offered by a corporation (e.g., Walker and Olson 1991) and because the

value associated with a global brand makes the globalness information itself useful. Hence, con-

sumers link MNCs’ PBG with functional and psychological value to make PBG a piece of diag-

nostic information regarding loyalty behavior toward an MNC. However, this relationship

might change because of boundary conditions. MNC origin acts as a diagnosticity multiplier,

i.e., an additional piece of accessible information that is used for MNC evaluation, because it

enables the use of specific information that is stored in consumers’ minds. Consumers’ ethno-
 7

centrism, i.e., personal beliefs that affect the likelihood of the use of PBG, acts as belief-con-

sistent information. PBG’s diagnosticity is primarily affected by two mechanisms: strength-

ened/weakened links of PBG with offered value and value pathways from PBG to loyalty.

Subsequently, the accessibility-diagnosticity theory and empirical studies on product

brands are used to hypothesize the main effect of PBG and the moderating effects.

Main effects of PBG

PBG directly affects intentional behavior only under certain conditions. Particularly highly affec-

tive associations are accessible and may become directly diagnostic (Verplanken, Hofstee, and

Janssen 1998). Additionally, consumers who appreciate globalness because it grants them mem-

bership in a worldwide segment consider global brands a “passport to global citizenship”

(Strizhakova, Coulter, and Price 2011, p. 342), which creates added value and may render PBG

diagnostic for loyalty. Based on such rationales, Steenkamp, Batra, and Alden (2003) hypothesize

a direct relationship between PBG and the purchase likelihood of global product brands (e.g.,

Winit et al. 2014 show it empirically), while Swoboda, Pennemann, and Taube (2012) deny the

existence of such a relationship for global (vs. domestic-only) retailer chains/corporations. We be-

lieve that although MNCs’ PBG is accessible, it is not directly relevant to behavioral intentions

toward MNCs for most consumers because, theoretically, it is not naturally diagnostic information

for loyalty decisions. Foreign or domestic MNCs’ PBG will seldom directly affect loyal behavior,

even if this assumption requires empirical evidence because recent studies show strong direct ef-

fects of MNCs’ reputation on loyalty behavior (e.g., Swoboda, Puchert, and Morschett 2015) and

because MNCs such as Procter & Gamble increasingly craft their global corporate brand with

specific value propositions to customers (Aaker 2004; Lafley 2009).

Theoretically, PBG becomes diagnostic when consumers link this information to a func-

tional and psychological value proposition of MNCs’ corporate brands because the offered value

makes a globalness impression a useful piece of information for decision making and is relevant
 8

with regard to loyalty intentions toward an MNC (following the rationale of Lynch, Marmorstein,

and Weigold 1988). Functional value represents the cognitively perceived utility derived from an

offer’s quality, performance, and reduction of perceived costs; psychological value represents the

affectively perceived utility derived from an offer’s feelings, states, and ability to enhance one’s

social self-concept (Sweeney and Soutar 2001). For global product brands, PBG has been shown

to reinforce functional value across nations because globalness stimulates cognitive perceptions of

quality or value for money (e.g., Holt, Quelch, and Taylor 2004; Schuiling and Kapferer 2004)

and psychological value by stimulating affective perceptions of emotional experience or prestige

(e.g., Alden, Steenkamp, and Batra 1999; Ger 1999). Similarly, when consumers perceive a foreign

or domestic MNC, they may regard its offers as being of higher quality, which is required for

global acceptance, and, in turn, of having greater credibility and lower risk (Keller 1998). Consum-

ers may also regard these corporate brands as superior because they perceive them to be more

scarce, expensive, sophisticated, or modern, which creates higher aspirational appeal and esteem

(e.g., Swoboda, Pennemann, and Taube 2012). As a result, MNCs’ PBG information becomes di-

agnostic and affects consumer behavior across nations.

We propose the following hypothesis for the indirect pathways of PBG to loyalty across

nations as the basis of subsequent rationales:

Hypothesis 1: The PBG of MNCs positively affects consumer loyalty (a) through functional

value and (b) through psychological value.

Moderating role of MNC origins

Next, we analyze how the total effect of PBG (the sum of the direct and indirect effects) and the

value pathways (the indirect effects via functional value and psychological value) through which

PBG translates into loyalty differ between foreign and domestic MNCs.

We suggest different total effects of PBG on consumer loyalty depending on the MNC’s

origin. Theoretically, consumers recall information on PBG from memory, and when functional
 9

value and psychological value are linked to PBG, such information may serve a diagnostic func-

tion for the creation of consumers’ loyalty intentions toward an MNC (Feldman and Lynch 1988).

Such information may become even more diagnostic depending on the MNC’s origin. Origin infor-

mation helps evaluate an MNC because its origin represents easily accessible information that

consumers can add to a list of salient MNC attributes and then use for MNC evaluation (e.g.,

Dimofte, Johansson, and Ronkainen 2008). A foreign MNC origin may function as a diagnosticity

multiplier for PBG, while a domestic MNC origin—even if a firm is global—may diminish the

diagnosticity of PBG.

Because foreignness is considered exciting and is seen to enhance social status (Eckhardt

2005), because foreignness tends to be perceived as global (e.g., Batra et al. 2000 across nations),

and because foreignness is thus likely to activate both cognitive and affective value mechanisms, a

foreign MNC origin should lead to a higher diagnosticity of PBG. However, a higher diagnosticity

of PBG does not make PBG directly useful in terms of loyalty because, theoretically, this information

alone will hardly convey sufficient direct diagnosticity to justify loyalty toward an MNC. By contrast,

domestic MNCs may be perceived as local because of their closeness to consumers or their cultural

embeddedness (e.g., Ger 1999; Özsomer 2012). Origin information may enhance local consumers’

specific feelings or even their pride (i.e., psychological vs. functional affect), but the diagnosticity

of PBG caused by origin information is likely to be weaker for domestic MNCs than for foreign

MNCs. Thus, domestic MNC origin is likely to increase the diagnosticity of PBG to a lesser extent.

Empirical studies show that the value-creation process of PBG is more relevant for for-

eign vs. local-only retail chains and corporations (Swoboda, Pennemann, and Taube 2012). We

expect to find a similar total effect for foreign vs. domestic MNCs because origin information—

which is linked to functional and psychological value—affects the diagnosticity of PBG. Con-

cerning country contexts, the mechanism has been assumed for Western retailers in an emerging

country, such as China, by arguing for a strengthened PBG-value link, that is, reinforced quality

and prestige perceptions (Swoboda, Pennemann, and Taube 2012). We might be skeptical if
 10

similar effects are also found in developed countries for foreign vs. domestic MNCs. However,

because foreignness is considered prevalent in developed countries and across different nations

and cultures (e.g., Alden, Steenkamp, and Batra 1999), we hypothesize the following across na-

tions:

Hypothesis 2: The total effect of PBG on consumer loyalty is higher for foreign than for domes-

tic MNCs.

We suggest that the value creation process via functional and psychological value depends on the

MNC origin. According to theory (Feldman and Lynch 1988), MNC origin information stimu-

lates and enhances the salience of the functional and psychological evaluation routes that affect

the creation of consumers’ loyalty intentions through PBG.

We have theoretically argued that MNCs’ globalness information becomes diagnostic via

functional and psychological value for consumers’ loyalty intentions and that a foreign MNC’s

origin functions as a diagnosticity multiplier for PBG. Because foreign origin affects both cogni-

tive and affective behavior (e.g., Roth and Diamantopoulos 2009; Askegaard and Ger 1998), both

value mechanisms are likely activated and equally render PBG diagnostic information when con-

sumers evaluate foreign (vs. domestic) MNCs and their corporate brands. On the one hand, for-

eignness may be associated with the successful internationalization of those MNCs (e.g., Masca-

renhas 2009), i.e., consumers might perceive their offers to be of superior quality, which makes

accessible PBG information more diagnostic via functional value. On the other hand, while for-

eignness also conveys a fascination with the MNC’s successful internationalization, it may nota-

bly intensify the glamor that consumers experience from their membership in the “global village”

because it is characterized by excitement and enchantment (Zhou and Belk 2004), stimulating

consumers’ feelings of esteem or prestige toward foreign MNCs. Thus, foreignness makes

MNCs’ PBG diagnostic via psychological value. Additionally, MNCs such as Procter & Gamble

increasingly consider their corporate brand across nations, evaluating local consumers’ cognitive

and emotional responses to draw on both value types abroad (Lafley 2009).
 11

Studies support this reasoning, as for Western foreign (vs. local-only) retailers, the pathways

through which PBG translates into consumer behavior have been shown to arise equally through

functional and psychological value (Swoboda, Pennemann, and Taube 2012). Additionally, con-

sumers have been shown to associate foreign product brands’ origins with quality or value for

money and with esteem or prestige (e.g., Dimofte, Johansson, and Ronkainen 2008 for the U.S.;

Holt, Quelch, and Taylor 2004 across nations; Zhou, Yang, and Hui 2010 for China).

Concerning domestic MNCs, we have theoretically argued that, although a domestic

origin increases the diagnosticity of PBG to a lesser extent, both value mechanisms are likely to

be activated. Although consumers may rely on domestic MNCs’ functional value (e.g., Kumar

and Steenkamp 2013) because emerging giants, such as China’s Lenovo or India’s Tata Group,

offer considerable quality as market leaders in their respective industries and home countries

(Banerjee, Prabhu, and Chandy 2015), we believe that consumers link domestic MNCs’ global-

ness impressions more strongly to psychological value. Quality evaluations can depend on a par-

ticular competitive situation with foreign MNCs and may be a valuable mechanism for the in-

creased diagnosticity of the PBG evoked by domestic origins. However, a substantial body of re-

search on global product brands theoretically argues and shows across several nations that con-

sumers feel particularly affectively connected to domestic offers because of their closeness to, for

example, consumers, trust, or cultural embeddedness (e.g., Ger 1999; Schuiling and Kapferer

2004; Winit et al. 2014). Additional origin information might predominantly enhance the psycho-

logical value pathway because of possible social self-enhancement and because specific feelings

or pride in the international success of a domestic MNC’s corporate brand will likely outweigh

functional value benefits (e.g., Verlegh 2007 on Western countries in the product brand context).

These value creation processes are assumed to be relevant across nations; therefore, we

propose the following hypotheses:

Hypothesis 3a: When MNCs are foreign, PBG equally contributes to consumer loyalty through

functional value and psychological value.

 12

Hypothesis 3b: When MNCs are domestic, PBG contributes more to consumer loyalty through

psychological value than through functional value.

Moderating role of consumer ethnocentrism

Consumers have higher or lower degrees of ethnocentrism (Shimp and Sharma 1987); less ethno-

centric consumers are more globally minded, cosmopolitan, and open to global offers, while more

ethnocentric consumers believe that buying foreign-made products damages the domestic econ-

omy. Both groups have feelings of belongingness and positive in-group identities (Brewer 1991).

We now analyze how the total effect of PBG differs between less and more ethnocentric consum-

ers and which value pathway (indirect effect) is dominant in each group.3

Concerning the total effect, we have theoretically argued that PBG linked with functional

and psychological value becomes diagnostic for loyalty intentions (Feldman and Lynch 1988),

while diagnosticity will likely differ depending on consumers’ ethnocentrism. A lower level of eth-

nocentrism may function as a diagnosticity multiplier for PBG, while a higher level of ethnocen-

trism may diminish the diagnosticity of PBG. Consumers prefer belief-consistent information

(e.g., Zhang and Khare 2009) and respond more favorably to information that allows belief con-

sistency (Wheeler, Petty, and Bizer 2005).

For less (vs. more) ethnocentric consumers, accessible PBG information becomes more

diagnostic for loyalty intentions via the value offered by an MNC because less ethnocentric beliefs

and the aspired-to belief consistency allow for the establishment of a mental link between PBG in-

formation and MNCs’ value propositions. Greater belief consistency increases the likelihood of a

favorable response to PBG information. By contrast, for more ethnocentric consumers, MNCs’

PBG is less diagnostic because ethnocentric beliefs and the aspired-to belief consistency hamper

the establishment of such mental links. Hence, the total effect of PBG is likely to be stronger for

consumers with lower (vs. higher) ethnocentrism, as indicated for product brands (e.g., Akram,

Merunka, and Akram 2011 for the PBG-value links; Klein 2002 for global brand attitudes in the
 13

U.S.).

Concerning the value paths, we believe that accessible PBG information becomes more

diagnostic via psychological (vs. functional) value for less ethnocentric consumers. The latter’s

beliefs and the aspired-to belief consistency particularly allow for the establishment of a mental

link between globalness information and the psychological benefits of MNCs. Thus, such con-

sumers may link MNCs’ PBG with functional and psychological value (in different contexts e.g.,

Akram, Merunka, and Akram 2011 and Brewer 1991), and both links may increase loyalty inten-

tions. However, within this group, the psychological value pathway is likely to be stronger, i.e.,

particularly belief consistent, because emotions, feelings of belongingness, and the quest for

global identity affect consumers’ emotional attachment to global offers more strongly than their

perceptions of the functional aspects offered by an MNC (similar to product brands, see

Oberecker, Riefler, and Diamantopoulos 2008).

For more ethnocentric consumers, we believe that accessible PBG information becomes

more diagnostic via psychological value because of the higher probability that this information

will establish a mental link. More ethnocentric consumers are not motivated to learn about global

brands because their brand-related cognitive processing is superficial (Supphellen and Grønhaug

2003), making PBG less diagnostic with regard to loyalty intentions via functional value. They

also tend to disregard the functional value benefits of global offers and feel animosity because

they do not conform to their beliefs. Hence, the functional value pathway is weak. Although the

psychological value pathway also contradicts the beliefs of more ethnocentric consumers, even

those consumers are psychologically affected by global brands (e.g., product brands, Dimofte, Jo-

hansson, and Ronkainen 2008). Furthermore, they are likely to rely more on affective (vs. cogni-

tive) evaluations in their purchase behavior (e.g., Verplanken, Hofstee, and Janssen 1998 on both

product and corporate brands). Self-concept theory-based reasoning supports our assumption of

stronger effects of psychological value paths for both more and less ethnocentric consumers; for
 14

both groups, self-congruence with a brand’s PBG can override cognitive assessments of func-

tional value benefits and instead foster psychological attachment (Malär et al. 2011).

We assume similar total effects and value paths across nations for three reasons: ethnocen-

trism is said to be similarly grounded because of similar belief structures across nations (Papado-

poulos, Heslop, and Bamossy 1990); ethnocentrism is prevalent across different cultures (Sharma

2015); and, for global products and more ethnocentric consumers, scholars have highlighted

weakened globalness-value links in various countries (e.g., Papadopoulos, Heslop, and Bamossy

1990; Akram, Merunka, and Akram 2011). Therefore, we hypothesize the following across na-

tions:

Hypothesis 4: The total effect of PBG on consumer loyalty is stronger for consumers with lower

ethnocentrism than for consumers with higher ethnocentrism.

Hypothesis 5: For consumers with (a) lower and (b) higher ethnocentrism, PBG contributes more

to consumer loyalty through psychological value than through functional value.

EMPIRICAL STUDY

Sample

To develop our sample, we cooperated with an MNC in the environmentally sensitive chemical

and pharmaceutical industry. In this industry, the likelihood of both leading foreign and domestic

MNCs in a country is high, and individuals are sensitive to the reputation of corporate brands

and less sensitive to product brands, which are likely to be affected by the corporate level

(Leisinger 2005). We identified the ten largest economies in the world (based on GDP; United

Nations 2014) because they are of paramount importance for most western MNCs and because

they have a high probability of having spawned domestic MNCs. Thus, consumers in those

countries are likely to be familiar with both foreign and domestic global MNCs as well as with

global brands in general (e.g., Strizhakova, Coulter, and Price 2011). The choice of MNCs oc-

curred in two steps: identifying the leading MNCs in terms of sales for each country (e.g.,
 15

GlobalData 2010) and then the six strongest competing MNCs in each country based on the ex-

pertise of managers from the subsidiaries and from the headquarters of our cooperation partner.

Those various subjectively chosen corporate brands of MNCs were analyzed, including, e.g.,

non-prescription and prescription drugs, crop and chemical consumer and personal care products,

and nutrition products, the relative importance of which remained unobserved in this study (for

globalness differences by product categories, see Strizhakova and Coulter 2015).

We ensured that all MNCs were global (i.e., all were present in at least 45 countries; their

foreign sales ratios and foreign employee ratios were mostly ≥ 50%; and they had substantial sales

of ≥ 10% of total sales in at least three continents each). India, Japan, and the U.S. were chosen for

this analysis for two reasons: first, they differed in economic strength (emerging vs. developed;

Alden et al. 2013) and culture and hence represented differing positions on an economic develop-

ment and culture continuum (e.g., Özsomer 2012). Studies report distinct differences between con-

sumers from emerging and developed countries and from different cultural contexts (Alden,

Steenkamp, and Batra 1999; Batra et al. 2000; Burgess and Steenkamp 2006). Therefore, our selec-

tion allows us to compare the results for each country and to generalize observations (Özsomer

2012). Second, India, Japan, and the U.S. hosted at least two domestic MNCs and two foreign

MNCs among the six competitors chosen, which was a prerequisite for our research design of for-

eign vs. domestic origin comparisons. Perhaps unsurprisingly, emerging MNCs were less global-

ized, though only slightly less so with regard to the mentioned criteria. Additional countries did not

host both foreign and domestic MNCs among the six competitors or were used for control purposes

in alternative models, e.g., China as an emerging country (with one domestic MNC) and Italy as a

developed but economically weaker country than the U.S. The foreign MNCs in each country were

selected from Germany and the German-speaking part of Switzerland to allow us a certain degree of

control over the potential differences in the MNCs’ home countries. The remaining two foreign

MNCs were used for control purposes in alternative models, as discussed in the results section.

After pre-tests, data were collected by a commercial marketing research agency that cross-
 16

nationally operates panels (mostly web surveys in those countries) with an average response rate

of 55%. The participants were compensated with cash rewards (Madden, Roth, and Dillon 2012).4

To select the respondents from each country, two screening criteria were used:

First, quota sampling according to gender and age distribution was applied based on in-

formation provided by national registration offices. The sampling was restricted to the urban

population between 18 and 65 years of age in large cities (e.g., due to reasons of familiarity with

the MNCs; Strizhakova, Coulter, and Price 2011). Because we have not controlled for the re-

gional locations of the respondents or further geographic characteristics within the countries,

our results are limited in this respect.

Second, in order to ensure comparable samples across nations (e.g., Özsomer 2012), the

respondents had to meet one of the following criteria: above-average income or a high level of

education or profession. To ensure panel quality, the agency was briefed on common tests: the

use of individualized survey links to prevent bogus respondents, the use of an instructional ma-

nipulation check, and the control/removal of respondents who needed little time or who applied

random clicking or straight-lining (Oppenheimer, Meyvis, and Davidenko 2009; Kaminska,

McCutcheon, and Billiet 2010). At the beginning of the survey, the respondents indicated their

unprompted and prompted awareness of up to six MNCs, followed by a question about their

knowledge of the MNCs (based on a 5-point Likert-type scale ranging from 1 = I do not know

the corporation to 5 = I know the corporation very well). The respondents who were at least

generally familiar (= 2) with three or more MNCs participated in the survey because general

knowledge is a precondition for evaluations (Keller 1993). A mixed design was used, i.e., the

respondents were questioned regarding two randomly assigned MNCs—most often our partner

MNC and another MNC—whose order in the survey was rotated to avoid repetition effects

(e.g., Chang, van Witteloostuijn, and Eden 2010; Tourangeau 2004).

This procedure led to 2,647 evaluations, which were quite similarly distributed over the

four competing firms in each of the three countries after we randomly reduced the responses to our
 17

partner to ensure a balanced design of the data across MNCs. After excluding incomplete cases

and detecting outliers based on the Mahalanobis distance, the sample included 704 responses

from India, 794 from Japan, and 662 from the U.S. (see Table 2). The final sample consisted of

2,160 responses; however, it is not representative of all countries’ population, as an ex post

comparison showed. Normality tests showed that the samples deviated from multivariate nor-

mality; therefore, the mean-adjusted maximum likelihood estimator was chosen to test the hy-

potheses, as it provides a robust chi-square test and handles potential threats within the data struc-

ture (Asparouhov 2005). Chi-square difference tests were conducted using scaling corrections

(Satorra and Bentler 2001).

Table 2

Measurement

Regarding the measurements, we considered general aspects (such as the hierarchy of effects in

all panels). We relied on scales from previous studies (using five-point Likert-type scales from

1 = strongly disagree to 5 = strongly agree; see Table 3) and measured the PBG of the corporate

brands using three items adapted from Steenkamp, Batra, and Alden (2003) and intentional loy-

alty toward the MNCs using three items adapted from Oliver (1999; 2015). To measure con-

sumers’ perceptions of value, we adapted the scale developed by Sweeney and Soutar (2001) to

the context of corporate brands (Alden et al. 2013; Swoboda, Pennemann, and Taube 2012) and

measured functional value and psychological value using four items each. We used MNCs’ ob-

jective origins, i.e., foreign vs. domestic. Consumer ethnocentrism was measured using four

items adapted from Shimp and Sharma (1987). The median split technique was used to group

consumers according to lower vs. higher ethnocentrism (following Akram, Merunka, and

Akram 2011 and Steenkamp, Batra, and Alden 2003 in the PBG context). Additionally, con-

sumer ethnocentrism was used as a continuous variable (e.g., Nijssen and Douglas 2011;

Sharma 2011).

The scales were quantitatively pre-tested in each country (average N = 154 per country).
 18

These pre-tests yielded satisfactory values for reliability and validity. All the scales in the main

studies were also tested for reliability and validity (see Tables 3 and 4). To ensure the reliability

of our measurements, we examined factor loadings (all ≥ .737) and Cronbach’s alpha (all values

are >.879) and reached the common thresholds of .700 or .600, respectively (Nunnally 1978;

Bagozzi and Yi 1988). For construct validity, all the CFA factor loadings were ≥ .720 (above

.500, see Hair et al. 2014). All values for Average Variance Extracted (AVE) were between

.767 and .943—thus above the threshold of .500 (Bagozzi and Yi 1988)—providing evidence of

convergent validity. Testing for discriminant validity (Fornell and Larcker 1981), we found all

AVE values to exceed the squared correlations, except for one case. Therefore, we verified dis-

criminant validity using a chi-square difference test (Anderson and Gerbing 1988). The re-

stricted model fit significantly worse (p < .001) than our proposed model; thus, discriminant va-

lidity can be assumed. The fit of our measurement model was good (CFI between .975 and

.991; TLI between .969 and 991; RMSEA between .030 and .056; SRMR between .031 and

.037).

To ensure semantic equivalence, the translation-back-translation method was applied by

commercial translation agencies that specialize in market research. These agencies were briefed

about the requirements of appropriate semantic equivalence processes (Hult et al. 2008).

We included covariates in the study. Gender (0 = male, 1 = female) and age were con-

trolled for because they are known to influence loyalty. MNCs’ familiarity was controlled for

and measured using a scaled item: “[MNC] is very familiar to me” (Steenkamp, Batra, and

Alden 2003).

Tables 3-4

Method

We tested for measurement equivalence to ensure comparability across the origin groups and
 19

the ethnocentric consumer groups (see Table 5; Steenkamp and Baumgartner 1998). No signifi-

cant differences in model fit between all configural and metric models were found, thus sup-

porting full metric invariance for all constructs in all countries (India: origin Δ²(14) = 12.902,

p > .05; ethnocentrism Δ²(14) = 15.788, p > .05; Japan: origin Δ²(14) = 11.739, p > .05; eth-

nocentrism Δ²(14) = 16.591, p > .05; U.S.: origin Δ²(14) = 25.732, p > .05; ethnocentrism

Δ²(14) = 23.104, p > .05). We also tested for the measurement equivalence of the PBG, func-

tional and psychological value, loyalty constructs and the entire measurement model across all

countries (establishing full and partial metric invariance) (see web appendix A, which is availa-

ble upon request).5

Common method variance (CMV) was addressed in three ways: we used an appropriate

questionnaire design and a single-factor test (which showed significantly worse fit values than

the proposed model, India: Δ²(6) = 1499.429, p < .001; Japan: Δ²(6) = 2407.771, p < .001;

U.S.: Δ²(6) = 3520.472, p < .001), and we applied the marker variable technique using the oc-

cupation variable, which is theoretically uncorrelated with our constructs (Rindfleisch, Bur-

roughs, and Wong 2009). The latter revealed no significant changes in the coefficients and cor-

relations (method variance was less than 8.50% in all countries; Williams, Hartman, and

Cavazotte 2010), which underscored the low probability of CMV in this study (for detailed re-

sults, see web appendix B).

We tested for endogeneity using the instrumental variable (IV) exposure to the MNCs’

marketing, especially their communication activities, which represents a core antecedent of

PBG (Steenkamp, Batra, and Alden 2003). For this purpose, we used one item: “When I have

read magazines or watched TV, I have often seen advertising for offers of [MNC]” (adapted

from Cleveland and Laroche 2007). We checked for the IV’s strength (F-values > 10; Antona-

kis et al. 2014) and calculated an efficient model that did not significantly differ from the con-

sistent model, which indicated that PBG was exogenous (Hausman 1978; z-values all < 1.96).
 20

The probability of the endogeneity of PBG seems to be reduced (for details, see web appendix

C).

We used Mplus 7.31 for hypothesis testing.

Table 5

RESULTS

Results of hypothesis tests

Next, the results of the hypothesis tests and the alternative models are presented (see Table 6).

Model 1 provides support for H1a and H1b in all three countries. PBG positively affects

loyalty through functional value (India: b = .294, p < .001; Japan: b = .065, p < .001; U.S.:

b = .163, p < .001) and psychological value (India: b = .310, p < .001; Japan: b = .100, p < .001;

U.S.: b = .299, p < .001). The direct effects of PBG on consumer loyalty in the rival models

were not significant in all three countries, as supported by a Wald test (Molenberghs and

Verbeke 2007) (India: Δ² = 3.529(1), p > .05; Japan: Δ² = 3.175(1), p > .05; U.S.:

Δ² = 2.396(1), p > .05) and by a bootstrap test (5,000 bootstrap samples, India/Japan/U.S.: b -

.072/-.028/-.088, standard error .028/.015/.049, lower interval limit -.108/-.056/-.183, upper in-

terval limit .001/.001/.008; for details, see web appendix D). The indirect effects of PBG on

consumer loyalty via functional and psychological value were significant in all three countries

(5,000 bootstrap samples, India/Japan/U.S.: via functional value: b .316/.068/.172, standard er-

ror .033/.013/.041, lower interval limit .254/.045/.094, upper interval limit .384/.094/.258; via

psychological value: b .333/.105/.318, standard error .035/.018/.050, lower interval limit

.271/.072/.224, upper interval limit .406/.142/.422; for details, see web appendix D). Because

the confidence intervals for the direct effects include zero and those for the indirect effects do

not, i.e., they are significant, indirect-only mediation is established (Zhao, Lynch Jr., and Chen

2010).

The results from Model 2 support H2 for India and Japan. The total effect of PBG on
 21

loyalty is stronger for foreign MNCs than for domestic MNCs in India (bforeign = .607, p < .001;

bdomestic = .415, p < .001; Δbforeign vs. domestic = .192, p < .001) and in Japan (bforeign = .311, p <

.001; bdomestic = .217, p < .001; Δbforeign vs. domestic = .094, p < .001). However, H2 is not supported

for the U.S. (bforeign = .469, p < .001; bdomestic = .440, p < .001; Δbforeign vs. domestic = .029, p > .05).

In H3a, we hypothesized that when an MNC is foreign, PBG contributes to loyalty equally

through functional value and psychological value. Our data support this assumption for India

(ΔbFV-PV = .023, p > .05) and Japan (ΔbFV-PV = .024, p > .05), but again not for the U.S., where

functional value shows stronger effects (ΔbFV-PV = .121, p < .001). H3b is supported. For do-

mestic MNCs, PBG contributes to loyalty more through psychological value than through func-

tional value in India (ΔbFV-PV = -.172, p < .001), Japan (ΔbFV-PV = -.077, p < .001), and the U.S.

(ΔbFV-PV = -.241, p < .001). These results will be discussed further below. The direct effects of

PBG on consumer loyalty were not significant (for details on the Wald test and the bootstrap

test, see the rival models in web appendix E).

Finally, the results from Model 3 support H4. The total effect of PBG on loyalty is

stronger for less (vs. more) ethnocentric consumers in all three countries (India: Δblower-higher =

.091, p < .001; Japan: Δblower-higher = .111, p < .001; U.S.: Δblower-higher = .077, p < .001). The re-

sults support H5a in Japan (ΔbFV-PV = -.042, p < .01) and the U.S. (ΔbFV-PV = -.063, p < .001),

as for less ethnocentric consumers, PBG contributes more to consumer loyalty through psycho-

logical than through functional value, though not in India, where both value types are equally

strong (ΔbFV-PV = .023, p > .05). H5b supports our assumptions concerning more ethnocentric

consumers in all countries (India ΔbFV-PV = -.050, p < .01, Japan ΔbFV-PV = -.025, p < .05, and

the U.S. ΔbFV-PV = -.243, p < .001). These results will be discussed further below. Again, a non-

significant direct effect of PBG on consumer loyalty was found (for details, see web appendix E).

Table 6

Alternative models and additional results

Alternative models were calculated to strengthen the observations and to reduce individual
 22

countries’ possible idiosyncrasies (for details, see web appendix E). First, all the hypotheses

were tested in a second emerging country (China, with one domestic MNC among the six com-

petitors; N = 658) and in a second developed country (Italy; N = 731). In both countries, the hy-

pothesis tests show identical results to those for India and the U.S., which allows for a certain

generalization of the observations for emerging and developed countries.

Second, in all the countries, alternative models were calculated by replacing the Ger-

man-speaking foreign MNCs with the two remaining competitors from the U.S. and France in

India and Japan and from the U.S. and the U.K. in China and Italy (India N = 735, Japan N =

701, China N = 632, and Italy; N = 725). In the U.S., domestic MNCs were replaced because no

other foreign MNCs were available (N = 658). These results supported all hypotheses reported

for the consciously chosen German-speaking MNCs, which allow a certain generalization and

control of, for example, foreign MNCs’ host country image influences.

Third, studies on PBG addressed ethnocentrism as a moderator using consumer groups

or a continuous variable. Therefore, we also tested ethnocentrism as a continuous variable (using

the residual centering approach, see Little, Bovaird, and Widaman 2006; Steinmetz, Davidov,

and Schmidt 2011). As ethnocentrism affects the value perceptions attributed to global brands,

we tested for moderation in the PBG-value link first. Significant negative effects were found for

India (China) and Japan but not for the U.S. (Italy; see web appendix F, Tables 11a-d). The in-

teraction effects concerning functional and psychological value are approximately equal in

strength (India: bPBG×CE-FV = -.187, p < .001; bPBG×CE-PV = -.198, p < .001; Japan: bPBG×CE-FV = -

.142, p < .01; bPBG×CE-PV = -.161, p < .01). Linearly increasing ethnocentrism reduces both PBG-

value links for MNCs to a similar extent (similar to the group comparisons, see Table 6). The in-

significant results in the U.S. (significant in the group comparison) can be regarded as more rig-

orous evidence. However, studies in Western countries show that attitudes toward global prod-

uct brands are affected by ethnocentrism (e.g., Kwak, Jaju, and Larsen 2006; Verlegh 2007). As

none of the 40 studies on ethnocentrism addresses corporate brands, we may conclude that, for
 23

U.S. (Italian) consumers, PBG is not affected by ethnocentrism because of their long participa-

tion in the global marketplace or the large number of MNCs in these countries (Interbrand

2014). However, multi-group models indicate stronger effects for less (vs. more) ethnocentric

consumers for all PBG-value links, total pathways, and for most indirect pathways, while testing

a third middle ethnocentric group led to inconclusive results. Such consumers might be indiffer-

ent regarding preferences/animosities against MNCs. This observation is not stable because of

small sample size, but it might indicate ethnocentrism’s non-linear role in the PBG effects of

MNCs. Therefore, we call for further research on this issue. In summary, the additional results

are conclusive for three countries.

We also tested consumer ethnocentrism as a continuous moderator for the full indirect

effect model, i.e., PBG-value and value-loyalty (see web appendix F, Tables 11e-h). The model

fits are poor, probably because of the insufficient sample size for stable interaction tests of those

models; therefore, the results need to be interpreted with caution. In India (China), ethnocen-

trism affects both PBG-value links negatively and equally (bPBG×CE-FV = -.163, p < .001;

bPBG×CE-PV = -.164, p < .001), while the value-loyalty links are affected differently (bFV×CE-LOY =

-.291, p < .001; bPV×CE-LOY = .289, p < .001). Similarly, in Japan, increasing ethnocentrism af-

fects both the PBG-value links negatively (slightly stronger functional value, bPBG×CE-FV = -.304,

p < .001; bPBG×CE-PV = -.225, p < .001), and the value-loyalty links differently (bFV×CE-LOY = -

.166, p > .05; bPV×CE-LOY = .148, p > .05). All interaction effects are insignificant in the U.S.

(and Italy), and we draw similar conclusions to those presented above. In summary, in three

countries, ethnocentrism equally reduces the effects of PBG on both value types and, at least ac-

cording to the observed trend, reduces the functional value-loyalty link and increases the psy-

chological value-loyalty link. For MNCs, increasing ethnocentrism weakens the indirect effects

of PBG via functional value twice, while the effects of psychological value are partly reduced

and partly reinforced. PBG information becomes more diagnostic when linked with psychologi-

cal value than when linked with functional value for loyalty intentions towards an MNC.
 24

Fourth, a three-way interaction is possible: foreign origin strengthens the positive PBG

effect, but this moderation is weakened by ethnocentrism, with less ethnocentric consumers be-

ing more ready to use the foreign origin information as a multiplier of PBG. Scholars may have

such a relationship in mind (when focusing only on consumer attitudes toward global product

brands depending on ethnocentrism, see Guo 2013; Dimofte, Johansson, and Ronkainen 2008).

Again, ethnocentrism’s effects on the PBG-value links for foreign vs. domestic MNCs are com-

pared first. We found significant negative interaction effects for foreign MNCs in all the models

for India (bPBG×CE-FV = -.152, p < .001; bPBG×CE-PV = -.169, p < .001, similarly, China) and Japan

(bPBG×CE-FV = -.362, p < .001; bPBG×CE-PV = -.426, p < .001), but not for domestic MNCs in these

countries (see web appendix F, Tables 12a-d for details). The effects for both value types are

relatively equal (stronger in Japan). We found insignificant effects in all cases for the U.S. (It-

aly), supporting the results discussed above. In three countries, ethnocentrism has a dampening

effect on foreign MNCs’ PBG-value links, and, because both value types are equally affected,

we refer to our initial models: foreign MNCs’ PBG equally contributes to loyalty through func-

tional and psychological value, while domestic MNCs’ PBG contributes more strongly through

psychological value (see H3a-b).

The results enhance those on product brands (discussible ethnocentrism effects in

emerging countries, e.g., Guo 2013, vs. verified effects in the U.S., e.g., Dimofte, Johansson,

and Ronkainen 2008) because we found a significant interaction not only for emerging, cultur-

ally distant countries but also for Japan, a developed, culturally distant country. The results are

not fully comparable with our total/indirect effect models, and they will be discussed further be-

low. Second, we could not test a three-way interaction for the PBG-value-loyalty pathways with

structural equations models (SEM) (nor by treating ethnocentrism as a continuous variable or by

using a four-tailed matrix with MNCs’ two origin groups * two ethnocentrism groups) because

of the small sample size in each group and country (N = 158-198 in the matrix fields). We call

for further research on this issue.

 25

DISCUSSION AND IMPLICATIONS

This study contributes to our understanding of the underlying pathways through which MNCs’

corporate brands transform PBG into consumers’ loyalty intentions in different country environ-

ments and helps us determine whether and how boundary conditions—MNCs’ foreign vs. do-

mestic origin and consumer ethnocentrism—change the total effect and the value pathways of

PBG into loyalty. As a result, we advance two findings on global corporate brands (Swoboda,

Pennemann, and Taube 2012; Alden et al. 2013) and add to the literature with a novel perspec-

tive, as corporate brands are a CEO’s task and are strategically important for MNCs that rely on

their corporate brands or value in international expansion and that want to consider the role of

PBG in competition with domestic MNCs. As our study consciously builds on extant research

(especially on the studies of Swoboda, Pennemann, and Taube 2012 and Steenkamp, Batra, and

Alden 2003, allowing comparisons) and covers only one industry context that is sensitive to

corporate brands in five of the ten largest economies in the world, we carefully provide major

implications for research and suggestions for managers and refer to studies on product brands to

help explain our findings.

Theoretical implications

First, regarding the initial research question, the results for MNCs’ corporate brands support our

theoretical reasoning and contribute to the literature, which is dominated by studies on global

product brands. The PBG of MNCs does not directly affect behavioral intentions (for direct

effects of product and service brands, see Winit et al. 2014) and constitutes accessible infor-

mation that becomes diagnostic for consumers’ intentional loyalty through pathways of both

functional and psychological value. We believe that this observation is notable because it en-

hances extant studies on corporate brands (Alden et al. 2013 and particularly Swoboda, Penne-

mann, and Taube 2012, who initially show similar results for global vs. local-only retail brands

in China). Analyzing manufacturing and leading local/foreign MNCs, we show that these global
 26

corporations do not participate in PBG directly but through offered value, which is attributed to

the whole corporation (H1). Corporate brands are strategically relevant for international expan-

sion (e.g., even for choices of green vs. brownfield investments) and for product brands (e.g.,

within an endorser branding strategy, Berens, van Riel, and van Bruggen 2005). The value asso-

ciated with corporate brands and transferred to product brands and categories is advantageous,

while product brands can, in turn, add meaning themselves (e.g., Aaker 2004). Therefore, we

call for more research on corporate brands and their relationships to product brands (e.g., for

such a reciprocity Swoboda, Berg, and Schramm-Klein 2013; for brand extensions Sichtmann

and Diamantopoulos 2013).

To provide a richer discussion, we also highlight interesting links between variables that

constitute the hypothesized pathways. In doing so, we wish to emphasize that PBG particularly

affects the value of MNCs in emerging countries (India, China) and in developed Western

countries (U.S., Italy). In Japan, PBG-value links are weaker than value-loyalty links, i.e.,

MNCs that provide potentially demanding Japanese consumers with quality, value for money,

and, in particular, emotional and social aspects of value increase loyalty behavior.

Second, reflecting MNCs’ competition in our next research question, for the first time,

we test how MNCs’ foreign vs. domestic origins affect the creation of loyalty intentions through

PBG. PBG does not directly determine loyalty in any case, but MNCs’ origins affect the total

effect and the value mechanisms by which PBG translates into loyalty. The results are not con-

sistent across developed countries, and we discuss two conclusions in greater detail.

(1) According to theory (Feldman and Lynch 1988), MNCs’ objective origins tend to act

as diagnosticity multipliers for PBG by affecting the total effects of PBG. Scholars have found

some benefits of PBG for foreign brands vs. local-only brands (e.g., product and corporate,

Zhou, Yang, and Hui 2010; Swoboda, Pennemann, and Taube 2012) but less for foreign vs. do-

mestic global product and service brands (Riefler 2012: no effects for domestic global brands in
 27

Austria; Winit et al. 2014: mixed effects for industries in Thailand). Although we observe

MNCs for which PBG will likely be generally beneficial, overall, foreign MNCs benefit more

than domestic MNCs from PBG. The differences in total effects (H2) between foreign and do-

mestic MNCs are significant in emerging countries, such as India (China), and in Japan. Addi-

tionally, foreign MNCs have mostly strong PBG-value links in these countries, i.e., consumers

are sensitive to the multiplier effects, and foreign (competing with domestic) MNCs can more

ably play the global card.

The same differences are insignificant in developed Western countries, such as the U.S.

(Italy), a country for which several studies demonstrate globalness advantages for foreign prod-

uct brands (e.g., Papadopoulos, Heslop, and Bamossy 1990; Sharma 2011). Observing similar

total effects and similar PBG-functional value links for both origins, we may conclude that

globalness information pays off for domestic MNCs and that the diagnosticity multiplier effect

is less relevant for foreign MNCs (for whom PBG-psychological value links are stronger). Con-

sumers’ lower sensitivity to foreignness may result from their long participation in and open-

ness to global markets and, in the case of the U.S., from its history of immigration (e.g.,

Guzmán and Paswan 2009) and diverse ethnic consumer groups (with different values, atti-

tudes, and social behaviors, Phinney 1996; Dimofte, Johansson, and Bagozzi 2010). A coun-

try’s economic importance is not the only factor that explains differences in the total effects of

PBG across nations (maybe culture difference or image perceptions of host countries are

relevant as well, Özsomer 2012; Sichtmann and Diamantopoulos 2013). Because we could not

control for this issue and because our models are already complex, we call for further investiga-

tions. However, in doing so, and contrary to the results for H1, a distinction between the acces-

sible information on globalness, i.e., being perceived as an MNC, and on foreignness, i.e., being

a foreign or domestic MNC (e.g., Winit et al. 2014), as well as the perspectives of different

countries’ contexts, should be applied in the future.

 28

(2) Regarding value pathways that translate PBG into loyalty for domestic MNCs, a sig-

nificantly stronger psychological (vs. functional) value pathway is evident in all the countries

and models (H3b). This result supports the assumption that domestic firms benefit largely from

the emotional and social aspects of value in their home countries (for domestic-only firms see

Alden et al. 2013; Swoboda, Pennemann, and Taube 2012). This effect may be even stronger

because of consumers’ specific feelings of pride for or social bonds with domestic firms that are

MNCs (e.g., Ger 1999) and because of relatively stronger PBG-psychological value and psy-

chological value-loyalty links in most countries. For domestic MNCs, psychological value has a

strong diagnostic function with regard to intention behavior.

Foreign MNCs attract consumers more through functional value in the U.S. (the

functional value was often shown to drive global product brands’ success, e.g., Holt, Quelch,

and Taylor 2004), while consumers are equally attracted through functional and psychological

value in all other countries. Additionally, for foreign (vs. domestic) MNCs, the functional value

pathways of PBG into loyalty are stronger in all countries, and the psychological value path-

ways are often similar (weaker only in the U.S.). These observations may contribute to partly

inconclusive results on product brands (e.g., affect, Dimofte, Johansson, and Ronkainen 2008,

vs. quality, Holt, Quelch, and Taylor 2004) and to the few results on corporate brands.

Swoboda, Pennemann, and Taube (2012) show stronger functional and psychological value for

foreign (vs. local-only) retail chains/corporations in China, while we present gradual differ-

ences, which we may explain with the aforementioned adaptation efforts of foreign MNCs

abroad and their brand-building capabilities. However, because foreign (vs. domestic) MNCs

also have stronger functional value-loyalty links in all countries, future research may focus on

the antecedents of such perceptions.

Third, our results on consumer ethnocentrism are novel for corporate brands, but they

can be compared to two moderation studies on PBG for product brands (Steenkamp, Batra, and
 29

Alden 2003 in the U.S. and Korea; Akram, Merunka, and Akram 2011 in Pakistan). MNCs’

PBG affects loyalty only through offered value for both less and more ethnocentric consumers,

though it affects the first group more strongly (H4). By contrast, Steenkamp, Batra, and Alden

(2003) only show a total effect of PBG on purchase intentions for less ethnocentric consumers,

and they do not specify the pathways. We may conclude that our specific industry context af-

fects our results and that MNCs do not represent local icons for which we have not controlled

(e.g., Steenkamp, Batra, and Alden 2003; Özsomer 2012). We can also conclude that providing

functional and (particularly) psychological value to more ethnocentric consumers reduces the

impact of negative consumer beliefs toward MNCs (e.g., Alden et al. 2013). High levels of eth-

nocentrism weaken—but do not completely erase—the total effect of the PBG of those corpo-

rate brands. Moreover, we find consistent results for a strong value pathway through psycholog-

ical (vs. functional) value for more ethnocentric consumers in all countries and for less ethno-

centric consumers (for them, the functional path is also equally strong in India and China, H5).

Observing the pathways is valuable because Steenkamp, Batra, and Alden (2003) mention that

stronger quality paths for consumers who are less (vs. more) ethnocentric, while Akram,

Merunka, and Akram (2011) show dominant PBG-quality and PBG-prestige links for consum-

ers who are less (vs. more) ethnocentric. Our results indicate different belief-accessibility mech-

anisms and stronger psychological (vs. functional) value-loyalty links for more ethnocentric

consumers.

We tested ethnocentrism as a continuous moderating variable for the first time in PBG

research (in other contexts, only by Nijssen and Douglas 2011; Winit et al. 2014; Sharma 2011).

The PBG-value link and the PBG-loyalty pathways are moderated in emerging countries and in

Japan, i.e., increasing ethnocentrism reduced PBG effects, but not in the U.S. (Italy). In the

U.S.—where H2 and H3a were not supported—the already mentioned reasons might explain

the insignificant result, e.g., long participation in and openness to global markets and less ethno-

centrism. The differences in multi-group results likely results from the lack of consideration of
 30

the gradual linear effects analyzed in continuous interaction models. As a result, the effects

might be overemphasized and may be related to unobserved factors (Irwin and McClelland

2003). We call for tests of nonlinear conceptualizations because of the insignificant results in

continuous models or in middle-ethnocentrism groups and because research has yet to address

this option.

Finally, we found a three-way interaction of MNCs’ origins and consumer ethnocen-

trism in India (China) and Japan, but not in the U.S. (Italy), for the PBG-value links. No interac-

tion in the U.S. indicates no further effects for MNCs with different origins. The interaction im-

plies that the foreign origin strengthens the positive PBG effect in three countries; however, this

moderation is further weakened by increasing ethnocentrism, while less ethnocentric consumers

use foreign origin information as a multiplier of PBG. For increasing ethnocentrism, the acces-

sible PBG information on foreign MNCs becomes less diagnostic via functional and psycholog-

ical value because ethnocentric beliefs and the aspired-to belief consistency hamper the estab-

lishment of mental links between globalness information and the value proposition of MNCs.

We have appropriately evaluated interactions in SEM, but the results are similar to product

brand research (e.g., negative effects on brand preferences for global and foreign brands, He and

Wang 2015). For domestic MNCs, no effects occur, i.e., theoretically, PBG becomes diagnostic

via value perceptions, irrespective of ethnocentric tendencies because mental links are not ham-

pered by inconsistent beliefs. However, we observe the three-way interaction only for the PBG-

value link, while future research may consider value-loyalty effects and, thus, important total

PBG-loyalty pathways.

Managerial implications

CEOs and other responsible managers may be aware of the role that MNCs’ PBG plays in for-

eign countries, but they may claim that PBG only pays off indirectly when attracting consumers

through functional and psychological value, i.e., globalness perceptions alone do not generate
 31

loyalty intentions. Because consumers link PBG and offered value to the entire MNC, the

knowledge on related effects adds to corporate brand management across nations. The results

may even constitute a starting point for internationalization steps, e.g., when entering new coun-

tries and deciding to rely on the advantages in local competition with domestic MNCs or do-

mestic-only firms. Globalness and the strong value offered by an MNC are important when at-

tracting customers (as is corporate reputation, Swoboda, Puchert, and Morschett 2015). A prod-

uct brand perspective draws a more fine-grained picture for decisions in direct competition but

is likely to be affected by a strategically important corporate brand (Balmer 2010; Berens, van

Riel, and van Bruggen 2005). The implications for managers of foreign and domestic MNCs

differ.

(1) For managers of foreign MNCs, this study reveals that PBG can be transformed into

consumer loyalty intentions via functional/psychological value and that PBG serves as a source

of strategic advantage in competition and in countries of domestic, and increasingly internation-

alizing, emerging MNCs (Kumar and Steenkamp 2013). PBG pathways are promising for

emerging countries such as India and for developed, culturally different countries such as Japan,

where both the psychological and functional pathways are equal or stronger than those of com-

peting domestic MNCs. Both advantages are equally diminished by increasing consumer ethno-

centrism. PBG does not provide foreign MNCs with a competitive advantage over domestic

MNCs in developed Western countries such as the U.S., where the functional pathways are

promising, while foreign MNCs seem to struggle with forming emotional bonds with local con-

sumers; both are rather independent of ethnocentric tendencies.

Therefore, foreign MNCs should manage their PBG-loyalty pathways to maintain their

advantages, particularly their perceived functional value across nations, but they should not ne-

glect the psychological value pathway that allows them to succeed with their PBG. Targeting

less ethnocentric consumers enhances success, and thus, the relevance of those consumers, par-

ticularly in important countries, might be considered (while a positive relationship exists

 32

between low levels of ethnocentrism and economic development, Sharma 2011). When target-

ing more ethnocentric consumers, foreign MNCs might want to disguise their foreignness, for

example. They might also run marketing campaigns that inform such consumers about how the

MNC contributes to the local economy by employing local citizens or investing in infrastructure

to encourage more ethnocentric consumers to develop less negative attitudes.

(2) Domestic MNCs may continue to rely on psychological value advantages in their

home countries. For example, CEOs of domestic global corporate brands should focus on em-

phasizing the cultural connectedness or a sense of belonging that consumers can experience

from the corporate brand, which foreign brands do not offer. However, they especially need to

enhance their functional value perceptions. Here, domestic MNCs might stress qualitative as-

pects, such as the reliability and longevity of their offers, success and commitment, while differ-

entiating themselves by stressing, for example, their greater satisfaction of local customer needs

due to their spatial proximity. They also may wish to place greater emphasis on their globalness

to increase perceptions of functional and psychological value in particular, e.g., with the slogan

“rooted in the country, but at home in the world”. Such positioning does not result in any addi-

tional disadvantages for more or less ethnocentric consumers because increasing ethnocentrism

in the home country does not cause changes in the PBG-value links. Domestic (vs. foreign)

MNCs profit within more ethnocentric consumer groups and should stress their origins when

targeting these consumers.

LIMITATIONS AND FURTHER RESEARCH

To better understand the role of the PBG of MNCs’ corporate brands, further analyses are re-

quired, as our study is not without limitations. In addition to the mentioned limitations, two spe-

cific issues will be highlighted.

Although we paid special attention to data collection, broadening the database would mit-

igate some of the limitations and allow for further conclusions. For example, local-only firms,
 33

specific MNCs (with different branding strategies; Schuiling and Kapferer 2004), additional in-

dustries (as the relevance of global product brands and categories vary, see Strizhakova and

Coulter 2015 and corporate and product brands may interact differently), and more countries

could be studied to analyze the role of global corporate brands’ PBG in different contexts. We

compare important countries that may be particularly attractive for MNCs, but this selection

likely affects our results. Countries differ in multiple dimensions, and analyzing country-level

predictors is promising (for multilevel approaches see Swoboda, Puchert, and Morschett 2015).

Furthermore, our selection of homogenous samples was aided by familiarity with the MNCs and

by comparisons between countries, but it may affect the results because corporate brand infor-

mation might be particularly diagnostic for urban and well-educated respondents. Broader sam-

ples or alternative interview techniques may extend the generalizability, but they will cause addi-

tional challenges for country comparisons (e.g., Sheth 2011).

Alternative measurements of value and of consumers’ loyalty types can extend our con-

clusions because, for example, our single-point-in-time measurement of intentional loyalty tends

to make the predictive influence of antecedents more associative (e.g., Oliver, 1999, 2015). Im-

portant areas for further research would integrate the mentioned antecedents, e.g., culture, and

more detailed origin information about the MNCs’ home countries into the framework because

consumers are known to prefer brands that originate from countries that are considered to have a

particular expertise or image (e.g., Sharma 2011; Sichtmann and Diamantopoulos 2013). We

have consciously used two domestic vs. two foreign MNCs in each model, but the results might

be affected by MNCs’ dissimilarities in terms of distribution, heritage, reputation, or local icon

value, i.e., by factors that go beyond consumer awareness and that need to be observed in future

research (e.g., Swoboda, Puchert, and Morschett 2015; Steenkamp, Batra, and Alden 2003). Be-

cause our models are already complex, such possibilities were not investigated in this study.

Furthermore, future studies may test our framework in a reciprocity design that investi-

gates globalness effects of the same foreign (e.g., Western) and domestic (e.g., emerging) MNCs
 34

in the respective home countries of both MNCs, switching the foreign vs. domestic roles. How-

ever, selecting several MNCs and countries in which MNCs are leading competitors is challeng-

ing but essential to achieve valid and generalizable results in such a design. Finally, consumers

may not behave exclusively ethnocentrically or geocentrically and may instead combine both

identities (e.g., Steenkamp and de Jong 2010). We treated consumer ethnocentrism in various

ways, but a fine-grained investigation showed inconsistent results for a new middle group, which

might indicate a challenging non-linear conceptualization of ethnocentrism.

 35

REFERENCES

Aaker, David A. (2004), "Leveraging the Corporate Brand," California Management Review,

46 (3), 6–18.

Akram, Aneela, Dwight Merunka, and Muhammad Shakaib Akram (2011), "Perceived Brand

Globalness in Emerging Markets and the Moderating Role of Consumer Ethnocentrism,"

International Journal of Emerging Markets, 6 (4), 291–303.

Alden, Dana L., James B. Kelley, Petra Riefler, Julie A. Lee, and Geoffrey N. Soutar (2013),

"The Effect of Global Company Animosity on Global Brand Attitudes in Emerging and

Developed Markets: Does Perceived Value Matter?," Journal of International Marketing, 21

(2), 17–38.

Alden, Dana L., Jan-Benedict E. M. Steenkamp, and Rajeev Batra (1999), "Brand Positioning

Through Advertising in Asia, North America, and Europe: The Role of Global Consumer

Culture," Journal of Marketing, 63 (1), 75–87.

---- (2006), "Consumer Attitudes Toward Marketplace Globalization: Structure, Antecedents

and Consequences," International Journal of Research in Marketing, 23 (3), 227–39.

Anderson, James C. and David W. Gerbing (1988), "Structural Equation Modeling in Practice:

A Review and Recommended Two-Step Approach," Psychological Bulletin, 103 (3), 411–

23.

Antonakis, John, Samuel Bendahan, Philippe Jacquart, and Rafael Lalive (2014), "Causality and

Endogeneity: Problems and Solutions," in The Oxford Handbook of Leadership and

Organizations: 93-117, David V. Day, ed. Oxford: Oxford University Press.

Askegaard, Soren and Guliz Ger (1998), "Product-Country Images: Towards a Contextualized

Approach," European Advances in Consumer Research, 3 (1), 50–58.

Asparouhov, Tihomir (2005), "Sampling Weights In Latent Variable Modeling," Structural

Equation Modeling: A Multidisciplinary Journal, 12 (3), 411–34.

Bagozzi, Richard P. and Youjae Yi (1988), "On the Evaluation of Structural Equation Models,"
 36

Journal of the Academy of Marketing Science, 16 (1), 74–94.

Balmer, John M. T. (2012), "Corporate Brand Management Imperatives: Custodianship,

Credibility, and Calibration," California Management Review, 54 (3), 6–33.

---- (2010), "Explicating Corporate Brands and Their Management: Reflections and Directions

from 1995," Journal of Brand Management, 18 (3), 180–96.

Banerjee, Sourindra, Jaideep C. Prabhu, and Rajesh K. Chandy (2015), "Indirect Learning: How

Emerging-Market Firms Grow in Developed Markets," Journal of Marketing, 79 (1), 10–28.

Batra, Rajeev, Venkatram Ramaswamy, Dana L. Alden, Jan-Benedict E. M. Steenkamp, and S.

Ramachander (2000), "Effects of Brand Local and Nonlocal Origin on Consumer Attitudes

in Developing Countries," Journal of Consumer Psychology, 9 (2), 83–95.

Berens, G., C. B. M. van Riel, and G. H. van Bruggen (2005), "Corporate Associations and

Consumer Product Responses: The Moderating Role of Corporate Brand Dominance,"

Journal of Marketing, 69 (3), 35–48.

Brewer, M. B. (1991), "The Social Self: On Being the Same and Different at the Same Time,"

Personality and Social Psychology Bulletin, 17 (5), 475–82.

Burgess, Steven Michael and Jan-Benedict E. M. Steenkamp (2006), "Marketing Renaissance:

How Research in Emerging Markets Advances Marketing Science and Practice,"

International Journal of Research in Marketing, 23 (4), 337–56.

Castiglioni, Laura, Klaus Pforr, and Ulrich Krieger (2008), "The Effect of Incentives on

Response Rates and Panel Attrition: Results of a Controlled Experiment," Survey Research

Methods, 2 (3), 151–58.

Chang, Sea-Jin, Arjen van Witteloostuijn, and Lorraine Eden (2010), "From The Editors:

Common Method Variance in International Business Research," Journal of International

Business Studies, 41 (2), 178–84.

Cioletti, Amanda (2013), "Always Coca-Cola," License Global, 3 (16), 100–01.

Cleveland, Mark and Michel Laroche (2007), "Acculturaton to the Global Consumer Culture:
 37

Scale Development and Research Paradigm," Journal of Business Research, 60 (3), 249–59.

Coulter, Robin A., Linda L. Price, and Lawrence Feick (2003), "Rethinking the Origins of

Involvement and Brand Commitment: Insights from Postsocialist Central Europe," Journal

of Consumer Research, 30 (2), 151–69.

Dimofte, Claudiu V., Johny K. Johansson, and Richard P. Bagozzi (2010), "Global Brands in

the United States: How Consumer Ethnicity Mediates the Global Brand Effect," Journal of

International Marketing, 18 (3), 81–106.

Dimofte, Claudiu V., Johny K. Johansson, and Ilkka A. Ronkainen (2008), "Cognitive and

Affective Reactions of U.S. Consumers to Global Brands," Journal of International

Marketing, 16 (4), 113–35.

Eckhardt, Giana M. (2005), "Local Branding in a Foreign Product Category in an Emerging

Market," Journal of International Marketing, 13 (4), 57–79.

Feldman, Jack M. and John G. Lynch (1988), "Self-Generated Validity and Other Effects of

Measurement on Belief, Attitude, Intention, and Behavior," Journal of Applied Psychology,

73 (3), 421–35.

Fornell, Claes and David F. Larcker (1981), "Evaluating Structural Equation Models with

Unobservable Variables and Measurement Error," Journal of Marketing Research, 39–50.

Ger, Güliz (1999), "Localizing in the Global Village: Local Firms Competing in Global

Markets," California Management Review, 41 (4), 64–83.

GlobalData (2010), "Top Pharma List," GlobalData.

Guo, Xiaoling (2013), "Living in a Global World: Influence of Consumer Global Orientation on

Attitudes Toward Global Brands From Developed Versus Emerging Countries," Journal of

International Marketing, 21 (1), 1–22.

Guzmán, Francisco and Audhesh K. Paswan (2009), "Cultural Brands From Emerging Markets:

Brand Image Across Host and Home Countries," Journal of International Marketing, 17 (3),

71–86.
 38

Hair, Joseph F., William C. Black, Barry J. Babin, and Rolph E. Anderson (2014), Multivariate

Data Analysis (7 ed.). Upper Saddle River: Pearson.

Hausman, Jerry A. (1978), "Specification Tests in Econometrics," Econometrica: Journal of the

Econometric Society, 46 (6), 1251–71.

He, Jiaxun and Cheng Lu Wang (2015), "Cultural Identity and Consumer Ethnocentrism

Impacts on Preference and Purchase of Domestic Versus Import Brands: An Empirical Study

in China," Journal of Business Research, 68 (6), 1225–33.

Heine, Steven J. and Darrin R. Lehman (1995), "Social Desirability Among Canadian and

Japanese Students," Journal of Social Psychology, 135 (6), 777–79.

Higgins, E. Tory (1996), "Knowledge Activation: Accessibility, Applicability, and Salience," in

Social Psychology, E. Tory Higgins and Arie W. Kruglanski, eds. New York: Guilford Press.

Holt, Douglas, John. A. Quelch, and Earl L. Taylor (2004), "How Global Brands Compete,"

Harvard Business Review, 82 (9), 68–75.

Hult, G. T. M., David J. Ketchen, David A. Griffith, Brian R. Chabowski, Mary K. Hamman,

Bernadine J. Dykes, Wesley A. Pollitte, and S. T. Cavusgil (2008), "An Assessment of the

Measurement of Performance in International Business Research," Journal of International

Business Studies, 39 (6), 1064-80.

Interbrand (2014), "Best Global Brands," (accessed October 13, 2015), [available at

http://interbrand.com/best-brands/best-global-brands/2014/ranking/].

Irwin, Julie R. and Gary H. McClelland (2003), "Negative Consequences of Dichotomizing

Continuous Predictor Variables," Journal of Marketing Research, 40 (3), 366–71.

Jobber, David, John Saunders, and Vince-Wayne Mitchell (2004), "Prepaid Monetary Incentive

Effects on Mail Survey Response," Journal of Business Research, 57 (1), 21–25.

Johansson, Johny K. and Ilkka A. Ronkainen (2005), "The Esteem of Global Brands," Journal

of Brand Management, 12 (5), 339.

Kaminska, Olena, Allan L. McCutcheon, and Jaak Billiet (2010), "Satisficing Among Reluctant
 39

Respondents in a Cross-National Context," Public Opinion Quarterly, 74 (5), 956–84.

Keller, Kevin L. (1993), "Conceptualizing, Measuring, and Managing Customer-Based Brand

Equity," Journal of Marketing, 57 (1), 1–22.

---- (1998), Strategic Brand Management: Building, Measuring, and Managing Brand Equity.

Upper Saddle River, NJ: Prentice-Hall.

Klein, Jill Gabrielle (2002), "Us versus Them, or Us versus Everyone? Delineating Consumer

Aversion to Foreign Goods," Journal of International Business Studies, 345-63.

Kumar, Nirmalya and Jan-Benedict E. M. Steenkamp (2013), Brand Breakout: How Emerging

Market Brands Will Go Global. Houndmills: Macmillan.

Kwak, Hyokjin, Anupam Jaju, and Trina Larsen (2006), "Consumer Ethnocentrism Offline and

Online: The Mediating Role of Marketing Efforts and Personality Traits in the United States,

South Korea, and India," Journal of the Academy of Marketing Science, 34 (3), 367.

Lafley, A. G. (2009), "What Only the CEO Can Do," Harvard Business Review, 87 (5), 54–62.

Leisinger, Klaus M. (2005), "The Corporate Social Responsibility of the Pharmaceutical

Industry: Idealism Without Illusion and Realism Without Resignation," Business Ethics

Quarterly, 15 (4), 577–94.

Little, Todd D., James A. Bovaird, and Keith F. Widaman (2006), "On The Merits of

Orthogonalizing Powered and Product Terms: Implications for Modeling Interactions

Among Latent Variables," Structural Equation Modeling: A Multidisciplinary Journal, 13

(4), 497–519.

Lynch, John G., Jr., Howard Marmorstein, and Michael F. Weigold (1988), "Choices from Sets

Including Remembered Brands: Use of Recalled Attributes and Prior Overall Evaluations,"

Journal of Consumer Research, 15 (2), 169–84.

Madden, Thomas J., Martin S. Roth, and William R. Dillon (2012), "Global Product Quality

and Corporate Social Responsibility Perceptions: A Cross-National Study of Halo Effects,"

Journal of International Marketing, 20 (1), 42–57.

 40

Malär, Lucia, Harley Krohmer, Wayne D. Hoyer, and Bettina Nyffenegger (2011), "Emotional

Brand Attachment and Brand Personality: The Relative Importance of the Actual and the

Ideal Self," Journal of Marketing, 75 (4), 35–52.

Mascarenhas, Briance (2009), "The Emerging CEO Agenda," Journal of International

Management, 15 (3), 245–50.

Mercer, Andrew, Andrew Caporaso, David Cantor, and Reanne Townsend (2015), "How Much

Gets You How Much? Monetary Incentives and Response Rates in Household Surveys,"

Public Opinion Quarterly, 79 (1), 105–29.

Molenberghs, Geert and Geert Verbeke (2007), "Likelihood Ratio, Score, and Wald Tests in a

Constrained Parameter Space," The American Statistician, 61 (1), 22–27.

Morgeson III, Forrest V., Sunil Mithas, Timothy L. Keiningham, and Lerzan Aksoy (2011),

"An Investigation of the Cross-National Determinants of Customer Satisfaction," Journal of

the Academy of Marketing Science, 39 (2), 198–215.

Nijssen, Edwin J. and Susan P. Douglas (2011), "Consumer World-Mindedness and Attitudes

Toward Product Positioning in Advertising: An Examination of Global Versus Foreign

Versus Local Positioning," Journal of International Marketing, 19 (3), 113–33.

Nunnally, Jum C. (1978), Psychometric Theory. New York: McGraw-Hill.

Oberecker, Eva M., Petra Riefler, and Adamantios Diamantopoulos (2008), "The Consumer

Affinity Construct: Conceptualization, Qualitative Investigation, and Research Agenda,"

Journal of International Marketing, 16 (3), 23–56.

Oliver, Richard L. (2015), Satisfaction: A Behavioral Perspective On The Consumer (2 ed.).

New York, NY: Routledge.

---- (1999), "Whence Consumer Loyalty?," Journal of Marketing, 63 (4), 33–43.

Oppenheimer, Daniel M., Tom Meyvis, and Nicolas Davidenko (2009), "Instructional

Manipulation Checks: Detecting Satisficing to Increase Statistical Power," Journal of

Experimental Social Psychology, 45 (4), 867–72.

 41

Özsomer, A. (2012), "The Interplay Between Global and Local Brands: A Closer Look at

Perceived Brand Globalness and Local Iconness," Journal of International Marketing, 20

(2), 72–95.

Pan, Yue and George M. Zinkhan (2006), "Determinants of Retail Patronage: A Meta-

Analytical Perspective," Journal of Retailing, 82 (3), 229–43.

Papadopoulos, Nicolas, Louise A Heslop, and Gary Bamossy (1990), "A Comparative Image

Analysis of Domestic Versus Imported Products," International Journal of Research in

Marketing, 7 (4), 283-94.

Phinney, Jean S. (1996), "When We Talk About American Ethnic Groups, What Do We

Mean?," American Psychologist, 51 (9), 918.

Riefler, Petra (2012), "Why Consumers Do (Not) Like Global Brands: The Role of

Globalization Attitude, GCO and Global Brand Origin," International Journal of Research

in Marketing, 29 (1), 25–34.

Rindfleisch, Aric, James E. Burroughs, and Nancy Wong (2009), "The Safety of Objects:

Materialism, Existential Insecurity, and Brand Connection," Journal of Consumer Research,

36 (1), 1-16.

Roth, K. P. and A. Diamantopoulos (2009), "Advancing The Country Image Construct,"

Journal of Business Research, 62 (7), 726–40.

Satorra, Albert and Peter M. Bentler (2001), "A Scaled Difference Chi-Square Test Statistic For

Moment Structure Analysis," Psychometrika, 66 (4), 507–14.

Schuiling, Isabelle and Jean-Noël Kapferer (2004), "Real Differences Between Local and

International Brands: Strategic Implications for International Marketers," Journal of

International Marketing, 12 (4), 97–112.

Sharma, Piyush (2015), "Consumer Ethnocentrism: Reconceptualization and Cross-Cultural

Validation," Journal of International Business Studies, 46 (3), 381–89.

---- (2011), "Country of Origin Effects in Developed and Emerging Markets: Exploring the
 42

Contrasting Roles of Materialism and Value Consciousness," Journal of International

Business Studies, 42 (2), 285–306.

Sheth, Jagdish N. (2011), "Impact of Emerging Markets on Marketing: Rethinking Existing

Perspectives And Practices," Journal of Marketing, 75 (4), 166–82.

Shimp, Terence A. and Subhash Sharma (1987), "Consumer Ethnocentrism: Construction and

Validation of the CETSCALE," Journal of Marketing Research, 24 (3), 280–89.

Sichtmann, Christina and Adamantios Diamantopoulos (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.

Steenkamp, Jan-Benedict E. M., Rajeev Batra, and Dana L. Alden (2003), "How Perceived

Brand Globalness Creates Brand Value," Journal of International Business Studies, 34 (1),

53–65.

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.

Steenkamp, Jan-Benedict E. M. and Martijn G. de Jong (2010), "A Global Investigation Into the

Constellation of Consumer Attitudes Toward Global and Local Products," Journal of

Marketing, 74 (6), 18–40.

Steinmetz, Holger, Eldad Davidov, and Peter Schmidt (2011), "Three Approaches to Estimate

Latent Interaction Effects: Intention and Perceived Behavioral Control in the Theory of

Planned Behavior," Methodological Innovations Online, 6 (1), 95–110.

Strizhakova, Yuliya and Robin A. Coulter (2015), "Drivers of Local Relative to Global Brand

Purchases: A Contingency Approach," Journal of International Marketing, 23 (1), 1–22.

Strizhakova, Yuliya, Robin A. Coulter, and Linda L. Price (2011), "Branding in a Global

Marketplace: The Mediating Effects of Quality and Self-Identity Brand Signals,"

International Journal of Research in Marketing, 28 (4), 342–51.

 43

Supphellen, Magne and Kjell Grønhaug (2003), "Building Foreign Brand Personalities in

Russia: The Moderating Effect of Consumer Ethnocentrism," International Journal of

Advertising, 22 (2), 203–26.

Sweeney, Jillian C. and Geoffrey N. Soutar (2001), "Consumer Perceived Value: The

Development of a Multiple Item Scale," Journal of Retailing, 77 (2), 203–20.

Swoboda, Bernhard, Bettina Berg, and Hanna Schramm-Klein (2013), "Reciprocal Effects of

the Corporate Reputation and Store Equity of Retailers," Journal of Retailing, 89 (4), 447–

59.

Swoboda, Bernhard and Stefan Elsner (2013), "Transferring the Retail Format Successfully into

Foreign Countries," Journal of International Marketing, 21 (1), 81–109.

Swoboda, Bernhard, Karin Pennemann, and Markus Taube (2012), "The Effects of Perceived

Brand Globalness and Perceived Brand Localness in China: Empirical Evidence on Western,

Asian, and Domestic Retailers," Journal of International Marketing, 20 (4), 72–95.

Swoboda, Bernhard, Cathrin Puchert, and Dirk Morschett (2015), "Explaining the Differing

Effects of Corporate Reputation Across Nations: A Multilevel Analysis," Journal of the

Academy of Marketing Science, 1–20.

Tourangeau, Roger (2004), "Experimental Design Considerations for Testing and Evaluating

Questionnaires," in Methods for Testing and Evaluating Survey Questionnaires, Stanley

Presser and Jennifer M. Rothger and Mick P. Couper and Judith T. Lessler and Elizabeth

Martin and Jean Martin and Eleanor Singer, eds. Hoboken, NJ: John Wiley & Sons.

United Nations (2014), "GDP and Its Breakdown at Current Prices in US Dollars," (accessed

13/04/2015, [available at http://unstats.un.org/unsd/snaama/dnllist.asp].

Verlegh, P. W. J. (2007), "Home Country Bias in Product Evaluation: The Complementary

Roles of Economic and Socio-Psychological Motives," Journal of International Business

Studies, 38 (3), 361–73.

Verplanken, Bas, Godelieve Hofstee, and Heidi J. W. Janssen (1998), "Accessibility of

 44

Affective Versus Cognitive Components of Attitudes," European Journal of Social

Psychology, 28 (1), 23–35.

Walker, Beth A. and Jerry C. Olson (1991), "Means-End Chains: Connecting Products With

Self," Journal of Business Research, 22 (2), 111–18.

Wheeler, S. Christian, Richard E. Petty, and George Y. Bizer (2005), "Self-Schema Matching

and Attitude Change: Situational and Dispositional Determinants of Message Elaboration,"

Journal of Consumer Research, 31 (4), 787–97.

Williams, Larry J., Nathan Hartman, and Flavia Cavazotte (2010), "Method Variance and

Marker Variables: A Review and Comprehensive CFA Marker Technique," Organizational

Research Methods, 13 (3), 477–514.

Winit, Warat, Gary Gregory, Mark Cleveland, and Peeter Verlegh (2014), "Global vs Local

Brands: How Home Country Bias and Price Differences Impact Brand Evaluations,"

International Marketing Review, 31 (2), 102–28.

Xie, Yi, Rajeev Batra, and Siqing Peng (2015), "An Extended Model of Preference Formation

Between Global and Local Brands: The Roles of Identity Expressiveness, Trust, and Affect,"

Journal of International Marketing, 23 (1), 50–71.

Zhang, Yinlong and Adwait Khare (2009), "The Impact of Accessible Identities on the

Evaluation of Global versus Local Products," Journal of Consumer Research, 36 (3), 524–37.

Zhao, Xinshu, John G. Lynch Jr., and Qimei Chen (2010), "Reconsidering Baron and Kenny:

Myths and Truths about Mediation Analysis," Journal of Consumer Research, 37 (2), 197–

206.

Zhou, Lianxi, Zhiyong Yang, and Michael K. Hui (2010), "Non-Local or Local Brands? A

Multi-Level Investigation into Confidence in Brand Origin Identification and its Strategic

Implications," Journal of the Academy of Marketing Science, 38 (2), 202–18.

Zhou, Nan and Russell W. Belk (2004), "Chinese Consumer Readings of Global and Local

Advertising Appeals," Journal of Advertising, 33 (3), 63–76.

 45

NOTES

1. We refer to a few studies, especially that of Swoboda, Pennemann, and Taube (2012), and we

refer to the differences between internationalizing retail firms and manufacturing firms (e.g.,

Swoboda and Elsner 2013) and to the differences between PBG effects in both contexts. In

doing so, we seek to address the distinct retail contexts in which corporate brands might be

conceptually similar for retail firms, chains, or stores (not within diversified retailers operating

through different branded channels). In the retailing context, consumer perceptions are known

to be mostly affected by the retail store and the functional and psychological value that it of-

fers (stronger than by corporate issues, e.g., reputation, Swoboda, Berg, and Schramm-Klein

2013). Therefore, store patronage is used as an outcome in retail research (e.g., Pan and

Zinkhan 2006; Swoboda, Pennemann, and Taube 2012; differently to loyalty intentions).

2. We cooperated with an MNC, identified the six strongest competing MNCs in each of the ten

countries and then successively chose the most economically important countries in which

two domestic MNCs and two foreign MNCs were present (U.S., Japan, and India). Two addi-

tional countries were used for control purposes (China, with one domestic MNC, and Italy, as

an economically weaker country; see the sample selection section for details). We

acknowledge that these countries differ in terms of a variety of economic, cultural, and other

characteristics that we do not cover in this study.

3. As previously mentioned, scholars only address ethnocentrism as a moderator in the product

context. Two studies examine PBG effects for less/more ethnocentric groups (Steenkamp, Ba-

tra, and Alden 2003; Akram, Merunka, and Akram 2011), and four studies investigate differ-

ent topics (e.g., Nijssen and Douglas 2011; Sharma 2011). In addition, some studies find in-

conclusive direct effects of ethnocentrism on brand attitudes (e.g., Guo 2013; Alden,

Steenkamp, and Batra 2006). We theoretically argue that MNCs’ PBG is not naturally diag-

nostic information for loyalty decisions regarding corporate brands, and the literature on prod-

uct brands does not find any direct effects of PBG on the purchase intentions of consumers
 46

who are less (or more) ethnocentric (e.g., Steenkamp, Batra, and Alden 2003). By contrast,

studies on corporate brands show mostly insignificant results when analyzing consumer iden-

tity as a moderator (for foreign vs. local-only firms, Swoboda, Pennemann, and Taube 2012).

4. Cash rewards are common in panel studies, but they may bias the results because of consumer

materialism, for example, what we could not control across nations and what limits this study

(e.g., by using treatments, control groups, or eliminating satisficers, Kaminska, McCutcheon,

and Billiet 2010). However, they offer paramount advantages in international studies. Cash

rewards are important for quota sampling and comparable samples across nations, i.e., they

are usually utilized in panel surveys to avoid panel attrition and to retain respondents over

time (e.g., Castiglioni, Pforr, and Krieger 2008), and they increase the response rate (com-

pared with lottery prizes or charitable donations; Mercer et al. 2015), which is crucial, espe-

cially in large-scale surveys in leading countries. Cash rewards build a psychological obliga-

tion to return the favor, which is accomplished by completing the questionnaire. Respondent

compliance might also increase because of uncomfortable levels of cognitive dissonance cre-

ated by accepting the incentive without responding seriously (Jobber, Saunders, and Mitchell

2004). Additionally, because of the random selection of the evaluated MNCs by each re-

spondent, we assume that the probability of biases resulting from cash incentives is reduced in

this study.

5. The means of PBG are relatively lower while the variances of PBG are greater (2.9-3.4, 1.1-

1.4) in India and Japan than in the U.S. (4.1, 1.0) (see Table 2). This observation might be re-

lated to the tendency of Asian cultures (e.g., Japanese, Chinese, and possibly Indian) to be

prone to moderacy responses or social desirability biases (e.g., Heine and Lehman 1995;

Steenkamp and de Jong 2010), which lead to more central means. The variance differences

have been captured by measurement invariance. An additional test between foreign vs. do-

mestic MNCs (including alternative MNCs) shows, perhaps unsurprisingly (Batra et al. 2000),

that domestic MNCs’ PBG is lower than that of foreign MNCs in all contexts and that it is
 47

lower in Japan and India than in the U.S. The latter may be related to the longstanding partici-

pation of U.S. consumers in the global marketplace and the large number of U.S.-based global

corporate brands (please see web appendix A for details).

 48

Table 1
Literature Review on Empirical Studies on Perceived Brand Globalness and Similar Con-
structs
Studies on Studies on similar constructs
perceived brand globalness (e.g., global brand attitude, global consumption orientation)
One country -- - Alden, Steenkamp, and Batra (2006) d
- developed - Dimofte, Johansson, and Bagozzi (2010) d
- Dimofte, Johansson, and Ronkainen (2008)
- Nijssen and Douglas (2011)
- Riefler (2012) 2, a
- emerging - Akram, Merunka, and Akram (2011) a - Batra et al. (2000)
- Swoboda, Pennemann, and Taube (2012) 1, c - Zhou, Yang, and Hui (2010) 1, b, c
- Winit et al. (2014) 1/2, b
- Xie, Batra, and Peng (2015)
Several countries (pooled) - Holt, Quelch, and Taylor (2004) - Schuiling and Kapferer (2004)
- Steenkamp and de Jong (2010) a
Comparing countries - Johansson and Ronkainen (2005) - Papadopoulos, Heslop, and Bamossy (1990)
- only developed - Steenkamp, Batra, and Alden (2003)

- also emerging - Sichtmann and Diamantopoulos (2013) b/d - Alden et al. (2013) d
- This study 2, c - Guo (2013) d
- Sharma (2011)
- Strizhakova, Coulter, and Price (2011)
- Strizhakova and Coulter (2015) a, d
Note: Studies in bold analyze corporate or retail brands (further analyze product brands). Studies in italics analyze consumer ethnocentrism as modera-
tor.
1
Studies compare domestic-only vs. foreign global brands or 2 domestic global vs. foreign global brands (further without comparison). Theory: a con-
sumer culture, b signaling, c accessibility-diagnosticity, d e.g., information processing, categorization, acculturation etc. (other studies without explicit
theory).

Table 2
Sample Characteristics
India (N=704) Japan (N=794) U.S. (N=662)
in % Male Female Total Male Female Total Male Female Total
Age 18-25 15.9 10.0 25.9 5.9 4.4 10.3 7.3 7.7 14.9
Age 26-35 13.1 15.8 28.8 10.3 8.4 18.7 11.3 10.8 22.0
Age 36-45 12.1 9.2 21.4 12.5 11.8 24.3 13.7 12.2 25.9
Age 46-55 7.9 7.2 15.1 12.5 9.7 22.3 10.8 10.4 21.1
Age 56-65 4.3 4.5 8.8 13.0 11.4 24.4 9.1 6.9 16.0
Total 53.3 46.7 100.0 54.3 45.7 100.0 52.1 47.9 100.0
 49

Table 3
Reliability and Validity of Measurements
India Japan U.S.
Item MV/Std FL KMO ItTC α CR λ MV/Std FL KMO ItTC α CR λ MV/Std FL KMO ItTC α CR λ
Perceived brand globalness
To me, [MNC] is a global brand. 3.3/1.3 .865 .831 .863 2.9/1.4 .753 .740 .759 4.1/1.0 .906 .881 .908
I do think consumers abroad buy 3.3/1.2 .916 .866 .922 3.2/1.1 .955 .873 .959 4.1/1.0 .948 .911 .948
[MNC’s] products. .759 .930 .931 .701 .911 .913 .771 .953 .953
[MNC] sells its products all over the 3.4/1.3 .935 .879 .931 3.2/1.1 .977 .887 .972 4.1/1.0 .946 .910 .946
world.
Functional value
[MNC] offers high-quality products and 3.7/1.1 .903 .870 .907 3.0/0.5 .844 .799 .855 3.7/0.9 .892 .840 .906
services.
[MNC] develops well-made innovative 3.6/1.1 .921 .885 .919 3.1/0.5 .899 .845 .900 3.6/0.9 .879 .830 .881
products and services. .871 .945 .945 .853 .918 .918 .845 .921 .921
With its products, [MNC] is a leader in 3.6/1.1 .886 .855 .887 3.2/0.6 .838 .796 .839 3.5/1.0 .895 .847 .880
research and technology.
[MNC] offers good value for money. 3.6/1.0 .894 .862 .890 3.1/0.5 .859 .813 .845 3.3/1.0 .800 .767 .797
Psychological value
[MNC] offers products I like and 3.4/1.1 .879 .834 .878 2.9/0.6 .841 .771 .831 3.8/1.0 .767 .727 .760
enjoy.
The appealing image of [MNC’s] prod- 3.2/1.1 .827 .791 .810 2.9/0.7 .737 .686 .720 3.4/1.0 .807 .761 .808
ucts makes me feel good.
[MNC’s] products improve the way I 3.4/1.1 .887 .860 .840 .925 .924 .890 3.1/0.7 .818 .819 .748 .879 .879 .819 3.8/0.9 .912 .823 .846 .905 .905 .900
am perceived.
With [MNC’s] products I make a better 3.5/1.0 .887 .841 .897 3.1/0.7 .828 .759 .847 3.5/0.9 .885 .823 .900
impression on other people than with
those of competitors.
Loyalty
I am a loyal customer of [MNC]. 3.6/1.1 .940 .902 .933 3.2/0.7 .857 .802 .859 2.9/1.2 .962 .929 .963

I have developed a good relationship 3.5/1.1 .930 .894 .925 3.0/0.6 .839 .789 .862 2.9/1.1 .892 .877 .895
.774 .949 .949 .744 .906 .907 .754 .960 .962
with [MNC].
I am certain that I will buy products/of- 3.6/1.1 .915 .884 .927 3.1/0.6 .929 .851 .908 2.8/1.1 .977 .941 .975
fers of [MNC].
Consumer ethnocentrism
Purchasing foreign-made products is un- 2.9/1.3 .886 .826 .887 2.5/0.9 .892 .840 .893 2.5/1.2 .896 .855 .896
[nationality].
[Nationality] should not buy foreign 3.0/1.3 .862 .809 .851 2.5/0.9 .894 .843 .894 2.7/1.2 .890 .850 .886
products because this hurts local busi-
ness/causes unemployment. .829 .907 .909 .856 .920 .920 .851 .934 .934
A real [nationality] should always buy 3.1/1.3 .766 .728 .768 2.6/1.0 .813 .776 .812 2.6/1.2 .873 .835 .874
[nationality]-made products.
It is not right to purchase foreign prod- 2.8/1.3 .859 .801 .866 2.3/0.9 .853 .808 .852 2.4/1.2 .870 .833 .873
ucts.
Model fits CFA: India: CFI .992; TLI .991; RMSEA .030; SRMR .033; ²(125) = 210.072; SCF = 1.2778; Japan: CFI .979; TLI .975; RMSEA .044; SRMR .031; ²(125) = 336.791; SCF = 1.2987; U.S.: CFI .975; TLI .969;
RMSEA .056; SRMR .037; ²(125) = 429.081; SCF = 1.1842.
Note: MV/Std = Mean value/standard deviation; FL = Factor loadings; KMO = Kaiser-Meyer-Olkin criterion (≥ .5); ItTC = Item-to-total correlation (≥ .5); α = Cronbach’s alpha (≥ .7); CR = Composite reliability (≥ .6); λ =
Standardized factor loadings (≥ .5).
 50

Table 4
Discriminant Validity
Constructs PBG FV PV LOY
PBG .776
FV .445 .825
India
PV .433 .741 .774
LOY .379 .780 .812a .861
PBG .767
FV .100 .943
Japan
PV .104 .643 .870
LOY .062 .615 .723 .910
PBG .899
FV .235 .814
U.S.
PV .252 .766 .805
LOY .087 .382 .434 .872
a
For this case we computed a chi-square difference test (Anderson and Gerbing 1988), which showed satisfactory results and thus discriminant valid-
ity can be assumed.
Note: Values in bold represent the AVE of the construct; values in italics represent squared correlations; AVE = Average variance extracted (≥ .5);
PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty.

Table 5
Measurement Invariance
² diff. CFI TLI RMSEA
Model ² (d.f.) SCF (p-value) (ΔCFI) (ΔTLI) (ΔRMSEA) SRMR
Invariance: MNC origin
Model 1: 168.369 1.3443 - .997 .996 .022 .019
Configural invariance (142) (-) (-) (-)
Model 2: 181.271 1.2932 12.902 .997 .997 .021 .040
Metric invariance (156) (.730) (.000) (-.001) (.001)
India
Invariance: Consumer ethnocentrism
Model 1: 179.748 1.3336 - .996 .995 .027 .019
Configural invariance (142) (-) (-) (-)
Model 2: 195.536 1.2821 15.788 .996 .995 .026 .070
Metric invariance (156) (.416) (.000) (.000) (.001)
Invariance: MNC origin
Model 1: 300.074 1.3502 - .979 .972 .050 .032
Configural invariance (142) (-) (-) (-)
Model 2: 311.813 1.3455 11.739 .978 .974 .048 .053
Metric invariance (156) (.679) (.001) (-.002) (.002)
Japan
Invariance: Consumer ethnocentrism
Model 1: 317.058 1.3470 - .977 .970 .053 .033
Configural invariance (142) (-) (-)
Model 2: 333.649 1.3393 16.591 .976 .972 .051 .082
Metric invariance (156) (.333) (.001) (-.002) (.002)
Invariance: MNC origin
Model 1: 410.965 1.1726 - .973 .965 .070 .043
Configural invariance (142) (-) (-) (-)
Model 2: 436.697 1.1530 25.732 .972 .967 .068 .053
Metric invariance (156) (.066) (.001) (-.002) (.002)
U.S.
Invariance: Consumer ethnocentrism
Model 1: 362.149 1.1895 - .977 .971 .063 .038
Configural invariance (142) (-) (-) (-)
Model 2: 385.253 1.1678 23.104 .976 .972 .062 .062
Metric invariance (156) (.125) (.001) (-.001) (.001)
 51
Table 6
Results
India Japan U.S.
1. General model Total sample Rival model (direct path) Total sample Rival model (direct path) Total sample Rival model (direct path)
Effects beta bp beta bp beta bp beta bp beta bp beta bp
PBG→FV .697 .590*** .702 .594*** .314 .128*** .317 .130*** .499 .459*** .500 .460***
PBG→PV .691 .583*** .697 .589*** .313 .164*** .322 .169*** .517 .455*** .520 .458***
FV→LOY .512 .498*** .552 .532*** .430 .508*** .443 .520*** .271 .356*** .287 .374***
PV→LOY .545 .531*** .585 .566*** .662 .610*** .678 .622*** .479 .656*** .511 .696***
Gender -.011 -.020ns -.011 -.020ns .045 .044ns .042 .041ns -.032 -.068ns -.036 -.077ns
Age .036 .026ns .038 .028ns .032 .012ns .040 .016ns .014 .011ns .021 .017ns
Familiarity .004 .002ns .001 .001ns .005 .001ns .002 .000ns .062 .038* .057 .035*
PBG→LOY - - -.088 -.072ns - - -.058 -.028ns - - -.074 -.088ns
Total effect
PBG→LOY .733 .604*** .342 .165*** .382 .462***
Indirect effects (H1)
PBG→FV→LOY .357 .294*** .135 .065*** .135 .163***
PBG→PV→LOY .376 .310*** .207 .100*** .247 .299***
2. Moderator: Diff. betw. countries Diff. betw. countries Diff. betw. countries
MNC origin Foreign Domestic Foreign vs. domestic Foreign Domestic Foreign vs. domestic Foreign Domestic Foreign vs. domestic
Effects beta bp beta bp bp beta bp beta bp bp beta bp beta bp bp
PBG→FV .740 .602*** .333 .251*** .351*** .395 .258*** .279 .138*** .120*** .487 .450*** .494 .450*** .000ns
PBG→PV .682 .549*** .648 .536*** .013ns .349 .270*** .331 .234*** .036* .551 .490*** .483 .424*** .066*
FV→LOY .529 .523*** .498 .484*** .039* .577 .649*** .371 .510*** .139*** .482 .655*** .170 .218*** .437***
PV→LOY .533 .532*** .618 .548*** -.016ns .560 .531*** .654 .626*** -.095** .251 .354*** .595 .803*** -.449***
Gender -.001 -.002ns -.027 -.046ns -.008 -.008ns .063 .059ns -.099 -.214* .037 .080ns
Age .009 .006ns .067 .045* .029 .012ns .047 .017ns .068 .056ns -.039 -.033ns
Familiarity .006 .004ns -.004 -.002ns -.009 -.002ns .039 .009ns .040 .024ns .099 .061**
Total effect (H2)
PBG→LOY .755 .607*** .566 .415*** .192*** .423 .311*** .320 .217*** .094*** .373 .469*** .371 .440*** .029ns
Diff. within countries Diff. within countries Diff. within countries
Foreign Domestic Foreign Domestic Foreign Domestic
Indirect effects (H3) bp bp bp bp bp bp
PBG→FV→LOY .392 .315*** .166 .122*** .023ns .228 .167*** .104 .070*** .024ns .235 .295*** .084 .099*** .121***
PBG→PV→LOY .363 .292*** .400 .294*** -.172*** .195 .143*** .217 .147*** -.077*** .138 .174*** .287 .340*** -.241***
3. Moderator: Diff. betw. groups Diff. betw. groups Diff. betw. groups
Consumer
Consumerethnocentrism Lower Higher Lower vs. higher Lower Higher Lower vs. higher Lower Higher Lower vs. higher
Effects beta bp beta bp bp beta bp beta bp bp beta bp beta bp bp
PBG→FV .752 .651*** .687 .563*** .088*** .378 .159*** .208 .078** .081*** .521 .510*** .485 .425*** .085***
PBG→PV .750 .632*** .677 .546*** .086*** .378 .210*** .214 .102*** .108*** .559 .518*** .468 .394*** .124***
FV→LOY .528 .527*** .482 .463*** .064*** .446 .535*** .392 .492*** .043ns .350 .427*** .149 .210*** .217***
PV→LOY .494 .506*** .583 .570*** -.064*** .669 .606*** .628 .615*** -.009ns .421 .542*** .576 .844*** -.302***
Gender -.017 -.033ns .000 -.001ns .002 .002ns .062 .055ns .007 .016ns -.080 -.173*
Age .041 .031ns .024 .017ns .055 .023ns .000 .000ns .016 .013ns .007 .006ns
Familiarity .021 .014ns -.008 -.004ns -.004 -.001ns .029 .007ns .002 .002ns .082 .048*
Total effect (H4)
PBG→LOY .767 .663*** .726 .572*** .091*** .422 .212*** .212 .101*** .111*** .418 .499*** .342 .422*** .077***
Diff. within groups Diff. within groups Diff. within groups
Lower Higher Lower Higher Lower Higher
Indirect effects (H5) bp bp bp bp bp bp
PBG→FV→LOY .397 .343*** .331 .261*** .023ns .169 .085*** .082 .038*** .182 .218*** .072 .089***
PBG→PV→LOY .370 .320*** .395 .311*** -.050** .253 .127*** .134 .063*** -.042** -.025* .235 .281*** .270 .332*** -.063*** -.243***
Model fits: General model: India: CFI .961; TLI .952; RMSEA .069; SRMR .109; ²(112) = 488.376; SCF = 1.2206; Japan: CFI .925; TLI .908; RMSEA .081; SRMR .175; ²(112) = 718.672; SCF = 1.2376; U.S.: CFI .915;
TLI .897; RMSEA .102; SRMR .158; ²(112) = 961.718; SCF = 1.1078. MNC origin: India: CFI .955; TLI .948; RMSEA .071; SRMR .172; ²(234) = 659.837; SCF = 1.2010; Japan: CFI .919; TLI .907; RMSEA .079;
SRMR .171; ²(234) = 881.099; SCF = 1.2419; U.S.: CFI .912; TLI .898; RMSEA .102; SRMR .165; ²(234) = 1120.059; SCF = 1.1071. CE: India: CFI .961; TLI .954; RMSEA .067; SRMR .102; ²(234) = 603.278;
SCF = 1.1999; Japan: CFI .925; TLI .913; RMSEA .078; SRMR .174; ²(234) = 826.986; SCF = 1.2417; U.S.: CFI .917; TLI .903; RMSEA .099; SRMR .162; ²(234) = 1060.998; SCF = 1.1176.
Note: * p < .05; ** p < .01; *** p < .001; ns=not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.
 52

Figure 1
Conceptual Framework
 WEB APPENDIX

]Does Being Perceived as Global Pay Off?

An Analysis of Leading Foreign and Domestic MNCs in India, Japan, and the United

States

Bernhard Swoboda and Johannes Hirschmann

Contents

A. Descriptive Statistics and Additional Measurement Invariance Tests 1

B. Common Method Variance 3

C. Exogeneity Test of Instrumental Variables 5

D. Direct Effect of Perceived Brand Globalness on Consumer Loyalty 6

E. Alternative Models 7

F. Latent Interaction Models of Consumer Ethnocentrism 28

References 36
 1

Web Appendix A. Descriptive Statistics and Additional Measurement Invariance Tests

Table A.1 shows mean values (MV) and standard deviations (Std.) of perceived brand globalness
for all foreign vs. domestic MNCs of both our main models and alternative models.
We also computed further measuring invariance tests. First, we computed measurement invari-
ance across three groups, i.e., India, Japan, and the U.S., for the single constructs (see Table A.2).
Based on the difference in the chi² values metric invariance could not be established, but the chi²
difference test is likely to be biased when sample size is large (the sample size is n > 2200,
compared to invariance tests for the multi-group moderation mod-els where n ≈ 700; Hair et al.
2014) and thus we followed the procedure suggested by Chen (2007) and compared changes in
CFI and RMSEA. Because CFI changed no more than .010 and RMSEA no more than .015, we
assumed metric invariance. Second, we followed the same procedure to compute measurement
invariance across the same three groups for the whole measurement model (see Table A.3).

Table A.1
MV and Std. of PBG for Foreign and Domestic MNCs Within Each Country
India Japan U.S.
Foreign Domestic Foreign Domestic Foreign Domestic
(altern. MNCs) (altern. MNCs) (altern. MNCs)
MV/Std. MV/Std. MV/Std. MV/Std. MV/Std. MV/Std.
Perceived brand globalness 3.5/1.3* 3.0/1.3 3.2/1.5* 2.7/1.3 4.2/1.0 4.0/0.9*
To me, [MNC] is a global brand. (3.4/1.3)* (2.8/1.4)* (3.5/1.1)*
I do think consumers abroad buy 3.5/1.2* 3.0/1.2 3.5/1.2* 2.9/1.0 4.2/1.0 4.0/1.0*
[MNC’s] products. (3.3/1.2)* (3.5/1.1)* (3.5/1.1)*
[MNC] sells its products all over the 3.6/1.2* 3.1/1.3 3.5/1.2* 2.9/1.0 4.2/1.0 4.0/1.0*
world. (3.4/1.2)* (3.3/1.1)* (3.6/1.1)*
China Italy
Foreign Domestic Foreign Domestic
(altern. MNCs) (altern. MNCs)
MV/Std. MV/Std. MV/Std. MV/Std.
Perceived brand globalness 4.2/1.1* 3.6/1.2 3.9/1.1* 3.0/1.1
To me, [MNC] is a global brand. (3.7/1.3) (3.8/1.1)*
I do think consumers abroad buy 4.1/1.1* 3.6/1.1 3.9/1.1* 2.9/1.1
[MNC’s] products. (3.7/1.2) (3.8/1.0)*
[MNC] sells its products all over the 4.2/1.1* 3.6/1.1 4.0/1.1* 3.0/1.1
world. (3.7/1.2) (3.9/1.1)*
Note: * p < .05; MV differences (t-tests).
 2

Table A.2
Measurement Invariance for PBG Across Countries (India, Japan, and U.S.) (single constructs)
² diff. CFI TLI RMSEA
Model ² (d.f.) SCF (p-value) (ΔCFI) (ΔTLI) (ΔRMSEA) SRMR
Invariance: Perceived brand globalness
Model 1: 43.723 .7494 - .994 .975 .161 .012
Configural invariance (2) (-) (-) (-)
Model 2: 51.316 .8649 7.593 .994 .991 .097 .038
Metric invariance (6) (.000) (.000) (-.016) (.064)
Invariance: Functional value
Model 1: 165.422 1.6596 - .976 .945 .156 .093
Configural invariance (8) (-) (-) (-)
Model 2: 285.681 1.3577 120.259 .958 .946 .155 .132
Metric invariance (14) (.000) (.018) (-.001) (.001)
Model 3: 217.629 1.5205 52.207 .962 .942 .161 .121
Partial metric invariancea (10) (.000) (.008) (.003) (-.005)
Invariance: Psychological value
Model 1: 161.505 1.4562 - .971 .935 .154 .068
Configural invariance (8) (-) (-) (-)
Model 2: 271.445 1.2346 109.940 .951 .937 .151 .138
Metric invariance (14) (.000) (.020) (-.002) (.003)
Model 3: 177.650 1.3523 16.145 .968 .943 .144 .098
Partial metric invarianceb (10) (.067) (.003) (-.008) (.010)
Invariance: Loyalty
Model 1: 82.428 .6668 - .992 .964 .224 .074
Configural invariance (2) (-) (-) (-)
Model 2: 94.518 .8437 12.090 .991 .987 .135 .079
Metric invariance (6) (.001) (-.023) (.089)
a
Factor loadings were freed for the following items: FV2, FV4.
b
Factor loadings were freed for the following items: FV2, FV3.

Table A.3
Measurement Invariance for PBG Across Countries (India, Japan, and U.S.)
² diff. CFI TLI RMSEA
Model ² (d.f.) SCF (p-value) (ΔCFI) (ΔTLI) (ΔRMSEA) SRMR
Invariance
Model 1: 686.641 1.2214 - .984 .980 .053 .029
Configural invariance (213) (-) (-) (-)
Model 2: 1085.089 1.1825 398.448 .972 .968 .066 .076
Metric invariance (241) (.000) (.012) (.012) (-.013)
Model 3: 1038.415 1.1886 351.774 .974 .969 .065 .074
Partial metric invariancea (235) (.000) (.010) (.011) (-.012)
a
Factor loadings were freed for the following items: FV2, PV3, LOY2.
 3

Web Appendix B. Common Method Variance

We addressed common-method variance (CMV) a priori by using an appropriate questionnaire. A posteriori we accounted for CMV first by calcu-
lating a single-factor test using confirmatory factor analysis and second by applying the correlational marker technique. The results of the single-
factor test showed that the models with all items loading on a single factor had a significantly worse fit than our proposed models did (see Table B.1).
Tables B.2 to B.4 show the results for the marker variable technique (Lindell and Whitney 2001) following the latent variable approach of Williams,
Hartman, and Cavazotte (2010). We chose to use occupation as the marker variable as this variable is theoretically unrelated to PBG (Rindfleisch,
Burroughs, and Wong 2009). The technique consists of three successive phases. The results of the model comparisons (phase I) point out that the
correlations between the latent constructs are not biased through the presence of the marker variable. The results of the following reliability decom-
position (phase II) indicate that the amount of method variance, associated with the measurement of the substantive latent constructs, is less than
6.00% in India (between 3.26% and 5.88%), less than 8.00% in Japan (between 3.02% and 7.02%), and less than 8.50% in the U.S. (between 4.82%
and 8.50%). As the impact of method variance in the study of Williams, Hartman, and Cavazotte (2010) was above 12.5%, we found that the present
results of below 8.50% overall could be decreased. The results of the sensitivity analysis (phase III) show that marker-based method variance has a
low effect on construct correlations.

Table B.1
Single-factor Test
India Japan U.S.
Model χ² ² diff. χ² ² diff. χ² ² diff.
CFI TLI RMSEA SCF (df) (p-value) CFI TLI RMSEA SCF (df) (p-value) CFI TLI RMSEA SCF (df) (p-value)
CFA .997 .996 .022 1.3494 97.335 .977 .971 .053 1.3509 248.414 .975 .968 .067 1.1770 314.098
(71) (71) (71)
*** *** ***
SFT .833 .803 .162 1.3614 1596.764 .668 .607 .194 1.3879 2656.185 .617 .548 .252 1.2057 3834.570
(77) (77) (77)
Note: * p < .05; ** p < .01; *** p < .001; ns Not significant; CFA = Model fit of the confirmatory factor analysis; SFT = Model fit of the single factor test; SCF = Scaling correction factor for MLM.
 4

Table B.2
Results of Model Comparisons (phase I)
India Japan U.S.
Model ² df SCF CFI TLI RMSEA ² df SCF CFI TLI RMSEA ² df SCF CFI TLI RMSEA
CFA 172.676 85 1.2108 .991 .989 .037 752.280 85 1.3275 .914 .894 .096 435.273 85 1.1511 .964 .956 .074
Baseline 202.362 90 1.2219 .988 .987 .041 769.037 90 1.3074 .913 .898 .094 453.357 90 1.1401 .963 .957 .073
Method-C 195.275 89 1.2252 .989 .987 .040 775.898 89 1.3495 .913 .897 .094 449.333 89 1.1410 .963 .957 .073
Method-U 166.111 76 1.2542 .991 .987 .041 735.097 76 1.3534 .915 .883 .100 403.172 76 1.1671 .967 .954 .076
Method-R 165.065 82 1.2670 .991 .989 .038 726.436 82 1.3777 .917 .894 .096 402.059 82 1.1822 .967 .958 .072
Chi-square differences of model comparison tests
ΔModels Δ² Δdf p Δ² Δdf p Δ² Δdf p
Baseline vs. Method-C 7.087 1 ** 6.861 1 *** 4.024 1 *
Method-C vs. Method-U 29.164 13 ** 40.801 13 *** 46.161 13 ***
Method-U vs. Method-R 1.046 6 ns 8.661 6 ns 1.113 6 ns
Note: * p < .05; ** p < .01; *** p < .001; ns Not significant; SCF = Scaling correction factor for MLM.

Table B.3
Results of the Reliability Decomposition (phase II)
India Japan U.S.
Reliability Reliability Reliability
baseline model Decomposed reliability from method-U model baseline model Decomposed reliability from method-U model baseline model Decomposed reliability from method-U model
Total Substantive re- Method % reliability Total Substantive re- Method % reliability Total Substantive re- Method % reliability
Latent variable reliability liability reliability marker reliability liability reliability marker reliability liability reliability marker
PBG .919 .865 .054 5.88 .928 .900 .028 3.02 .964 .886 .078 8.09
FV .951 .920 .031 3.26 .980 .940 .040 4.08 .943 .897 .046 4.88
PV .928 .897 .031 3.34 .959 .918 .041 4.28 .934 .889 .045 4.82
LOY .944 .890 .054 5.72 .968 .900 .068 7.02 .953 .872 .081 8.50
Note: PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty.

Table B.4
Results of the Sensitivity Analyses (phase III)
India Japan U.S.
CFA Baseline Method-U Method-S Method-S CFA Baseline Method-U Method-S Method-S CFA Baseline Method-U Method-S Method-S
Construct correlations (.05) (.01) (.05) (.01) (.05) (.01)
PBG with FV .666 .666 .668 .665 .665 .474 .474 .475 .484 .486 .578 .578 .585 .580 .581
PBG with PV .637 .637 .638 .635 .634 .479 .479 .481 .479 .481 .594 .594 .603 .596 .596
PBG with LOY .535 .535 .526 .519 .516 .443 .443 .446 .436 .438 .349 .349 .416 .352 .353
PV with FV .891 .891 .891 .892 .892 .942 .942 .942 .930 .930 .907 .907 .906 .907 .907
PV with LOY .820 .820 .825 .825 .825 .952 .952 .952 .939 .939 .647 .647 .681 .648 .648
FV with LOY .715 .715 .723 .720 .720 .935 .935 .935 .951 .951 .614 .614 .653 .615 .616
PBG with marker -.129 .000 .000 .000 .000 -.202 .000 .000 .000 .000 -.250 .000 .000 .000 .000
FV with marker -.092 .000 .000 .000 .000 -.276 .000 .000 .000 .000 -.208 .000 .000 .000 .000
PV with marker -.084 .000 .000 .000 .000 -.259 .000 .000 .000 .000 -.156 .000 .000 .000 .000
LOY with marker -.121 .000 .000 .000 .000 -.248 .000 .000 .000 .000 -.027 .000 .000 .000 .000
Note: PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty.
 5

Web Appendix C. Exogeneity Test of Instrumental Variables

We tested for endogeneity by using the instrumental variable (IV) exposure to the MNCs’ market-
ing, especially their communication activities, which represents a core antecedent of PBG
(Steenkamp, Batra, and Alden 2003). For this purpose, we used one item: ‘When I have read
magazines or watched TV, I have often seen advertising for offers of [MNC]’ (adapted from
Cleveland and Laroche 2007).
We first checked whether exposure to communications of MNCs is a strong predictor of per-
ceived brand globalness in the three countries using F-tests. As the calculated F-values exceeded
the recommended threshold of 10 in all countries (see Table C.1), the IV can be considered to be
a strong predictor (Antonakis et al. 2014). Additionally to our efficient (i.e., proposed) models
we estimated consistent models (see Table C.2) which included the IV (Antonakis et al. 2010)
and verified whether there was a change in path estimates using the Hausman (1978) test. The
respective t-values were below the critical ratio of 1.96, indicating that the probability of endoge-
neity of PBG seems to be reduced.

Table C.1
F-test of Strong Instrument Variables
India Japan U.S.
IV → PBG 259.708 25.723 146.383
Note: IV = Instrumental variable; PBG = Perceived brand globalness; F-value > 10 indicates strong predictor.

Table C.2
Results of the Efficient and Consistent Models
India Japan U.S.
Consistent Proposed / effi- Consistent Proposed / effi- Consistent Proposed / effi-
model cient model model cient model model cient model
Effects beta p beta p beta p beta p beta p beta p
PBG → FV .708 *** .697 *** .311 *** .314 *** .506 *** .499 ***
PBG → PV .705 *** .691 *** .307 *** .313 *** .527 *** .517 ***
FV → LOY .511 *** .512 *** .435 *** .430 *** .271 *** .271 ***
PV → LOY .542 *** .545 *** .652 *** .662 *** .477 *** .479 ***
IV → PBG .833 *** -- .151 *** -- .470 *** --
Gender -.011 ns -.011 ns .029 ns .045 ns -.032 ns -.032 ns
Age .036 * .036 ns .033 ns .032 ns .014 ns .014 ns
Brand familiarity .004 ns .004 ns .009 ns .005 ns .061 * .062 *
Total effect
PBG → LOY .745 *** .733 *** .336 *** .342 *** .388 *** .382 ***
Indirect effects
PBG → FV → LOY .362 *** .357 *** .135 *** .135 *** .137 *** .135 ***
PBG → PV → LOY .383 *** .376 *** .200 *** .207 *** .251 *** .247 ***
Model fits:
India: Consistent model: CFI .955; TLI .947; RMSEA .070; SRMR .107; ²(126) = 587.555; SCF = 1.1954; Proposed model: CFI .961;
TLI .952; RMSEA .069; SRMR .109; ²(112) = 488.376; SCF = 1.2206.
Japan: Consistent model: CFI .909; TLI .893; RMSEA .084; SRMR .094; ²(126) = 894.989; SCF = 1.2110; Proposed model: CFI .925;
TLI .908; RMSEA .081; SRMR .115; ²(112) = 718.672; SCF = 1.2376.
U.S.: Consistent model: CFI .901; TLI .885; RMSEA .101; SRMR .115; ²(126) = 1245.687; SCF = 1.0982; Proposed model: CFI .915;
TLI .897; RMSEA .102; SRMR .118; ²(112) = 961.718; SCF = 1.1078.
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psycholog-
ical value; LOY = Loyalty; IV = Instrumental variable ‘Exposure to marketing activities of MNCs’; SCF = Scaling correction factor for
MLM; b = unstandardized coefficients; beta = standardized coefficients.
 6

Web Appendix D. Direct Effect of Perceived Brand Globalness on Consumer Loyalty

Table D.1
Test for Direct Effect Using the Wald Test
India Japan U.S.
1. General model W df p W df p W df p
PBG → LOY 3.529 1 ns 3.175 1 ns 2.396 1 ns
2. Moderator: Foreign Domestic Foreign Domestic Foreign Domestic
MNC origin W df p W df p W df p W df p W df p W df p
PBG → LOY 1.980 1 ns .894 1 ns .782 1 ns 3.716 1 ns .295 1 ns 2.755 1 ns

3. Moderator: Lower Higher Lower Higher Lower Higher

Consumer ethnocentrism W df p W df p W df p W df p W df p W df p
PBG → LOY 3.300 1 ns 1.502 1 ns 3.441 1 ns 2.679 1 ns .748 1 ns .280 1 ns
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; PBG = Perceived brand globalness; LOY = Loyalty; W = Wald test statistic; df = degrees of freedom.

Table D.2
Test for Direct and Indirect Effects using Bootstrapping
India Japan U.S.
1. General model S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p
Direct effect
PBG → LOY .028 -.072 -.108 .001 ns .015 -.028 -.056 .001 ns .049 -.088 -.183 .008 ns
Indirect effects
PBG → FV → LOY .033 .316 .254 .384 *** .013 .068 .045 .094 *** .041 .172 .094 .258 ***
PBG → PV → LOY .035 .333 .271 .406 *** .018 .105 .072 .142 *** .050 .318 .224 .422 ***
2. Moderator:
MNC origin Foreign Domestic Foreign Domestic Foreign Domestic
S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p
Direct effect
PBG → LOY .037 -.043 -.112 .028 ns .035 -.033 -.097 .038 ns .030 -.021 -.081 .038 ns .026 -.055 -.106 .001 ns .070 -.042 -.184 .092 ns .064 -.089 -.217 .033 ns
Indirect effects
PBG → FV → LOY .016 .265 .235 .295 *** .017 .122 .084 .151 *** .026 .168 .118 .219 *** .019 .074 .039 .112 *** .065 .298 .177 .433 *** .009 .100 .081 .118 ***
PBG → PV → LOY .028 .347 .295 .398 *** .044 .303 .214 .396 *** .012 .147 .123 .170 *** .017 .151 .117 .181 *** .071 .186 .053 .333 ** .051 .365 .269 .467 ***
3. Moderator:
Consumer Lower Higher Lower Higher Lower Higher
ethnocentrism S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p
Direct effect
PBG → LOY .042 -.083 -.169 .006 ns .034 -.060 -.136 .001 ns .025 -.042 -.106 .001 ns .024 -.034 -.082 .020 ns .070 -.070 -.214 .063 ns .074 -.043 -.191 .099 ns
Indirect effects
PBG → FV → LOY .029 .246 .190 .307 *** .036 .275 .203 .335 *** .013 .089 .065 .110 *** .015 .040 .009 .067 ** .059 .225 .119 .332 *** .049 .095 .006 .193 *
PBG → PV → LOY .044 .317 .228 .395 *** .050 .326 .235 .435 *** .023 .136 .090 .182 *** .022 .068 .030 .106 ** .067 .300 .168 .443 *** .066 .339 .214 .468 ***
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty; S.E. = standard error; b = unstandardized coefficients;
Lower: Lower limit of confidence interval; Upper: Upper limit of confidence interval; 95% confidence interval limits are shown; bootstrap samples = 5,000.
 7

Web Appendix E. Alternative Models

Subsequently the results of the alternative models are presented after sample characteristics and tests:
 Sample Characteristics (see Tables E.1a-c) 8
 Reliability and Validity (see Tables E.2a-c) 9
 Discriminant Validity (see Tables E.3a-b) 12
 Measurement Invariance (see Tables E.4a-b) 12
 Common Method Variance (see Tables E.5a-l) 14
 Exogeneity Test of Instrumental Variables (see Tables E.6a-d) 18
 Results (see Tables E.7a-c) 20
 Test for Direct Effect using the Wald Test (see Tables E.8a-c) 23
 Test for Direct and Indirect Effects using Bootstrapping (see Tables E.9a-c) 25

First, the hypotheses were tested in a second emerging country (China) and in a second developed country (Italy). The results are stable compared to
those for India and the U.S. (see Table E.7a).
Second, hypotheses were tested replacing the German-speaking foreign MNCs with the remaining foreign competitors in all countries (India, Japan,
China, Italy)—except in the U.S., where only domestic MNCs were present among the major competitors and where the domestic MNCs were
replaced with other domestic MNCs instead. The results are stable compared to those presented in the manuscript (see Tables E.7b-c).
 8

Table E.1a
Sample Characteristics of Alternative Models (Further Countries)
China (N=658) Italy (N=731)
in % Male Female Total Male Female Total
Age 18-25 14.4 10.4 24.8 7.5 7.3 14.8
Age 26-35 12.8 11.4 24.2 9.7 10.9 20.5
Age 36-45 13.5 11.3 24.8 12.5 12.7 25.2
Age 46-55 9.0 7.4 16.4 11.6 9.9 21.5
Age 56-65 7.1 2.9 9.9 8.6 9.4 18.0
Total 56.8 43.2 100.0 49.8 50.2 100.0

Table E.1b
Sample Characteristics of Alternative Models (Alternative MNCs in the Countries)
India (N=735) Japan (N=701) U.S. (N=658)
in % Male Female Total Male Female Total Male Female Total
Age 18-25 14.7 12.4 27.2 7.1 6.8 13.8 7.8 7.1 14.9
Age 26-35 12.0 14.6 26.7 11.9 9.3 21.2 9.5 9.5 19.1
Age 36-45 11.1 11.1 22.2 12.1 11.6 23.7 11.7 11.4 23.1
Age 46-55 7.6 7.8 15.4 11.7 9.2 20.9 13.2 11.6 24.7
Age 56-65 4.2 4.3 8.5 10.0 10.3 20.3 8.5 9.7 18.2
Total 49.7 50.3 100.0 52.8 47.2 100.0 50.7 49.3 100.0

Table E.1c
Sample Characteristics of Alternative Models (Alternative MNCs in the Countries)
China (N=632) Italy (N=725)
in % Male Female Total Male Female Total
Age 18-25 12.8 12.7 25.5 7.5 6.7 14.3
Age 26-35 11.2 11.6 22.8 10.3 8.9 19.2
Age 36-45 13.7 11.8 25.5 12.1 13.0 25.1
Age 46-55 8.2 8.3 16.5 11.4 11.9 23.3
Age 56-65 7.7 1.9 9.7 8.4 9.7 18.1
Total 53.7 46.3 100.0 49.7 50.3 100.0
 9

Table E.2a
Reliability and Validity of Alternative Models (Further Countries)
China Italy
Item MV/SD FL KMO ItTC α CR λ MV/SD FL KMO ItTC α CR λ
Perceived brand globalness
To me, [MNC] is a global brand. 3.9/1.2 .920 .895 .920 3.4/1.2 .882 .862 .884
I do think consumers abroad buy 3.8/1.1 .944 .912 .944 3.4/1.2 .960 .918 .960
[MNC’s] products. .776 .956 .956 .759 .951 .952
[MNC] sells its products all over 3.9/1.1 .950 .916 .951 3.5/1.2 .952 .913 .952
the world.
Functional value
[MNC] offers high-quality prod- 3.8/0.9 .905 .879 .911 3.5/1.0 .882 .851 .912
ucts and services.
[MNC] develops well-made inno- 3.7/0.9 .918 .890 .924 3.4/0.9 .933 .896 .918
vative products and services.
.876 .954 .957 .865 .945 .948
With its products, [MNC] is a 3.7/0.9 .915 .887 .925 3.4/0.9 .893 .861 .892
leader in research and technology.
[MNC] offers good value for 3.6/0.9 .926 .897 .923 3.4/0.9 .897 .865 .900
money.
Psychological value
[MNC] offers products I like and 3.5/1.0 .895 .852 .894 3.2/1.0 .846 .811 .858
enjoy.
The appealing image of [MNC’s] 3.5/0.9 .844 .809 .844 3.2/1.0 .875 .835 .873
products makes me feel good.
[MNC’s] products improve the 3.7/0.9 .865 .853 .825 .930 .930 .877 3.4/1.0 .886 .864 .844 .931 .931 .886
way I am perceived.
With [MNC’s] products I make a 3.6/0.9 .902 .857 .892 3.2/1.0 .904 .860 .897
better impression on other people
than with those of competitors.
Loyalty
I am a loyal customer of [MNC]. 3.2/1.0 .958 .925 .959 2.7/1.1 .961 .931 .960

I have developed a good relation- 3.1/1.0 .936 .909 .932 2.8/1.0 .924 .903 .927
.777 .960 .960 .776 .962 .963
ship with [MNC].
I am certain that I will buy prod- 3.2/1.0 .934 .908 .936 2.7/1.1 .954 .926 .953
ucts/offers of [MNC].
Consumer ethnocentrism
Purchasing foreign-made products 2.6/1.1 .894 .857 .895 2.6/1.2 .903 .862 .901
is un-[nationality].
[Nationality] should not buy for- 2.6/1.1 .935 .892 .931 2.7/1.2 .897 .858 .895
eign products because this hurts lo-
cal business/causes unemployment. .859 .939 .939 .852 .935 .936
A real [nationality] should always 2.8/1.2 .818 .793 .819 2.7/1.2 .875 .839 .876
buy [nationality]-made products.
It is not right to purchase foreign 2.4/1.2 .919 .877 .921 2.4/1.2 .866 .830 .870
products.
Model fits CFA: China: CFI .984; TLI .980; RMSEA .048; SRMR .029; ²(125) = 316.476; SCF = 1.2426. Italy: CFI .983; TLI .979; RMSEA .049;
SRMR .028; ²(125) = 342.202; SCF = 1.1704.
Note: MV/SD = Mean value/standard deviation; FL = Factor loadings; KMO = Kaiser-Meyer-Olkin criterion (≥ .5); ItTC = Item-to-total correlation
(≥ .5); α = Cronbach’s alpha (≥ .7); CR = Composite reliability (≥ .6); λ = Standardized factor loadings (≥ .5).
 10

Table E.2b
Reliability and Validity of Alternative Models (Alternative MNCs in the Countries)
India Japan U.S.
Item MV/SD FL KMO ItTC α CR λ MV/SD FL KMO ItTC α CR λ MV/SD FL KMO ItTC α CR λ
Perceived brand globalness
To me, [MNC] is a global brand. 3.2/1.3 .835 .800 .835 2.7/1.3 .674 .665 .676 3.6/1.1 .825 .802 .824
I do think consumers abroad buy 3.1/1.2 .925 .862 .926 3.1/1.0 .958 .851 .964 3.5/1.1 .949 .891 .947
[MNC’s] products. .750 .921 .921 .666 .887 .889 .740 .932 .933
[MNC] sells its products all over the 3.2/1.2 .919 .858 .918 3.1/1.0 .989 .870 .983 3.7/1.1 .946 .888 .949
world.
Functional value
[MNC] offers high-quality products and 3.6/1.0 .895 .863 .899 3.2/0.6 .808 .776 .878 3.5/0.9 .941 .919 .947
services.
[MNC] develops well-made innovative 3.5/1.1 .907 .874 .910 3.2/0.6 .890 .844 .912 3.4/0.9 .945 .922 .934
products and services. .867 .945 .947 .857 .926 .935 .846 .964 .965
With its products, [MNC] is a leader in 3.5/1.0 .896 .863 .896 3.2/0.6 .895 .848 .874 3.5/0.9 .927 .906 .934
research and technology.
[MNC] offers good value for money. 3.5/1.0 .903 .869 .911 3.2/0.6 .891 .845 .873 3.4/0.9 .922 .901 .919
Psychological value
[MNC] offers products I like and 3.4/1.0 .893 .843 .889 3.0/0.6 .815 .752 .803 3.2/1.0 .948 .917 .948
enjoy.
The appealing image of [MNC’s] prod- 3.1/1.1 .793 .760 .793 3.0/0.7 .758 .707 .738 3.2/0.9 .881 .860 .881
ucts makes me feel good.
[MNC’s] products improve the way I 3.4/1.1 .894 .856 .843 .921 .922 .895 3.1/0.7 .815 .830 .748 .883 .883 .829 3.3/1.0 .913 .871 .888 .956 .957 .921
am perceived.
With [MNC’s] products I make a better 3.3/1.1 .875 .828 .878 3.0/0.6 .861 .789 .872 3.2/0.9 .937 .908 .930
impression on other people than with
those of competitors.
Loyalty
I am a loyal customer of [MNC]. 3.0/1.2 .941 .906 .939 2.3/0.9 .962 .948 .962 2.6/1.2 .985 .971 .986

I have developed a good relationship 3.0/1.1 .934 .901 .933 2.4/0.9 .962 .948 .962 2.6/1.1 .952 .945 .952
.777 .953 .953 .782 .978 .978 .770 .983 .984
with [MNC].
I am certain that I will buy products/of- 3.1/1.2 .924 .894 .927 2.3/0.9 .981 .962 .981 2.6/1.2 .989 .974 .989
fers of [MNC].
Consumer ethnocentrism
Purchasing foreign-made products is un- 2.8/1.3 .889 .830 .891 2.5/0.9 .900 .849 .899 2.5/1.1 .904 .855 .912
[nationality].
[Nationality] should not buy foreign 2.9/1.3 .855 .804 .844 2.5/0.9 .902 .852 .900 2.9/1.1 .845 .809 .834
products because this hurts local busi-
ness/causes unemployment. .836 .909 .910 .857 .922 .923 .841 .928 .927
A real [nationality] should always buy 3.0/1.3 .777 .738 .779 2.6/1.0 .814 .779 .815 2.8/1.2 .856 .818 .844
[nationality]-made products.
It is not right to purchase foreign prod- 2.7/1.3 .862 .807 .869 2.4/0.9 .855 .811 .857 2.4/1.1 .888 .841 .899
ucts.
Model fits CFA: India: CFI .994; TLI .992; RMSEA .027; SRMR .029; ²(125) = 193.634; SCF = 1.2401. Japan: CFI .984; TLI .981; RMSEA .041; SRMR .031; ²(125) = 275.873; SCF = 1.3223. U.S.: CFI .985; TLI .982;
RMSEA .048; SRMR .021; ²(125) = 322.016; SCF = 1.2356.
Note: MV/SD = Mean value/standard deviation; FL = Factor loadings; KMO = Kaiser-Meyer-Olkin criterion (≥ .5); ItTC = Item-to-total correlation (≥ .5); α = Cronbach’s alpha (≥ .7); CR = Composite reliability (≥ .6);
λ = Standardized factor loadings (≥ .5).
 11

Table E.2c
Reliability and Validity of Alternative Models (Alternative MNCs in the Countries)
China Italy
Item MV/SD FL KMO ItTC α CR λ MV/SD FL KMO ItTC α CR λ
Perceived brand globalness
To me, [MNC] is a global brand. 3.7/1.3 .921 .896 .921 3.4/1.2 .882 .857 .882
I do think consumers abroad buy 3.6/1.2 .947 .915 .945 3.4/1.2 .952 .909 .951
[MNC’s] products. .776 .956 .957 .761 .947 .948
[MNC] sells its products all over 3.7/1.2 .948 .915 .950 3.5/1.2 .945 .903 .946
the world.
Functional value
[MNC] offers high-quality prod- 3.6/0.9 .902 .873 .907 3.4/0.9 .868 .838 .905
ucts and services.
[MNC] develops well-made inno- 3.6/0.9 .917 .886 .916 3.4/0.9 .916 .879 .920
vative products and services.
.872 .951 .950 .865 .942 .947
With its products, [MNC] is a 3.6/0.9 .919 .888 .914 3.3/0.9 .912 .875 .897
leader in research and technology.
[MNC] offers good value for 3.5/0.9 .904 .874 .898 3.3/0.9 .892 .858 .892
money.
Psychological value
[MNC] offers products I like and 3.4/0.9 .876 .836 .869 3.2/1.0 .891 .849 .893
enjoy.
The appealing image of [MNC’s] 3.4/0.9 .863 .825 .850 3.1/1.0 .853 .817 .849
products makes me feel good.
[MNC’s] products improve the 3.6/0.9 .883 .860 .841 .930 .929 .894 3.3/1.0 .865 .864 .827 .931 .931 .874
way I am perceived.
With [MNC’s] products I make a 3.5/0.9 .884 .842 .891 3.2/0.9 .907 .862 .900
better impression on other people
than with those of competitors.
Loyalty
I am a loyal customer of [MNC]. 3.0/1.0 .954 .918 .951 2.6/1.1 .963 .933 .961

I have developed a good relation- 3.0/1.0 .912 .887 .912 2.7/1.1 .924 .905 .928
.773 .955 .955 .775 .964 .964
ship with [MNC].
I am certain that I will buy prod- 3.0/1.0 .943 .910 .947 2.6/1.1 .956 .929 .955
ucts/offers of [MNC].
Consumer ethnocentrism
Purchasing foreign-made products 2.7/1.1 .880 .841 .875 2.5/1.2 .906 .868 .903
is un-[nationality].
[Nationality] should not buy for- 2.7/1.1 .931 .882 .929 2.7/1.2 .916 .877 .915
eign products because this hurts lo-
cal business/causes unemployment. .848 .931 .931 .852 .939 .940
A real [nationality] should always 2.8/1.2 .816 .785 .815 2.6/1.3 .879 .845 .883
buy [nationality]-made products.
It is not right to purchase foreign 2.5/1.1 .891 .847 .899 2.4/1.2 .867 .834 .868
products.
Model fits CFA: China: CFI .988; TLI .985; RMSEA .040; SRMR .030; ²(125) = 257.886; SCF = 1.2615. Italy: CFI .990; TLI .987; RMSEA .038;
SRMR .025; ²(125) = 253.875; SCF = 1.1954.
Note: MV/SD = Mean value/standard deviation; FL = Factor loadings; KMO = Kaiser-Meyer-Olkin criterion (≥ .5); ItTC = Item-to-total correlation
(≥ .5); α = Cronbach’s alpha (≥ .7); CR = Composite reliability (≥ .6); λ = Standardized factor loadings (≥ .5).
 12

Table E.3a
Discriminant Validity of Alternative Models (Further Countries)
Constructs PBG FV PV LOY
PBG .860
China FV .503 .885
PV .426 .828a .817
LOY .201 .462 .619 .889
PBG .850
Italy FV .359 .865
PV .309 .757 .824
LOY .142 .425 .557 .884
a
For the case where the criterion of Fornell and Larcker (1981) was slightly failed (squared correlation exceeded the AVE) we computed chi-square
difference tests (Anderson and Gerbing (1988). The comparison of a more restricted model with the proposed model shows satisfactory results, i.e. the
restricted model fitted significantly poorer (p < .001) than our proposed model and thus discriminant validity can be assumed.
Note: Values in bold represent the AVE of the construct; values in italics represent squared correlations; AVE = Average variance extracted (≥ .5);
PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty.

Table E.3b
Discriminant Validity of Alternative Models (Alternative MNCs in the Countries)
Constructs PBG FV PV LOY
PBG .766
FV .460 .832
India
PV .416 .773a .766
LOY .278 .496 .696 .844
PBG .782
FV .155 .920
Japan
PV .120 .701 .874
LOY .017 .118 .269 .950
PBG .850
U.S. FV .263 .903
PV .197 .773 .867
LOY .092 .421 .602 .937
PBG .844
FV .504 .877
China
PV .444 .834 .836
LOY .202 .417 .566 .895
PBG .851
FV .287 .876
Italy PV .231 .774 .814
LOY .087 .416 .581 .889
a
For the cases where the criterion of Fornell and Larcker (1981) was slightly failed (squared correlation exceeded the AVE) we computed chi-square
difference tests (Anderson and Gerbing (1988). The comparison of a more restricted model with the proposed model shows satisfactory results, i.e. the
restricted model fitted significantly poorer (p < .001) than our proposed model and thus discriminant validity can be assumed.
Note: Values in bold represent the AVE of the construct; values in italics represent squared correlations; AVE = Average variance extracted (≥ .5);
PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty.

Table E.4a
Measurement Invariance of Alternative Models (Further Countries)
² diff. CFI TLI RMSEA
Model ² (d.f.) SCF (p-value) (ΔCFI) (ΔTLI) (ΔRMSEA) SRMR
Invariance: MNC origin
Model 1: 292.331 1.2922 - .984 .979 .056 .030
Configural invariance (142) (-) (-) (-)
Model 2: 310.349 1.2510 18.018 .983 .980 .055 .045
China Metric invariance (156) (.558) (.001) (-.001) (.001)
Invariance: Consumer ethnocentrism
Model 1: 276.919 1.2981 - .985 .981 .053 .027
Configural invariance (142) (-) (-) (-)
Model 2: 301.558 1.2516 24.639 .984 .981 .053 .066
Metric invariance (156) (.060) (.001) (.000) (.000)
Invariance: MNC origin
Model 1: 279.982 1.1718 - .987 .983 .051 .031
Configural invariance (142) (-) (-) (-)
Model 2: 301.699 1.1474 21.717 .986 .984 .050 .046
Metric invariance (156) (.127) (.001) (-.001) (.001)
Italy
Invariance: Consumer ethnocentrism
Model 1: 303.550 1.1914 - .984 .980 .056 .033
Configural invariance (142) (-) (-) (-)
Model 2: 317.305 1.1630 13.755 .984 .981 .053 .048
Metric invariance (156) (.866) (.000) (-.001) (.003)
 13

Table E.4b
Measurement Invariance of Alternative Models (Alternative MNCs in the Countries)
² diff. CFI TLI RMSEA
Model ² (d.f.) SCF (p-value) (ΔCFI) (ΔTLI) (ΔRMSEA) SRMR
Invariance: MNC origin
Model 1: 205.523 1.2321 - .993 .991 .035 .021
Configural invariance (142) (-) (-) (-)
Model 2: 219.300 1.1932 13.777 .993 .992 .033 .042
Metric invariance (156) (.719) (.000) (-.001) (.002)
India
Invariance: Consumer ethnocentrism
Model 1: 186.301 1.2312 - .995 .994 .029 .020
Configural invariance (142) (-) (-) (-)
Model 2: 204.813 1.1920 18.512 .995 .994 .029 .054
Metric invariance (156) (.181) (.000) (.000) (.000)
Invariance: MNC origin
Model 1: 264.419 1.3301 - .984 .980 .049 .037
Configural invariance (142) (-) (-) (-)
Model 2: 277.063 1.3002 12.644 .984 .981 .048 .053
Metric invariance (156) (.858) (.000) (-.001) (.001)
Japan
Invariance: Consumer ethnocentrism
Model 1: 273.750 1.3250 - .983 .979 .051 .035
Configural invariance (142) (-) (-) (-)
Model 2: 288.572 1.3065 14.822 .983 .981 .049 .054
Metric invariance (156) (.544) (.000) (-.002) (.002)
Invariance: MNC origin
Model 1: 240.777 1.3283 - .982 .977 .063 .029
Configural invariance (142) (-) (-) (-)
Model 2: 262.019 1.2931 21.242 .981 .977 .063 .060
Metric invariance (156) (.121) (.001) (.000) (.000)
U.S.
Invariance: Consumer ethnocentrism
Model 1: 231.340 1.3376 - .983 .979 .060 .024
Configural invariance (142) (-) (-) (-)
Model 2: 253.487 1.3027 22.147 .982 .979 .060 .062
Metric invariance (156) (.081) (.001) (.000) (.000)
Invariance: MNC origin
Model 1: 237.050 1.2923 - .989 .986 .045 .028
Configural invariance (142) (-) (-) (-)
Model 2: 256.858 1.2539 19.808 .989 .987 .044 .047
Metric invariance (156) (.198) (.000) (-.001) (.001)
China
Invariance: Consumer ethnocentrism
Model 1: 267.754 1.3041 - .986 .982 .051 .027
Configural invariance (142) (-) (-) (-)
Model 2: 290.512 1.2655 22.758 .985 .982 .051 .069
Metric invariance (156) (.098) (.001) (.000) (.000)
Invariance: MNC origin
Model 1: 239.217 1.1837 - .991 .988 .043 .028
Configural invariance (142) (-) (-) (-)
Model 2: 260.679 1.1597 21.462 .990 .989 .043 .047
Metric invariance (156) (.104) (.001) (-.001) (.000)
Italy
Invariance: Consumer ethnocentrism
Model 1: 244.803 1.2013 - .990 .987 .045 .031
Configural invariance (142) (-) (-) (-)
Model 2: 264.590 1.1764 19.787 .990 .988 .044 .085
Metric invariance (156) (.181) (.000) (-.001) (.001)
 14

Table E.5a
Single-factor Test of Alternative Models (Further Countries)
China Italy
Model χ² ² diff. χ² ² diff.
CFI TLI RMSEA SCF (df) (p-value) CFI TLI RMSEA SCF (df) (p-value)
CFA .983 .979 .057 1.3056 255.246 .984 .979 .057 1.1849 238.442
(71) (71)
*** ***
SFT .727 .677 .222 1.3205 2614.570 .672 .613 .246 1.2015 3495.067
(77) (77)
Note: * p < .05; ** p < .01; *** p < .001; ns Not significant; CFA = Model fit of the confirmatory factor analysis; SFT = Model fit of the
single factor test; SCF = Scaling correction factor for MLM.

Tables E.5b-d
Marker Variable Technique of Alternative Models (Further Countries)
Table E.5b
Results of Model Comparisons (Phase I)
China Italy
Model ² df SCF CFI TLI RMSEA ² df SCF CFI TLI RMSEA
CFA 294.679 85 1.2308 .977 .972 .062 317.277 85 1.1497 .978 .973 .062
Baseline 334.582 90 1.2444 .974 .969 .065 337.941 90 1.1430 .976 .972 .062
Method-C 325.265 89 1.2413 .974 .970 .064 336.987 89 1.1449 .976 .972 .062
Method-U 295.528 76 1.2778 .976 .967 .067 304.310 76 1.1596 .978 .970 .065
Method-R 293.457 82 1.2884 .977 .971 .063 302.310 82 1.1710 .979 .973 .061
Chi-square differences of model comparison tests
ΔModels Δ² Δdf p Δ² Δdf p
Baseline vs. Method-C 9.317 1 ** .954 1 ns
Method-C vs. Method-U 29.737 13 * 32.677 13 **
Method-U vs. Method-R 2.071 6 ns 2.000 6 ns
Note: * p < .05; ** p < .01; *** p < .001; ns Not significant; SCF = Scaling correction factor for MLM.

Table E.5c
Results of the Reliability Decomposition (Phase II)
China Italy
Reliability Reliability
baseline Decomposed reliability from method-U baseline Decomposed reliability from method-U
model model model model
Total Substantive Method relia- % reliability Total Substantive Method relia- % reliability
Latent variable reliability reliability bility marker reliability reliability bility marker
PBG .951 .896 .055 5.78 .947 .893 .053 5.60
FV .965 .934 .031 3.21 .960 .931 .029 3.02
PV .943 .913 .030 3.18 .945 .916 .029 3.07
LOY .961 .910 .051 5.31 .959 .909 .050 5.21
Note: PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty.

Table E.5d
Results of the Sensitivity Analyses (Phase III)
China Italy
CFA Baseline Method-U Method-S Method-S CFA Baseline Method-U Method-S Method-S
Construct correlations (.05) (.01) (.05) (.01)
PBG with FV .714 .714 .712 .710 .710 .622 .622 .628 .630 .631
PBG with PV .669 .669 .674 .672 .671 .589 .589 .608 .623 .625
PBG with LOY .486 .486 .499 .493 .491 .412 .412 .435 .466 .473
PV with FV .926 .926 .924 .923 .923 .895 .895 .896 .898 .898
PV with LOY .806 .806 .807 .806 .806 .752 .752 .751 .754 .755
FV with LOY .716 .716 .720 .717 .716 .670 .670 .677 .691 .695
PBG with marker -.164 .000 .000 .000 .000 .024 .000 .000 .000 .000
FV with marker -.171 .000 .000 .000 .000 .005 .000 .000 .000 .000
PV with marker -.180 .000 .000 .000 .000 -.035 .000 .000 .000 .000
LOY with marker -.230 .000 .000 .000 .000 .020 .000 .000 .000 .000
Note: PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty.
 15

Table E.5e
Single-factor Test of Alternative Models (Alternative MNCs in the Countries)
India Japan U.S.
Model χ² ² diff. χ² ² diff. χ² ² diff.
CFI TLI RMSEA SCF (df) (p-value) CFI TLI RMSEA SCF (df) (p-value) CFI TLI RMSEA SCF (df) (p-value)
CFA .996 .995 .027 1.2367 108.114 .984 .980 .050 1.3151 195.797 .986 .983 .055 1.3030 220.020
(71) (71) (71)
*** *** ***
SFT .767 .725 .195 1.2469 2251.432 .488 .395 .273 1.3137 4154.244 .636 .570 .274 1.3015 4067.317
(77) (77) (77)
Note: * p < .05; ** p < .01; *** p < .001; ns Not significant; CFA = Model fit of the confirmatory factor analysis; SFT = Model fit of the single factor test; SCF = Scaling correction factor for MLM.

Tables E.5f-D.5h
Marker Variable Technique of Alternative Models (Alternative MNCs in the Countries)
Table E.5f
Results of Model Comparisons (Phase I)
India Japan U.S.
Model ² df SCF CFI TLI RMSEA ² df SCF CFI TLI RMSEA ² df SCF CFI TLI RMSEA
CFA 143.735 85 1.2035 .994 .992 .031 463.120 85 1.2674 .953 .942 .081 310.279 85 1.2378 .979 .975 .063
Baseline 164.203 90 1.2153 .992 .991 .034 490.477 90 1.2500 .950 .942 .081 334.902 90 1.2223 .978 .974 .064
Method-C 160.934 89 1.2170 .992 .991 .033 489.119 89 1.2529 .950 .941 .081 332.059 89 1.2245 .978 .974 .064
Method-U 137.964 76 1.2258 .993 .990 .034 456.083 76 1.2846 .952 .934 .085 291.545 76 1.2409 .980 .973 .065
Method-R 136.681 82 1.2392 .994 .992 .031 455.780 82 1.2861 .953 .940 .082 289.100 82 1.2555 .981 .976 .061
Chi-square differences of model comparison tests
ΔModels Δ² Δdf p Δ² Δdf p Δ² Δdf p
Baseline vs. Method-C 3.269 1 ns 1.358 1 ns 2.843 1 ns
Method-C vs. Method-U 22.970 13 * 33.036 13 * 40.514 13 ***
Method-U vs. Method-R 1.283 6 ns .303 6 ns 2.445 6 ns
Note: * p < .05; ** p < .01; *** p < .001; ns Not significant; SCF = Scaling correction factor for MLM.

Table E.5g
Results of the Reliability Decomposition (Phase II)
India Japan U.S.
Reliability Reliability Reliability
baseline model Decomposed reliability from method-U model baseline model Decomposed reliability from method-U model baseline model Decomposed reliability from method-U model
Total Substantive re- Method % reliability Total Substantive re- Method % reliability Total Substantive re- Method % reliability
Latent variable reliability liability reliability marker reliability liability reliability marker reliability liability reliability marker
PBG .913 .859 .054 5.91 .929 .876 .053 5.70 .942 .896 .046 4.88
FV .950 .919 .031 3.26 .976 .947 .029 2.97 .973 .949 .024 2.47
PV .925 .894 .032 3.46 .956 .928 .029 3.03 .962 .937 .025 2.60
LOY .945 .893 .052 5.50 .983 .934 .049 4.99 .978 .934 .045 4.60
Note: PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty.
 16

Table E.5h
Results of the Sensitivity Analyses (Phase III)
India Japan U.S.
CFA Baseline Method-U Method-S Method-S CFA Baseline Method-U Method-S Method-S CFA Baseline Method-U Method-S Method-S
Construct correlations (.05) (.01) (.05) (.01) (.05) (.01)
PBG with FV .676 .676 .676 .672 .672 .581 .581 .578 .576 .575 .583 .583 .574 .568 .566
PBG with PV .644 .644 .647 .644 .644 .551 .551 .550 .550 .549 .519 .519 .520 .514 .512
PBG with LOY .527 .527 .522 .516 .515 .326 .326 .340 .356 .358 .335 .335 .319 .310 .307
PV with FV .887 .887 .889 .888 .888 .939 .939 .938 .937 .937 .898 .898 .899 .899 .899
PV with LOY .823 .823 .825 .824 .824 .663 .663 .669 .677 .679 .736 .736 .740 .740 .740
FV with LOY .700 .700 .702 .698 .697 .565 .565 .572 .584 .587 .622 .622 .618 .616 .616
PBG with marker -.138 .000 .000 .000 .000 -.162 .000 .000 .000 .000 -.137 .000 .000 .000 .000
FV with marker -.089 .000 .000 .000 .000 -.073 .000 .000 .000 .000 -.082 .000 .000 .000 .000
PV with marker -.049 .000 .000 .000 .000 -.086 .000 .000 .000 .000 -.088 .000 .000 .000 .000
LOY with marker -.081 .000 .000 .000 .000 -.028 .000 .000 .000 .000 -.118 .000 .000 .000 .000
Note: PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty.
 17

Table E.5i
Single-factor Test of Alternative Models (Alternative MNCs in the Countries)
China Italy
Model χ² ² diff. χ² ² diff.
CFI TLI RMSEA SCF (df) (p-value) CFI TLI RMSEA SCF (df) (p-value)
CFA .988 .984 .048 1.2998 180.179 .991 .989 .041 1.1836 157.439
(71) (71)
*** ***
SFT .724 .674 .217 1.3307 2514.483 .668 .607 .245 1.1946 3445.484
(77) (77)
Note: * p < .05; ** p < .01; *** p < .001; ns Not significant; CFA = Model fit of the confirmatory factor analysis; SFT = Model fit of the
single factor test; SCF = Scaling correction factor for MLM.

Tables E.5j-D.5l
Marker Variable Technique of Alternative Models (Alternative MNCs in the Countries)
Table E.5j
Results of Model Comparisons (Phase I)
China Italy
Model ² df SCF CFI TLI RMSEA ² df SCF CFI TLI RMSEA
CFA 285.100 85 1.2225 .977 .972 .060 239.209 85 1.1469 .985 .981 .051
Baseline 326.454 90 1.2414 .973 .969 .063 280.734 90 1.1437 .981 .978 .055
Method-C 319.465 89 1.2420 .974 .969 .063 278.565 89 1.1446 .981 .978 .055
Method-U 283.421 76 1.2888 .976 .967 .064 244.253 76 1.1501 .983 .976 .057
Method-R 279.589 82 1.3146 .978 .971 .061 243.041 82 1.1578 .984 .979 .053

Chi-square differences of model comparison tests

ΔModels Δ² Δdf p Δ² Δdf p
Baseline vs. Method-C 6.989 1 ** 2.169 1 ns
Method-C vs. Method-U 36.044 13 ** 34.312 13 **
Method-U vs. Method-R 3.832 6 ns 1.212 6 ns
Note: * p < .05; ** p < .01; *** p < .001; ns Not significant; SCF = Scaling correction factor for MLM.

Table E.5k
Results of the Reliability Decomposition (Phase II)
China Italy
Reliability Reliability
baseline Decomposed reliability from method-U baseline Decomposed reliability from method-U
model model model model
Total Substantive Method relia- % reliability Total Substantive Method relia- % reliability
Latent variable reliability reliability bility marker reliability reliability bility marker
PBG .946 .885 .061 6.45 .947 .896 .051 5.39
FV .963 .930 .033 3.43 .961 .941 .020 2.08
PV .949 .918 .031 3.27 .944 .914 .030 3.18
LOY .962 .910 .052 5.41 .960 .910 .050 5.21
Note: PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty.

Table E.5l
Results of the Sensitivity Analyses (Phase III)
China Italy
CFA Baseline Method-U Method-S Method-S CFA Baseline Method-U Method-S Method-S
Construct correlations (.05) (.01) (.05) (.01)
PBG with FV .706 .706 .715 .714 .713 .573 .573 .586 .585 .586
PBG with PV .672 .672 .701 .711 .710 .531 .531 .540 .550 .551
PBG with LOY .499 .499 .531 .546 .544 .352 .352 .357 .368 .369
PV with FV .934 .934 .935 .935 .934 .905 .905 .901 .903 .902
PV with LOY .800 .800 .801 .802 .802 .772 .772 .772 .771 .771
FV with LOY .715 .715 .720 .724 .723 .671 .671 .667 .670 .669
PBG with marker -.148 .000 .000 .000 .000 .026 .000 .000 .000 .000
FV with marker -.128 .000 .000 .000 .000 .070 .000 .000 .000 .000
PV with marker -.123 .000 .000 .000 .000 .032 .000 .000 .000 .000
LOY with marker -.170 .000 .000 .000 .000 .078 .000 .000 .000 .000
Note: PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty.
 18

Table E.6a
F-test of Strong Instrument Variables (Further Countries)
China Italy
IV → PBG 124.201 176.685
Note: IV = Instrumental variable; PBG = Perceived brand globalness; F-value > 10 indicates strong predictor.

Table E.6b
Results of the Efficient and Consistent models (Further Countries)
China Italy
Consistent model Proposed / efficient Consistent model Proposed / efficient
model model
Effects beta p beta p beta p beta p
PBG → FV .726 *** .720 *** .370 *** .362 ***
PBG → PV .679 *** .671 *** .590 *** .579 ***
FV → LOY .328 *** .328 *** .360 *** .360 ***
PV → LOY .715 *** .713 *** .562 *** .562 ***
IV → PBG .489 *** - .586 *** -
Gender -.031 ns -.031 ns -.055 * -.055 *
Age -.003 ns -.003 ns .011 ns .011 ns
Brand familiarity -.036 ns -.036 ns .025 ns .025 ns
Total effect
PBG → LOY .723 *** .714 *** .465 *** .456 ***
Indirect effects
PBG → FV → LOY .238 *** .236 *** .133 *** .130 ***
PBG → PV → LOY .485 *** .478 *** .332 *** .326 ***
Model fits:
China: Consistent model: CFI .930; TLI .918; RMSEA .089; SRMR .133; ²(126) = 788.347; SCF = 1.2280; Proposed model: CFI .943;
TLI .930; RMSEA .086; SRMR .106; ²(112) = 628.061; SCF = 1.2650.
Italy: Consistent model: CFI .927; TLI .914; RMSEA .087; SRMR .101; ²(126) = 830.222; SCF = 1.1427; Proposed model: CFI .947;
TLI .935; RMSEA .079; SRMR .103; ²(112) = 597.515; SCF = 1.1559.
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psycholog-
ical value; LOY = Loyalty; IV = Instrumental variable ‘Exposure to marketing activities of MNCs’; SCF = Scaling correction factor for
MLM; b = unstandardized coefficients; beta = standardized coefficients.
 19

Table E.6c
F-test of Strong Instrument Variables (Alternative MNCs in the Countries)
India Japan U.S. China Italy
IV → PBG 219.381 10.389 156.841 158.763 115.442
Note: IV = Instrumental variable; PBG = Perceived brand globalness; F-value > 10 indicates strong predictor.

Table E.6d
Results of the Efficient and Consistent Models (Alternative MNCs in the Countries)
India Japan U.S. China Italy
Consistent Proposed / effi- Consistent Proposed / effi- Consistent Proposed / effi- Consistent Proposed / effi- Consistent Proposed / effi-
model cient model model cient model model cient model model cient model model cient model
Effects beta p beta p beta p beta p beta p beta p beta p beta p beta p beta p
PBG → FV .681 *** .671 *** .382 *** .379 *** .418 *** .404 *** .735 *** .730 *** .553 *** .545 ***
PBG → PV .697 *** .681 *** .333 *** .330 *** .477 *** .462 *** .693 *** .686 *** .501 *** .499 ***
FV → LOY .460 *** .462 *** .306 *** .306 *** .149 *** .150 *** .255 *** .255 *** .323 *** .323 ***
PV → LOY .562 *** .566 *** .728 *** .728 *** .695 *** .696 *** .756 *** .756 *** .530 *** .533 ***
IV → PBG .779 *** - .174 *** - .629 *** - .545 *** - .458 *** -
Gender .000 ns .000 ns .049 ns .049 ns -.006 ns -.006 ns -.006 ns -.006 ns -.029 ns -.029 ns
Age .045 * .045 * .042 ns .042 ns -.074 ** -.074 ** -.016 ns -.016 ns -.037 ns -.037 ns
Brand familiarity .040 ns .041 ns .016 ns .016 ns -.040 ns -.040 ns -.020 ns -.020 ns .028 ns .028 ns
Total effect
PBG → LOY .705 *** .695 *** .360 *** .356 *** .394 *** .382 *** .712 *** .705 *** .444 *** .442 ***
Indirect effects
PBG → FV → LOY .314 *** .310 *** .117 *** .116 *** .062 *** .061 *** .187 *** .186 *** .178 *** .176 ***
PBG → PV → LOY .392 *** .385 *** .243 *** .240 *** .331 *** .321 *** .524 *** .519 *** .266 *** .266 ***
Model fits:
India: Consistent model: CFI .945; TLI .935; RMSEA .073; SRMR .110; ²(126) = 620.404; SCF = 1.2188; Proposed model: CFI .951; TLI .940; RMSEA .072; SRMR .114; ²(112) = 525.333; SCF = 1.2456.
Japan: Consistent model: CFI .926; TLI .914; RMSEA .071; SRMR .094; ²(126) = 636.565; SCF = 1.2894; Proposed model: CFI .931; TLI .916; RMSEA .074; SRMR .092; ²(112) = 513.843; SCF = 1.3471.
U.S.: Consistent model: CFI .935; TLI .923; RMSEA .087; SRMR .102; ²(126) = 784.329; SCF = 1.2650; Proposed model: CFI .950; TLI .939; RMSEA .081; SRMR .109; ²(112) = 586.119; SCF = 1.2968.
China: Consistent model: CFI .939; TLI .928; RMSEA .083; SRMR .122; ²(126) = 709.187; SCF = 1.2081; Proposed model: CFI .948; TLI .937; RMSEA .081; SRMR .100; ²(112) = 581.879; SCF = 1.2408.
Italy: Consistent model: CFI .917; TLI .904; RMSEA .093; SRMR .107; ²(126) = 1030.307; SCF = 1.1314; Proposed model: CFI .939; TLI .926; RMSEA .088; SRMR .105; ²(112) = 707.659; SCF = 1.1497.
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty; IV = Instrumental variable ‘Exposure to market-
ing activities of MNCs’; SCF = Scaling correction factor for MLM; b = unstandardized coefficients; beta = standardized coefficients.
 20

Table E.7a
Results of Alternative Models (Further Countries)
China Italy
1. General model Total sample Rival model (direct path) Total sample Rival model (direct path)
Effects beta bp beta bp beta bp beta bp
PBG→FV .720 .546*** .721 .547*** .362 .273*** .364 .274***
PBG→PV .671 .533*** .675 .536*** .579 .469*** .584 .473***
FV→LOY .328 .327*** .351 .349*** .360 .440*** .357 .435***
PV→LOY .713 .679*** .727 .689*** .562 .641*** .610 .691***
Gender -.031 -.052ns -.032 -.053ns -.055 -.109* -.060 -.118*
Age -.003 -.002ns -.003 -.002ns .011 .008ns .015 .011ns
Familiarity -.036 -.019ns -.037 -.020ns .025 .015ns .027 .016ns
PBG→LOY - - -.041 -.031ns - - -.067 -.061ns
Total effect
PBG→LOY .714 .541*** .456 .421***
Indirect effects (H1)
PBG→FV→LOY .236 .179*** .130 .120***
PBG→PV→LOY .478 .362*** .326 .301***
2. Moderator: Diff. betw. countries Diff. betw. countries
MNC origin Foreign Domestic Foreign vs. domestic Foreign Domestic Foreign vs. domestic
Effects beta bp beta bp bp beta bp beta bp bp
PBG→FV .695 .533*** .516 .327*** .206*** .486 .365*** .417 .332*** .033ns
PBG→PV .628 .532*** .686 .529*** .003ns .658 .532*** .508 .417*** .115***
FV→LOY .528 .559*** .261 .289*** .270*** .625 .676*** .450 .481*** .195***
PV→LOY .539 .516*** .789 .718*** -.202*** .301 .302*** .570 .590*** -.288***
Gender -.038 -.063ns -.019 -.029ns -.017 -.027ns .057 .097ns
Age .018 .012ns -.021 -.012ns .119 .072*** .071 .047*
Familiarity -.032 -.017ns -.048 -.024ns .006 .003ns -.007 -.004ns
Total effect (H2)
PBG→LOY .705 .573*** .676 .474*** .099*** .502 .407*** .477 .406*** .001ns
Diff. within countries Diff. within countries
Foreign Domestic Foreign Domestic
Indirect effects (H3) bp bp bp bp
PBG→FV→LOY .367 .298*** .135 .095*** .023ns .304 .247*** .187 .159*** .086***
PBG→PV→LOY .338 .275*** .541 .380*** -.285*** .198 .161*** .290 .246*** -.087***
3. Moderator Diff. betw. groups Diff. betw. groups
Consumer
Consumer ethnocentrism Lower Higher Lower vs. higher Lower Higher Lower vs. higher
Effects beta bp beta bp bp beta bp beta bp bp
PBG→FV .752 .581*** .670 .490*** .091*** .418 .327*** .325 .252*** .075*
PBG→PV .723 .582*** .601 .489*** .093*** .642 .499*** .566 .388*** .111***
FV→LOY .503 .513*** .413 .409*** .104*** .522 .543*** .338 .361*** .182***
PV→LOY .504 .493*** .692 .616*** -.123*** .587 .614*** .659 .798*** -.184***
Gender -.022 -.040ns -.040 -.059ns .017 .032ns .016 .028ns
Age .018 .012ns -.013 -.008ns .075 .052* .056 .036ns
Familiarity -.055 -.032* -.019 -.009ns -.041 -.022ns .060 .029*
Total effect (H4)
PBG→LOY .742 .585*** .694 .502*** .083*** .595 .484*** .483 .401*** .083*
Diff. within groups Diff. within groups
Lower Higher Lower Higher
Indirect effects (H5) bp bp bp bp
PBG→FV→LOY .378 .298*** .277 .200*** .011ns .218 .178*** .110 .091***
PBG→PV→LOY .364 .287*** .416 .301*** -.101*** .377 .307*** .373 .310*** -.129*** -.219***
Model fits: General model: China: CFI .943; TLI .930; RMSEA .086; SRMR .106; ²(112) = 628.061; SCF = 1.2650. Italy: CFI .947; TLI .935;
RMSEA .079; SRMR .103; ²(112) = 597.515; SCF = 1.1559. MNC origin: China: CFI .934; TLI .924; RMSEA .089; SRMR .106;
²(234) = 818.085; SCF = 1.2421. Italy: CFI .950; TLI .942; RMSEA .073; SRMR .100; ²(234) = 674.058; SCF = 1.2022. CE: China: CFI .942;
TLI .933; RMSEA .084; SRMR .092; ²(234) = 759.345; SCF = 1.2394. Italy: CFI .947; TLI .939; RMSEA .076; SRMR .095; ²(234) = 708.840;
SCF = 1.2086.
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; SCF = Scaling correction factor for MLM; b = unstandardized coefficients; beta = stand-
ardized coefficients.
 21
Table E.7b
Results of Alternative Models (Alternative MNCs in the Countries)
India Japan U.S.
1. General model Total sample Rival model (direct path) Total sample Rival model (direct path) Total sample Rival model (direct path)
Effects beta bp beta bp beta bp beta bp beta bp beta bp
PBG→FV .671 .561*** .674 .564*** .379 .189*** .382 .191*** .404 .312*** .405 .313***
PBG→PV .681 .555*** .685 .558*** .330 .202*** .339 .208*** .462 .437*** .466 .441***
FV→LOY .462 .472*** .485 .494*** .306 .380*** .326 .401*** .150 .238*** .155 .246***
PV→LOY .566 .594*** .597 .623*** .728 .734*** .742 .744*** .696 .900*** .724 .932***
Gender .000 .000ns .001 .002ns .049 .053ns .047 .050ns -.006 -.013ns -.012 -.026ns
Age .045 .033* .048 .035* .042 .017ns .047 .020ns -.074 -.062** -.070 -.059**
Familiarity .041 .024ns .038 .022ns .016 .005ns .014 .004ns -.040 -.022ns -.043 -.024ns
**
PBG→LOY - - -.061 -.052ns - - -.058 -.036ns - - -.059 -.072ns
Total effect
PBG→LOY .695 .595*** .356 .221*** .382 .468***
Indirect effects (H1)
PBG→FV→LOY .310 .265*** .116 .072*** .061 .074***
PBG→PV→LOY .385 .330*** .240 .149*** .321 .394***
2. Moderator: Diff. betw. countries Diff. betw. countries Diff. betw. countries
MNC origin Foreign Domestic Foreign vs. domestic Foreign Domestic Foreign vs. domestic Foreign Domestic Foreign vs. domestic
Effects beta bp beta bp bp beta bp beta bp bp beta bp beta bp bp
PBG→FV .723 .566*** .365 .268*** .298*** .426 .264*** .375 .195*** .069*** .462 .385*** .463 .359*** .026ns
PBG→PV .709 .555*** .658 .539*** .016ns .359 .270*** .341 .234*** .036* .542 .600*** .470 .488*** .112**
FV→LOY .514 .547*** .453 .485*** .062*** .490 .649*** .344 .444*** .205*** .751 .930*** .433 .527*** .403***
PV→LOY .499 .531*** .630 .605*** -.074*** .530 .577*** .723 .708*** -.131*** .260 .242*** .649 .589*** -.347***
Gender .021 .038ns -.028 -.050ns .025 .032ns .052 .049ns .027 .045ns .043 .073ns
Age .030 .021ns .042 .029ns .024 .011ns .059 .021ns -.031 -.020ns .025 .015ns
Familiarity .054 .032ns -.009 -.005ns .028 .009ns .023 .005ns -.008 -.004ns -.020 -.008ns
Total effect (H2)
PBG→LOY .726 .604*** .579 .456*** .148*** .399 .327*** .375 .252*** .075*** .488 .503*** .505 .477*** .026ns
Diff. within countries Diff. within countries Diff. within countries
Foreign Domestic Foreign Domestic Foreign Domestic
Indirect effects (H3) bp bp bp bp bp bp
PBG→FV→LOY .372 .310*** .165 .130*** .015ns .209 .171*** .129 .087*** .015ns .347 .358*** .201 .189*** .213***
PBG→PV→LOY .354 .295*** .414 .326*** -.196*** .190 .156*** .246 .166*** -.079*** .141 .145*** .305 .288*** -.099***
3. Moderator Diff. betw. groups Diff. betw. groups Diff. betw. groups
Consumer
Consumer ethnocentrism Lower Higher Lower vs. higher Lower Higher Lower vs. higher Lower Higher Lower vs. higher
Effects beta bp beta bp bp beta bp beta bp bp beta bp beta bp bp
PBG→FV .762 .686*** .706 .595*** .091*** .433 .214*** .283 .129*** .085*** .474 .377*** .392 .287*** .090***
PBG→PV .744 .607*** .678 .532*** .075*** .363 .212*** .187 .112*** .100*** .488 .523*** .406 .365*** .158***
FV→LOY .499 .504*** .394 .391*** .113*** .327 .405*** .291 .367*** .038ns .399 .626*** .176 .261*** .365***
PV→LOY .487 .542*** .616 .656*** -.114*** .683 .715*** .759 .728*** -.013ns .501 .583*** .652 .791*** -.208***
Gender -.023 -.045ns .012 .021ns -.004 -.005ns .074 .072* -.052 -.106ns .007 .015ns
Age .059 .043* -.003 -.002ns .053 .022ns .040 .015ns -.107 -.082* -.112 -.086*
Familiarity .021 .014ns .034 .018ns .104 .032* -.001 .000ns -.027 -.014ns -.038 -.020ns
Total effect (H4)
PBG→LOY .742 .675*** .696 .582*** .093*** .390 .238*** .224 .129*** .109** .434 .541*** .334 .364*** .177***
Diff. within groups Diff. within groups Diff. within groups
Lower Higher Lower Higher Lower Higher
Indirect effects (H5) bp bp bp bp bp bp
PBG→FV→LOY .380 .346*** .278 .233*** .017ns .142 .086*** .082 .047*** .189 .236*** .069 .075***
PBG→PV→LOY .362 .329*** .418 .349*** -.116*** .248 .152*** .142 .082*** -.066** -.035*** .245 .305*** .265 .289*** -.069* -.214***
Model fits: General model: India: CFI .951; TLI .940; RMSEA .072; SRMR .114; ²(112) = 525.333; SCF = 1.2456. Japan: CFI .931; TLI .916; RMSEA .074; SRMR .092; ²(112) = 513.843; SCF = 1.3471. U.S.: CFI .950; TLI .939;
RMSEA .081; SRMR .109; ²(112) = 586.119; SCF = 1.2968. MNC origin: India: CFI .949; TLI .941; RMSEA .072; SRMR .103; ²(234) = 667.876; SCF = 1.2342. Japan: CFI .923; TLI .911; RMSEA .076; SRMR .081; ²(234) =
706.273; SCF = 1.3494. U.S.: CFI .911; TLI .897; RMSEA .107; SRMR .101; ²(234) = 670.119; SCF = 1.2814. CE: India: CFI .954; TLI .947; RMSEA .070; SRMR .097; ²(234) = 630.192; SCF = 1.2121. Japan: CFI .931; TLI
.920; RMSEA .073; SRMR .091; ²(234) = 652.836; SCF = 1.3456. U.S.: CFI .947; TLI .939; RMSEA .082; SRMR .108; ²(234) = 488.269; SCF = 1.2654.
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; SCF = Scaling correction factor for MLM; b = unstandardized coefficients; beta = standardized coefficients.
 22

Table E.7c
Results of Alternative Models (Alternative MNCs in the Countries)
China Italy
1. General model Total sample Rival model (direct path) Total sample Rival model (direct path)
Effects beta bp beta bp beta bp beta bp
PBG→FV .730 .521*** .731 .522*** .545 .446*** .545 .446***
PBG→PV .686 .473*** .691 .476*** .499 .492*** .499 .492***
FV→LOY .255 .252*** .282 .276*** .323 .416*** .326 .419***
PV→LOY .756 .473*** .775 .787*** .533 .568*** .532 .568***
Gender -.006 -.010ns -.008 -.013ns -.029 -.059ns -.029 -.059ns
Age -.016 -.010ns -.018 -.011ns -.037 -.029ns -.037 -.029ns
Familiarity -.020 -.012ns -.022 -.013ns .028 .018ns .028 .018ns
PBG→LOY - - -.048 -.034ns - - -.003 -.003ns
Total effect
PBG→LOY .705 .496*** .442 .467***
Indirect effects (H1)
PBG→FV→LOY .186 .131*** .176 .186***
PBG→PV→LOY .519 .365*** .266 .281***
2. Moderator: Diff. betw. countries Diff. betw. countries
MNC origin Foreign Domestic Foreign vs. domestic Foreign Domestic Foreign vs. domestic
Effects beta bp beta bp bp beta bp beta bp bp
PBG→FV .699 .495*** .668 .429*** .066* .665 .508*** .593 .486*** .022ns
PBG→PV .677 .502*** .685 .447*** .055ns .649 .569*** .531 .492*** .077*
FV→LOY .537 .569*** .228 .244*** .325*** .517 .581*** .231 .265*** .316***
PV→LOY .534 .540*** .775 .816*** -.276*** .395 .388*** .742 .755*** -.367***
Gender .000 -.001ns -.006 -.010ns .024 .039ns .046 .087ns
Age .009 .005ns -.018 -.012ns .127 .079*** .077 .059*
Familiarity -.018 -.010ns -.008 -.005ns .024 .011ns -.015 -.009ns
Total effect (H2)
PBG→LOY .737 .553*** .684 .470*** .087*** .600 .516*** .531 .501*** .015ns
Diff. within countries Diff. within countries
Foreign Domestic Foreign Domestic
Indirect effects (H3) bp bp bp bp
PBG→FV→LOY .375 .281*** .153 .105*** .010ns .343 .295*** .137 .129*** .074**
PBG→PV→LOY .362 .271*** .531 .365*** -.260*** .257 .221*** .394 .372*** -.243***
3. Moderator Diff. betw. groups Diff. betw. groups
Consumer
Consumer ethnocentrism Lower Higher Lower vs. higher Lower Higher Lower vs. higher
Effects beta bp beta bp bp beta bp beta bp bp
PBG→FV .767 .563*** .723 .464*** .099*** .471 .347*** .269 .206*** .141***
PBG→PV .696 .509*** .675 .473*** .036** .614 .476*** .458 .337*** .139***
FV→LOY .468 .504*** .333 .358*** .146*** .450 .474*** .334 .354*** .120***
PV→LOY .510 .551*** .636 .626*** -.075*** .638 .638*** .718 .789*** -.151***
Gender .020 .037ns .000 .000ns .018 .031ns -.003 -.006ns
Age .059 .041* -.009 -.006ns .064 .043* .057 .038ns
Familiarity -.070 -.044** .048 .027ns -.014 -.007ns .039 .020ns
Total effect (H4)
PBG→LOY .714 .565*** .670 .462*** .103*** .604 .468*** .418 .339*** .129***
Diff. within groups Diff. within groups
Lower Higher Lower Higher
Indirect effects (H5) bp bp bp bp
PBG→FV→LOY .359 .284*** .241 .166*** .003ns .212 .165*** .090 .073***
PBG→PV→LOY .355 .281*** .429 .296*** -.130*** .392 .304*** .329 .266*** -.139*** -.193***
Model fits: General model: China: CFI .948; TLI .937; RMSEA .081; SRMR .100; ²(112) = 581.879; SCF = 1.2408. Italy: CFI .939; TLI .926;
RMSEA .088; SRMR .105; ²(112) = 707.659; SCF = 1.1497. MNC origin: China: CFI .955; TLI .948; RMSEA .072; SRMR .109;
²(234) = 624.332; SCF = 1.2434. Italy: CFI .950; TLI .943; RMSEA .075; SRMR .093; ²(234) = 693.784; SCF = 1.1974. CE: China: CFI .940;
TLI .931; RMSEA .084; SRMR .105; ²(234) = 770.028; SCF = 1.2443. Italy: CFI .941; TLI .931; RMSEA .082; SRMR .107; ²(234) = 840.350;
SCF = 1.1762.
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value;
LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.
 23

Table E.8a
Test for Direct Effect using the Wald Test (Alternative Models; Further Countries)
China Italy
I. General model W df p W df p
PBG → LOY .732 1 ns 2.612 1 ns
II. Moderator: Foreign Domestic Foreign Domestic
MNC origin W df p W df p W df p W df p
PBG → LOY 2.781 1 ns .275 1 ns .636 1 ns 3.777 1 ns

III. Moderator: Lower Higher Lower Higher

Consumer ethnocentrism W df p W df p W df p W df p
PBG → LOY .355 1 ns 3.497 1 ns 3.042 1 ns 1.550 1 ns
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; PBG = Perceived brand globalness; LOY = Loyalty; W = Wald test statistic; df
= degrees of freedom.
 24

Table E.8b
Test for Direct Effect using the Wald Test (Alternative Models; Alternative MNCs in the Countries)
India Japan U.S.
I. General model W df p W df p W df p
PBG → LOY 1.401 1 ns 2.017 1 ns 1.767 1 ns
II. Moderator: Foreign Domestic Foreign Domestic Foreign Domestic
MNC origin W df p W df p W df p W df p W df p W df p
PBG → LOY 2.116 1 ns .774 1 ns .925 1 ns 3.685 1 ns .056 1 ns .000 1 ns

III. Moderator: Lower Higher Lower Higher Lower Higher

Consumer ethnocentrism W df p W df p W df p W df p W df p W df p
PBG → LOY 2.358 1 ns 3.714 1 ns .359 1 ns 2.953 1 ns .016 1 ns .938 1 ns
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; PBG = Perceived brand globalness; LOY = Loyalty; W = Wald test statistic; df = degrees of freedom.
 25

Table E.8c
Test for Direct Effect using the Wald Test (Alternative Models; Alternative MNCs in the
Countries)
China Italy
I. General model W df p W df p
PBG → LOY .903 1 ns .003 1 ns
II. Moderator: Foreign Domestic Foreign Domestic
MNC origin W df p W df p W df p W df p
PBG → LOY 3.612 1 ns .218 1 ns .038 1 ns 3.100 1 ns

III. Moderator: Lower Higher Lower Higher

Consumer ethnocentrism W df p W df p W df p W df p
PBG → LOY .002 1 ns .058 1 ns 3.625 1 ns 1.996 1 ns
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; PBG = Perceived brand globalness; LOY = Loyalty; W = Wald test statistic; df
= degrees of freedom.

Table E.9a
Test for Direct and Indirect Effects using Bootstrapping (Alternative Models; Further Coun-
tries)
China Italy
I. General model S.E. b Lowe Up- p S.E. b Lowe Up- p
r per r per
Direct effect
PBG → LOY .025 -.031 -.081 .018 ns .038 -.061 -.136 .010 ns
Indirect effects
PBG → FV → LOY .030 .191 .133 .249 *** .020 .119 .083 .160 ***
PBG → PV → LOY .034 .370 .307 .439 *** .035 .327 .261 .400 ***
II. Moderator:
MNC origin Foreign Domestic Foreign Domestic
S.E. b Lowe Up- p S.E. b Lowe Up- p S.E. b Lowe Up- p S.E. b Lowe Up- p
r per r per r per r per
Direct effect
PBG → LOY .031 -.042 -.107 .033 ns .041 -.020 -.073 .101 ns .043 .025 -.053 .106 ns .037 -.063 -.137 .011 ns
Indirect effects
PBG → FV → LOY .025 .298 .244 .344 *** .026 .098 .047 .163 *** .018 .222 .186 .254 *** .011 .086 .064 .107 ***
PBG → PV → LOY .037 .287 .231 .388 *** .043 .387 .316 .471 *** .013 .169 .145 .196 *** .039 .330 .260 .410 ***
III. Moderator:
Consumer Lower Higher Lower Higher
ethnocentrism S.E. b Lowe Up- p S.E. b Lowe Up- p S.E. b Lowe Up- p S.E. b Lowe Up- p
r per r per r per r per
Direct effect
PBG → LOY .035 -.018 -.082 .051 ns .031 -.054 -.120 .003 ns .038 -.069 -.140 .007 ns .036 -.038 -.100 .033 ns
Indirect effects
PBG → FV → LOY .018 .298 .259 .334 *** .030 .288 .229 .348 *** .027 .254 .205 .305 *** .029 .134 .074 .186 ***
PBG → PV → LOY .032 .295 .226 .356 *** .024 .238 .186 .279 *** .020 .203 .163 .241 *** .019 .189 .148 .225 ***
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychologi-
cal value; LOY = Loyalty; S.E. = standard error; b = unstandardized coefficients; Lower: Lower limit of confidence interval; Upper: Upper
limit of confidence interval; 95% confidence interval limits are shown; bootstrap samples = 5,000.
 26

Table E.9b
Test for Direct and Indirect Effects using Bootstrapping (Alternative Models; Alternative MNCs in the Countries)
India Japan U.S.
I. General model S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p
Direct effect
PBG → LOY .031 -.052 -.112 .010 ns .023 -.036 -.080 .013 ns .047 -.072 -.169 .011 ns
Indirect effects
PBG → FV → LOY .037 .278 .207 .354 *** .015 .077 .044 .105 *** .024 .077 .034 .135 ***
PBG → PV → LOY .037 .348 .279 .424 *** .027 .155 .105 .210 *** .055 .411 .307 .525 ***
II. Moderator:
MNC origin Foreign Domestic Foreign Domestic Foreign Domestic
S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p
Direct effect
PBG → LOY .048 -.088 -.190 .002 ns .039 -.037 -.114 .039 ns .051 -.042 -.139 .049 ns .031 -.058 -.122 .003 ns .067 .013 -.115 .139 ns .064 .000 -.136 .123 ns
Indirect effects
PBG → FV → LOY .052 .339 .244 .448 *** .020 .132 .092 .172 *** .033 .172 .108 .237 *** .024 .094 .052 .148 *** .061 .351 .236 .491 *** .050 .187 .099 .302 ***
PBG → PV → LOY .052 .323 .222 .432 *** .050 .342 .250 .448 *** .018 .162 .126 .199 *** .019 .169 .135 .209 *** .024 .150 .100 .200 *** .055 .289 .178 .384 ***
III. Moderator:
Consumer Lower Higher Lower Higher Lower Higher
ethnocentrism S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p S.E. b Lower Upper p
Direct effect
PBG → LOY .081 -.209 -.577 .048 ns .073 -.083 -.417 .021 ns .033 .020 -.051 .080 ns .056 -.095 -.222 .030 ns .075 .009 -.169 .134 ns .093 -.074 -.240 .112 ns
Indirect effects
PBG → FV → LOY .171 .494 .306 .599 *** .168 .291 .238 .470 *** .017 .092 .060 .124 *** .025 .097 .044 .152 *** .049 .227 .154 .332 *** .045 .073 .021 .165 ***
PBG → PV → LOY .145 .631 .335 .974 *** .113 .397 .288 .862 *** .035 .159 .089 .228 *** .018 .159 .121 .195 *** .049 .306 .201 .374 *** .052 .288 .210 .420 ***
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value; LOY = Loyalty; S.E. = standard error; b = unstandardized coefficients;
Lower: Lower limit of confidence interval; Upper: Upper limit of confidence interval; 95% confidence interval limits are shown; bootstrap samples = 5,000.
 27

Table E.9c
Test for Direct and Indirect Effects using Bootstrapping (Alternative Models; Alternative
MNCs in the Countries)
China Italy
I. General model S.E. b Lowe Up- p S.E. b Lowe Up- p
r per r per
Direct effect
PBG → LOY .026 -.034 -.083 .016 ns .044 -.003 -.089 .082 ns
Indirect effects
PBG → FV → LOY .027 .144 .090 .196 *** .033 .187 .123 .252 ***
PBG → PV → LOY .035 .374 .308 .445 *** .028 .281 .226 .337 ***
II. Moderator:
MNC origin Foreign Domestic Foreign Domestic
S.E. b Lowe Up- p S.E. b Lowe Up- p S.E. b Lowe Up- p S.E. b Lowe Up- p
r per r per r per r per
Direct effect
PBG → LOY .027 -.048 -.100 .005 ns .033 -.020 -.086 .043 ns .051 .008 -.100 .111 ns .044 -.076 -.163 .010 ns
Indirect effects
PBG → FV → LOY .024 .286 .238 .334 *** .027 .109 .056 .163 *** .042 .291 .206 .368 *** .027 .355 .302 .408 ***
PBG → PV → LOY .025 .282 .232 .331 *** .041 .372 .295 .455 *** .017 .212 .179 .249 *** .010 .082 .062 .101 ***
III. Moderator:
Consumer Lower Higher Lower Higher
ethnocentrism S.E. b Lowe Up- p S.E. b Lowe Up- p S.E. b Lowe Up- p S.E. b Lowe Up- p
r per r per r per r per
Direct effect
PBG → LOY .058 .005 -.107 .112 ns .038 .007 -.074 .082 ns .037 -.063 -.141 .004 ns .037 -.044 -.128 .019 ns
Indirect effects
PBG → FV → LOY .063 .282 .146 .403 *** .031 .162 .104 .226 *** .030 .158 .101 .219 *** .025 .092 .044 .141 ***
PBG → PV → LOY .072 .280 .117 .400 *** .025 .296 .252 .352 *** .012 .122 .099 .148 *** .033 .234 .161 .301 ***
Note: * p < .05; ** p < .01; *** p < .001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychologi-
cal value; LOY = Loyalty; S.E. = standard error; b = unstandardized coefficients; Lower: Lower limit of confidence interval; Upper: Upper
limit of confidence interval; 95% confidence interval limits are shown; bootstrap samples = 5,000.
 28

Web Appendix F. Latent Interaction Models of Consumer Ethnocentrism

We additionally tested the effect of consumer ethnocentrism on the PBG-value links (see Tables
F.1a-d) and PBG-value-loyalty pathways (see Tables F.1e-h) as a continuous variable. We further
tested for a three-way interaction comparing the PBG-value links for foreign vs. domestic
MNCs (see Tables F.2a-d). We used the residual centering approach suggested by Little,
Bovaird, and Widaman (2006). The residual centering approach has been shown to be of meth-
odological superiority compared to other approaches in terms of model fit and standard errors
(Steinmetz, Davidov, and Schmidt 2011).
Table F.1a
Results of Latent Interaction (Moderated PBG-value Links)
India Japan U.S.
beta bp beta bp beta bp
Moderator:
Consumer ethnocentrism
PBG→FV .667 .935 *** .309 .329 *** .511 .597 ***
PBG→PV .642 .938 *** .311 .331 *** .527 .620 ***
FV→LOY .516 .907 *** .432 .711 *** .269 .291 ***
PV→LOY .533 .900 *** .648 1.065 *** .482 .518 ***
CE → FV .166 .233 *** -.017 -.019 ns .137 .160 ***
CE → PV .268 .392 *** .023 .025 ns .109 .128 **
PBG × CE → FV -.133 -.187 *** -.134 -.142 ** .002 .002 ns
PBG × CE → PV -.135 -.198 *** -.151 -.161 ** -.011 -.012 ns
Gender -.012 -.059 ns .030 .104 ns -.032 -.081 ns
Age .038 .074 * .032 .043 ns .013 .013 ns
Familiarity .002 .002 ns .009 .008 ns .061 .044 *
Total effect
PBG→LOY .686 1.694 *** .335 .586 *** .391 .496 ***
Indirect effects
PBG→FV→LOY .344 .849 *** .134 .234 *** .137 .174 ***
PBG→PV→LOY .342 .845 *** .201 .352 *** .254 .322 ***
Model fits: India: CFI 0.983; TLI 0.981; RMSEA 0.030; SRMR 0.059; ²(450) = 750.174; SCF = 1.3421; Japan: CFI 0.975; TLI 0.970; RMSEA
0.034; SRMR 0.094; ²(450) = 907.254; SCF = 1.4685; U.S.: CFI 0.965; TLI 0.959; RMSEA 0.044; SRMR 0.087; ²(450) = 1103.849; SCF = 1.3647.
Note: * p<0.05; ** p<0.01; *** p<0.001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value;
LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.

Table F.1b
Results of Latent Interaction (Moderated PBG-value Links; Alternative MNCs in the Coun-
tries)
India Japan U.S.
beta bp beta bp beta bp
Moderator:
Consumer ethnocentrism
PBG→FV .632 .870 *** .382 .403 *** .411 .456 ***
PBG→PV .626 .904 *** .328 .343 *** .469 .540 ***
FV→LOY .496 .758 *** .304 .517 *** .147 .199 ***
PV→LOY .477 .696 *** .742 1.275 *** .698 .911 ***
CE → FV .198 .273 *** .074 .078 ns .144 .160 ***
CE → PV .270 .390 *** .077 .080 ns .176 .203 ***
PBG × CE → FV -.117 -.162 *** -.236 -.249 *** .061 .068 ns
PBG × CE → PV -.130 -.187 *** -.142 -.148 * .045 .052 ns
Gender .001 .006 ns .038 .135 ns -.006 -.018 ns
Age .049 .082 * .045 .061 ns -.074 -.084 **
Familiarity .037 .049 ns .047 .047 ns -.040 -.030 ns
Total effect
PBG→LOY .612 1.289 *** .359 .646 *** .388 .583 ***
Indirect effects
PBG→FV→LOY .313 .660 *** .116 .209 *** .060 .091 ***
PBG→PV→LOY .299 .629 *** .243 .437 *** .327 .492 ***
Model fits: India: CFI 0.981; TLI 0.978; RMSEA 0.030; SRMR 0.060; ²(450) = 752.576; SCF = 1.3826; Japan: CFI 0.979; TLI 0.976; RMSEA
0.030; SRMR 0.099; ²(450) = 723.181; SCF = 1.6041; U.S.: CFI 0.979; TLI 0.975; RMSEA 0.034; SRMR 0.100; ²(450) = 800.677; SCF = 1.5035.
Note: * p<0.05; ** p<0.01; *** p<0.001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value;
LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.
 29

Table F.1c
Results of Latent Interaction (Moderated PBG-value Links; Further Countries)
China Italy
beta bp beta bp
Moderator:
Consumer ethnocentrism
PBG→FV .724 1.021 *** .361 .396 ***
PBG→PV .680 .897 *** .578 .732 ***
FV→LOY .332 .614 *** .356 .482 ***
PV→LOY .717 1.414 *** .561 .659 ***
CE → FV .122 .172 *** .188 .206 ***
CE → PV .191 .251 *** .197 .249 ***
PBG × CE → FV -.129 -.181 *** -.041 -.045 ns
PBG × CE → PV -.148 -.196 *** -.036 -.045 ns
Gender -.032 -.166 ns -.054 -.159 *
Age -.002 -.004 ns .012 .013 ns
Familiarity -.035 -.060 ns .026 .022 ns
Total effect
PBG→LOY .728 1.894 *** .453 .673 ***
Indirect effects
PBG→FV→LOY .241 .627 *** .128 .191 ***
PBG→PV→LOY .487 1.268 *** .324 .482 ***
Model fits: China: CFI 0.977; TLI 0.973; RMSEA 0.036; SRMR 0.075; ²(450) = 827.285; SCF = 1.4551; Italy: CFI 0.981; TLI 0.977; RMSEA
0.033; SRMR 0.074; ²(450) = 804.944; SCF = 1.4108.
Note: * p<0.05; ** p<0.01; *** p<0.001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value;
LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.

Table F.1d
Results of Latent Interaction (Moderated PBG-value Links; Alternative MNCs in the Coun-
tries)
China Italy
beta bp beta bp
Moderator:
Consumer ethnocentrism
PBG→FV .744 1.073 *** .555 .669 ***
PBG→PV .701 .962 *** .403 .447 ***
FV→LOY .261 .444 *** .339 .377 ***
PV→LOY .755 1.351 *** .501 .606 ***
CE → FV .108 .156 ** .095 .114 ***
CE → PV .122 .168 *** .183 .203 ***
PBG × CE → FV -.110 -.159 ** -.041 -.049 ns
PBG × CE → PV -.182 -.249 *** -.030 -.034 ns
Gender -.006 -.032 ns -.029 -.079 ns
Age -.016 -.030 ns -.038 -.039 ns
Familiarity -.020 -.035 ns .029 .024 ns
Total effect
PBG→LOY .723 1.777 *** .390 .523 ***
Indirect effects
PBG→FV→LOY .194 .477 *** .188 .252 ***
PBG→PV→LOY .529 1.300 *** .202 .271 ***
Model fits: China: CFI 0.977; TLI 0.973; RMSEA 0.036; SRMR 0.075; ²(450) = 827.285; SCF = 1.4551; Italy: CFI 0.979; TLI 0.976; RMSEA
0.034; SRMR 0.086; ²(450) = 816.498; SCF = 1.4622.
Note: * p<0.05; ** p<0.01; *** p<0.001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value;
LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.
 30

Table F.1e
Results of Latent Interaction (Moderated PBG-value-loyalty Pathways)
India Japan U.S.
beta bp beta bp beta bp
Moderator:
Consumer ethnocentrism
PBG→FV .543 .586 *** .345 .326 *** .287 .314 ***
PBG→PV .744 .695 *** .883 .406 *** .567 .259 ***
FV→LOY .429 1.049 *** .291 .616 *** .150 .192 *
PV→LOY .540 1.524 *** .608 2.642 *** .521 1.595 ***
CE → FV .703 .758 *** .418 .394 *** .421 .461 ***
CE → PV -.526 -.492 *** -1.747 -.803 *** -1.942 -.887 ***
PBG × CE → FV -.151 -.163 *** -.322 -.304 *** .001 .001 ns
PBG × CE → PV -.175 -.164 *** -.490 -.225 *** -.012 -.006 ns
CE → LOY -.031 -.081 ns -.031 -.061 ns .148 .207 ***
FV × CE → LOY -.110 -.291 * -.083 -.166 ns -.050 -.071 ns
PV × CE → LOY .109 .289 * .074 .148 ns .072 .100 ns
Gender -.007 -.036 ns .027 .108 ns -.032 -.090 ns
Age .032 .066 * .020 .031 ns .010 .011 ns
Familiarity .004 .008 ns .008 .009 ns .042 .034 ns
Total effect
PBG→LOY .634 1.674 *** .637 1.272 *** .339 .473 ***
Indirect effects
PBG→FV→LOY .233 .614 *** .101 .201 *** .043 .060 *
PBG→PV→LOY .402 1.060 *** .536 1.072 *** .295 .413 ***
Model fits: India: CFI 0.866; TLI 0.853; RMSEA 0.060; SRMR 0.057; ²(450) = 6880.144; SCF = 1.7626; Japan: CFI 0.873; TLI 0.861; RMSEA
0.054; SRMR 0.053; ²(450) = 6734.889; SCF = 1.9565; U.S.: CFI 0.845; TLI 0.831; RMSEA 0.066; SRMR 0.048; ²(450) = 8174.487; SCF =
1.8599.
Note: * p<0.05; ** p<0.01; *** p<0.001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value;
LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.

Table F.1f
Results of Latent Interaction (Moderated PBG-value-loyalty Pathways; Alternative MNCs)
India Japan U.S.
beta bp beta bp beta bp
Moderator:
Consumer ethnocentrism
PBG→FV .490 .519 *** .408 .343 *** .349 .314 ***
PBG→PV .752 .674 *** .281 .198 *** .373 .259 ***
FV→LOY .378 .804 *** .130 .316 * .166 .297 ***
PV→LOY .556 1.399 *** .753 2.186 ** .566 1.314 ***
CE → FV .740 .784 *** .793 .666 *** .733 .659 ***
CE → PV -.584 -.524 *** -.969 -.681 *** -.968 -.672 ***
PBG × CE → FV -.136 -.144 *** -.226 -.190 *** .055 .049 ns
PBG × CE → PV -.147 -.131 *** -.191 -.134 *** .034 .024 ns
CE → LOY -.021 -.047 ns -.005 -.009 ns .144 .233 ***
FV × CE → LOY -.152 -.342 ** -.103 -.211 ns .038 .061 ns
PV × CE → LOY .152 .344 ** .079 .160 ns .025 .040 ns
Gender .002 .008 ns .038 .154 ns -.008 -.027 ns
Age .043 .077 * .045 .069 ns -.066 -.081 **
Familiarity .037 .053 ns .037 .042 ns -.044 -.036 ns
Total effect
PBG→LOY .603 1.360 *** .265 .541 *** .269 .433 ***
Indirect effects
PBG→FV→LOY .185 .417 *** .053 .108 * .058 .093 ***
PBG→PV→LOY .418 .943 *** .212 .433 *** .211 .340 ***
Model fits: India: CFI 0.877; TLI 0.866; RMSEA 0.056; SRMR 0.053; ²(450) = 6314.994; SCF = 1.7495; Japan: CFI 0.858; TLI 0.845; RMSEA
0.055; SRMR 0.063; ²(450) = 5824.988; SCF = 2.1022; U.S.: CFI 0.833; TLI 0.818; RMSEA 0.075; SRMR 0.044; ²(450) = 9084.574; SCF =
1.7752.
Note: * p<0.05; ** p<0.01; *** p<0.001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value;
LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.
 31

Table F.1g
Results of Latent Interaction (Moderated PBG-value-loyalty Pathways; Further Countries)
China Italy
beta bp beta bp
Moderator:
Consumer ethnocentrism
PBG→FV .744 .763 *** .300 .318 ***
PBG→PV .600 .564 *** .420 .262 ***
FV→LOY .163 .447 ** .283 .438 ***
PV→LOY .782 2.341 *** .538 1.416 ***
CE → FV .684 .702 *** .498 .529 ***
CE → PV -.670 -.630 *** -1.228 -.766 ***
PBG × CE → FV -.126 -.130 *** -.036 -.038 ns
PBG × CE → PV -.118 -.111 *** -.032 -.020 ns
CE → LOY .017 .047 ns .072 .118 **
FV × CE → LOY -.086 -.242 * -.008 -.013 ns
PV × CE → LOY .095 .268 ** .033 .055 ns
Gender -.032 -.181 * -.047 -.154 ns
Age .002 .004 ns .017 .021 ns
Familiarity -.037 -.068 * .026 .025 ns
Total effect
PBG→LOY .591 1.661 *** .311 .510 ***
Indirect effects
PBG→FV→LOY .121 .341 ** .085 .139 ***
PBG→PV→LOY .470 1.320 *** .226 .371 ***
Model fits: China: CFI 0.861; TLI 0.848; RMSEA 0.065; SRMR 0.068; ²(450) = 7102.444; SCF = 1.8297; Italy: CFI 0.836; TLI 0.821; RMSEA
0.069; SRMR 0.054; ²(450) = 8372.568; SCF = 1.8811.
Note: * p<0.05; ** p<0.01; *** p<0.001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value;
LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.

Table F.1h
Results of Latent Interaction (Moderated PBG-value-loyalty Pathways; Alternative MNCs in
the Countries)
China Italy
beta bp beta bp
Moderator:
Consumer ethnocentrism
PBG→FV .840 .756 *** .406 .301 ***
PBG→PV .579 .609 *** .254 .239 ***
FV→LOY .237 .682 *** .327 .648 ***
PV→LOY .700 1.724 *** .425 .663 ***
CE → FV .883 .794 *** 1.145 .848 ***
CE → PV -.464 -.488 *** -.381 -.359 ***
PBG × CE → FV -.161 -.145 *** -.037 -.027 ns
PBG × CE → PV -.168 -.177 *** -.025 -.023 ns
CE → LOY .007 .018 ns .109 .160 ***
FV × CE → LOY -.105 -.273 * -.023 -.033 ns
PV × CE → LOY .128 .332 *** .041 .061 ns
Gender -.009 -.047 ns -.021 -.062 ns
Age -.013 -.025 ns -.027 -.030 ns
Familiarity -.021 -.040 ns .027 .024 ns
Total effect
PBG→LOY .604 1.567 *** .241 .353 ***
Indirect effects
PBG→FV→LOY .199 .516 *** .133 .195 ***
PBG→PV→LOY .405 1.051 *** .108 .158 ***
Model fits: China: CFI 0.835; TLI 0.820; RMSEA 0.069; SRMR 0.069; ²(450) = 7917.185; SCF = 1.8489; Italy: CFI 0.858; TLI 0.845; RMSEA
0.065; SRMR 0.051; ²(450) = 7485.939; SCF = 1.8645.
Note: * p<0.05; ** p<0.01; *** p<0.001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value;
LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.
 32

Table F.2a
Results of Latent Three-Way Interaction (Moderated PBG-value Links)
India Japan U.S.
beta bp beta bp beta bp
Moderator:
Consumer ethnocentrism Foreign Foreign Foreign
PBG→FV .726 1.086 *** .360 .410 *** .474 .567 ***
PBG→PV .677 .999 *** .362 .422 *** .541 .650 ***
FV→LOY .522 .918 *** .470 .828 *** .429 .461 ***
PV→LOY .530 .944 *** .620 1.065 *** .327 .350 ***
CE → FV .128 .192 ** .099 .112 ns .256 .306 ***
CE → PV .222 .328 *** .126 .147 * .104 .125 ***
PBG × CE → FV -.102 -.152 * -.318 -.362 *** .059 .070 ns
PBG × CE → PV -.114 -.169 * -.365 -.426 *** -.020 -.024 ns
Gender -.002 -.013 ns .000 -.002 ns -.098 -.252 **
Age .011 .024 ns .017 .027 ns .069 .068 ns
Familiarity .005 .009 ns -.036 -.038 ns .037 .027 ns
Total effect
PBG→LOY .738 1.941 *** .394 .790 *** .381 .489 ***
Indirect effects
PBG→FV→LOY .379 .997 *** .169 .340 *** .203 .261 ***
PBG→PV→LOY .359 .944 *** .224 .450 *** .177 .228 ***
Moderator:
Consumer ethnocentrism Domestic Domestic Domestic
PBG→FV .604 .824 *** .278 .295 *** .496 .572 ***
PBG→PV .589 .875 *** .248 .265 *** .488 .558 ***
FV→LOY .501 .855 *** .387 .599 *** .038 .046 ns
PV→LOY .536 .839 *** .642 .986 *** .684 .826 ***
CE → FV .201 .274 *** .156 .165 ** .000 .000 ns
CE → PV .330 .491 *** .238 .255 *** .035 .040 ns
PBG × CE → FV .002 .002 ns .121 .128 ns -.024 -.028 ns
PBG × CE → PV .013 .020 ns .074 .079 ns -.012 -.014 ns
Gender -.025 -.118 ns .053 .173 ns .034 .096 ns
Age .061 .113 * .055 .069 ns -.028 -.030 ns
Familiarity -.006 -.009 ns .051 .047 ns .093 .075 **
Total effect
PBG→LOY .619 1.438 *** .267 .438 *** .352 .487 ***
Indirect effects
PBG→FV→LOY .303 .704 *** .107 .176 *** .019 .026 ns
PBG→PV→LOY .316 .734 *** .159 .261 *** .333 .461 ***
Model fits: Foreign: India: CFI 0.966; TLI 0.961; RMSEA 0.045; SRMR 0.081; ²(450) = 785.809; SCF = 1.3122; Japan: CFI 0.887; TLI 0.869;
RMSEA 0.084; SRMR 0.108; ²(450) = 1647.892; SCF = 1.3616; U.S.: CFI 0.959; TLI 0.952; RMSEA 0.050; SRMR 0.093; ²(450) = 865.345; SCF
= 1.2857; Domestic: India: CFI 0.966; TLI 0.961; RMSEA 0.043; SRMR 0.087; ²(450) = 760.806; SCF = 1.3013; Japan: CFI 0.924; TLI 0.911;
RMSEA 0.058; SRMR 0.059; ²(450) = 1181.575; SCF = 1.4818; U.S.: CFI 0.964; TLI 0.958; RMSEA 0.045; SRMR 0.107; ²(450) = 799.084; SCF
= 1.4016.
Note: * p<0.05; ** p<0.01; *** p<0.001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value;
LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.
 33

Table F.2b
Results of Latent Three-Way Interaction (Moderated PBG-value Links; Alternative MNCs in
the Countries)
India Japan U.S.
beta bp beta bp beta bp
Moderator:
Consumer ethnocentrism Foreign Foreign Foreign
PBG→FV .691 .993 *** .406 .442 *** .431 .481 ***
PBG→PV .669 .965 *** .377 .406 *** .518 .618 ***
FV→LOY .442 .647 *** .227 .374 *** .135 .184 ***
PV→LOY .559 .815 *** .766 1.274 *** .709 .908 ***
CE → FV .136 .195 ** .119 .130 * .106 .118 *
CE → PV .185 .266 *** .099 .107 ns .182 .217 ***
PBG × CE → FV -.095 -.137 * -.328 -.357 *** .073 .081 ns
PBG × CE → PV -.093 -.134 * .-228 -.245 ** .037 .045 ns
Gender .035 .146 ns .033 .120 ns .030 .091 ns
Age .042 .070 ns .013 .017 ns -.027 -.032 ns
Familiarity .070 .095 ns .075 .076 ns -.047 -.037 ns
Total effect
PBG→LOY .680 1.429 *** .381 .683 *** .425 .650 ***
Indirect effects
PBG→FV→LOY .306 .643 *** .092 .165 *** .058 .089 ***
PBG→PV→LOY .374 .786 *** .288 .517 *** .367 .561 ***
Moderator:
Consumer ethnocentrism Domestic Domestic Domestic
PBG→FV .589 .805 *** .366 .399 *** .363 .397 ***
PBG→PV .600 .900 *** .279 .297 *** .399 .442 ***
FV→LOY .472 .748 *** .352 .605 *** .165 .225 ***
PV→LOY .545 .785 *** .715 1.263 *** .687 .927 ***
CE → FV .229 .312 *** .141 .153 * .181 .197 **
CE → PV .324 .486 *** .156 .166 * .172 .191 **
PBG × CE → FV .013 .018 ns .104 .114 ns .065 .071 ns
PBG × CE → PV .014 .021 ns .096 .102 ns .047 .052 ns
Gender -.038 -.165 ns .045 .170 ns -.053 -.159 ns
Age .054 .092 ns .066 .094 * -.124 -.137 ***
Familiarity .007 .010 ns .025 .026 ns -.045 -.033 ns
Total effect
PBG→LOY .605 1.309 *** .328 .616 *** .334 .499 ***
Indirect effects
PBG→FV→LOY .278 .602 *** .129 .242 *** .060 .089 ***
PBG→PV→LOY .327 .707 *** .199 .375 *** .274 .409 ***
Model fits: Foreign: India: CFI 0.976; TLI 0.972; RMSEA 0.035; SRMR 0.081; ²(450) = 645.686; SCF = 1.4127; Japan: CFI 0.918; TLI 0.904;
RMSEA 0.068; SRMR 0.101; ²(450) = 1033.728; SCF = 1.6087; U.S.: CFI 0.971; TLI 0.966; RMSEA 0.042; SRMR 0.102; ²(450) = 715.194; SCF
= 1.4239; Domestic: India: CFI 0.964; TLI 0.958; RMSEA 0.044; SRMR 0.085; ²(450) = 778.130; SCF = 1.2976; Japan: CFI 0.953; TLI 0.946;
RMSEA 0.045; SRMR 0.106; ²(450) = 824.987; SCF = 1.4838; U.S.: CFI 0.959; TLI 0.952; RMSEA 0.050; SRMR 0.124; ²(450) = 819.455; SCF =
1.4097.
Note: * p<0.05; ** p<0.01; *** p<0.001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value;
LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.
 34

Table F.2c
Results of Latent Three-Way Interaction (Moderated PBG-value Links; Further Countries)
China Italy
beta bp beta bp
Moderator:
Consumer ethnocentrism Foreign Foreign
PBG→FV .700 .969 *** .543 .664 ***
PBG→PV .636 .814 *** .654 .901 ***
FV→LOY .371 .668 *** .254 .333 ***
PV→LOY .686 1.336 *** .640 .746 ***
CE → FV .058 .081 ns .165 .202 ***
CE → PV .079 .101 ns .178 .245 ***
PBG × CE → FV -.128 -.177 ** .010 .012 ns
PBG × CE → PV -.126 -.162 * .000 .000 ns
Gender -.042 -.210 ns -.066 -.211 *
Age .015 .030 ns -.054 -.064 ns
Familiarity -.030 -.049 ns .047 .042 ns
Total effect
PBG→LOY .696 1.735 *** .556 .893 ***
Indirect effects
PBG→FV→LOY .260 .647 *** .138 .221 ***
PBG→PV→LOY .436 1.088 *** .419 .672 ***
Moderator:
Consumer ethnocentrism Domestic Domestic
PBG→FV .719 1.034 *** .270 .295 ***
PBG→PV .719 1.040 *** .509 .628 ***
FV→LOY .403 .648 *** .451 .599 ***
PV→LOY .621 .994 *** .457 .538 ***
CE → FV .034 .048 ns .259 .283 ***
CE → PV .085 .123 ns .258 .318 ***
PBG × CE → FV -.001 -.002 ns -.090 -.098 ns
PBG × CE → PV -.038 -.056 ns -.097 -.120 ns
Gender -.019 -.091 ns -.033 -.097 ns
Age -.023 -.041 ns .090 .102 *
Familiarity -.044 -.066 ns -.012 -.011 ns
Total effect
PBG→LOY .736 1.704 *** .355 .515 ***
Indirect effects
PBG→FV→LOY .289 .670 *** .122 .177 ***
PBG→PV→LOY .446 1.034 *** .233 .338 ***
Model fits: Foreign: China: CFI 0.962; TLI 0.955; RMSEA 0.047; SRMR 0.095; ²(450) = 816.848; SCF = 1.4652; Italy: CFI 0.966; TLI 0.960;
RMSEA 0.046; SRMR 0.104; ²(450) = 780.794; SCF = 1.3672; Domestic: China: CFI 0.969; TLI 0.964; RMSEA 0.044; SRMR 0.092; ²(450) =
699.937; SCF = 1.3541; Italy: CFI 0.941; TLI 0.931; RMSEA 0.060; SRMR 0.107; ²(450) = 1051.120; SCF = 1.3527.
Note: * p<0.05; ** p<0.01; *** p<0.001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value;
LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.
 35

Table F.2d
Results of Latent Three-Way Interaction (Moderated PBG-value Links; Alternative MNCs in
the Countries)
China Italy
beta bp beta bp
Moderator:
Consumer ethnocentrism Foreign Foreign
PBG→FV .761 1.142 *** .688 .954 ***
PBG→PV .728 1.043 *** .558 .687 ***
FV→LOY .210 .327 *** .392 .385 ***
PV→LOY .773 1.258 *** .418 .463 ***
CE → FV .073 .110 ns .013 .019 ns
CE → PV .153 .220 ** .151 .186 **
PBG × CE → FV -.093 -.139 * -.077 -.106 ns
PBG × CE → PV -.095 -.136 * -.015 -.018 ns
Gender -.006 -.028 ns -.038 -.103 ns
Age -.011 -.020 ns -.074 -.075 *
Familiarity -.011 -.019 ns .047 .036 ns
Total effect
PBG→LOY .723 1.685 *** .503 .686 ***
Indirect effects
PBG→FV→LOY .160 .373 *** .269 .368 ***
PBG→PV→LOY .562 1.312 *** .233 .318 ***
Moderator:
Consumer ethnocentrism Domestic Domestic
PBG→FV .719 1.034 *** .456 .519 ***
PBG→PV .706 1.000 *** .201 .210 ***
FV→LOY .468 .742 *** .300 .350 ***
PV→LOY .561 .903 *** .549 .698 ***
CE → FV .016 .023 ns .161 .183 **
CE → PV .084 .119 ns .213 .222 ***
PBG × CE → FV .002 .003 ns -.121 -.137 ns
PBG × CE → PV -.008 -.011 ns -.048 -.050 ns
Gender -.010 -.045 ns -.017 -.045 ns
Age -.025 -.044 ns .005 .005 ns
Familiarity -.039 -.059 ns -.002 -.002 ns
Total effect
PBG→LOY .733 1.669 *** .247 .328 ***
Indirect effects
PBG→FV→LOY .336 .766 *** .137 .182 ***
PBG→PV→LOY .396 .903 *** .110 .146 ***
Model fits: Foreign: China: CFI 0.976; TLI 0.972; RMSEA 0.036; SRMR 0.082; ²(450) = 671.043; SCF = 1.5038; Italy: CFI 0.966; TLI 0.960;
RMSEA 0.046; SRMR 0.104; ²(450) = 780.794; SCF = 1.3672; Domestic: China: CFI 0.977; TLI 0.973; RMSEA 0.038; SRMR 0.084; ²(450) =
639.610; SCF = 1.3392; Italy: CFI 0.958; TLI 0.951; RMSEA 0.052; SRMR 0.105; ²(450) = 879.592; SCF = 1.3281.
Note: * p<0.05; ** p<0.01; *** p<0.001; ns = Not significant; PBG = Perceived brand globalness; FV = Functional value; PV = Psychological value;
LOY = Loyalty; b = unstandardized coefficients; beta = standardized coefficients.
 36

REFERENCES

Anderson, James C. and David W. Gerbing (1988), "Structural Equation Modeling in

Practice: A Review and Recommended Two-Step Approach," Psychological Bulletin, 103
(3), 411–23.

Antonakis, John, Samuel Bendahan, Philippe Jacquart, and Rafael Lalive (2014), "Causality
and Endogeneity: Problems and Solutions," in The Oxford Handbook of Leadership and
Organizations, David D. Day (Ed.). New York: USA: Oxford University Press.

---- (2010), "On Making Causal Claims: A Review and Recommendations," The Leadership
Quarterly, 21 (6), 1086–120.

Chen, Fang Fang (2007), "Sensitivity of Goodness of Fit Indexes to Lack of Measurement
Invariance," Structural Equation Modeling: A Multidisciplinary Journal, 14 (3), 464–504.

Cleveland, Mark and Michel Laroche (2007), "Acculturaton to the Global Consumer Culture:
Scale Development and Research Paradigm," Journal of Business Research, 60 (3), 249–59.

Fornell, Claes and David F. Larcker (1981), "Evaluating Structural Equation Models with
Unobservable Variables and Measurement Error," Journal of Marketing Research, 39–50.

Hair, Joseph F., William C. Black, Barry J. Babin, and Rolph E. Anderson (2014),
Multivariate Data Analysis (7 ed.). Upper Saddle River: Pearson.

Hausman, Jerry A. (1978), "Specification Tests in Econometrics," Econometrica, 46 (6),

1251-71.

Lindell, Michael K and David J Whitney (2001), "Accounting for Common Method Variance
in Cross-Sectional Research Designs," Journal of Applied Psychology, 86 (1), 114.

Little, Todd D., James A. Bovaird, and Keith F. Widaman (2006), "On The Merits of
Orthogonalizing Powered and Product Terms: Implications for Modeling Interactions Among
Latent Variables," Structural Equation Modeling: A Multidisciplinary Journal, 13 (4), 497–
519.

Rindfleisch, Aric, James E. Burroughs, and Nancy Wong (2009), "The Safety of Objects:
Materialism, Existential Insecurity, and Brand Connection," Journal of Consumer Research,
36 (1), 1-16.
 37

Steenkamp, Jan-Benedict E. M., Rajeev Batra, and Dana L. Alden (2003), "How Perceived
Brand Globalness Creates Brand Value," Journal of International Business Studies, 34 (1),
53–65.

Steinmetz, Holger, Eldad Davidov, and Peter Schmidt (2011), "Three Approaches to Estimate
Latent Interaction Effects: Intention and Perceived Behavioral Control in the Theory of
Planned Behavior," Methodological Innovations Online, 6 (1), 95–110.

Williams, Larry J., Nathan Hartman, and Flavia Cavazotte (2010), "Method Variance and
Marker Variables: A Review and Comprehensive CFA Marker Technique," Organizational
Research Methods, 13 (3), 477–514.