Brand Meaning and its Creation in a Cross-Cultural Context

DISSERTATION
of the University of St. Gallen,
Graduate School of Business Administration,
Economics, Law and Social Sciences (HSG)
to obtain the title of
Doctor Oeconomiae

submitted by

Ivana First

from

Croatia

Approved on the application of

Prof. Dr. Torsten Tomczak

and

Prof. Dr. Marcus Schögel

Dissertation no. 3616

Gutenberg AG, Schaan 2009

The University of St. Gallen, Graduate School of Business
Administration, Economics, Law and social Sciences (HSG) hereby
consents to the printing of the present dissertation, without hereby
expressing any opinion on the views herein expressed.

St. Gallen, May 19, 2009.

The President:

Prof. Ernst Mohr, PhD

Preface

Looking back at the process of writing my doctoral thesis, I feel it was a

challenging but rewarding task that required devotion, curiosity, and
diligence. It was a process that would have been impossible without the
professional and personal support of several people. I would like to take
this opportunity to express my gratitude to them.

Firstly, I would like to thank my supervisor, Prof. Dr. Torsten Tomczak,

and co-supervisor, Prof. Dr. Marcus Schögel, for giving me freedom as
well as guidance in this research. I would also like to thank Prof. Dr.
Bruno Grbac from the University of Rijeka (Croatia). Being a part of his
team provided me with continuous constructive discussion and helped me
to develop my research ideas.

Furthermore, I am very grateful to my colleague and friend Deepali Sinha,

a doctoral student herself, who was always available for ongoing concerns
ranging from prosaic to complex issues. I would also like to thank Dr.
Daniel Wentzel for his insight into various issues throughout the process
of writing this thesis, Dr. Darko Lonþariü for his advice on statistical
analyses, and Joanna Niederer for correcting the English side of things.

In addition, I would like to express my heartfelt appreciation to all the

participants in this research and especially to the interviewees. Without
their time and effort this research would not have been accomplished.

Finally, I am deeply grateful to my family, who gave me the opportunity

to work on this thesis and, together with my boyfriend Andrej, provided
me with backing and encouragement whenever I needed it. I dedicate this
work to them.

Rijeka, July 31, 2009.

Ivana First
 Table of Contents
1. INTRODUCTION ................................................................................ 1
1.1. SETTING THE SCENE ........................................................................ 1
1.2. PROBLEM DEFINITION AND PURPOSE OF THE STUDY ....................... 2
1.3. STRUCTURE OF THE THESIS ............................................................. 8
2. DEFINITIONS AND THEORETICAL BACKGROUND ............. 11
2.1. THE CONCEPT OF BRAND .............................................................. 11
2.1.1. Branding Concepts................................................................ 16
2.1.1.1. Brand Awareness ..................................................................... 16
2.1.1.2. Brand Associations .................................................................. 17
2.1.1.3. Brand Attributes....................................................................... 18
2.1.1.4. Brand Personality..................................................................... 19
2.1.1.5. Brand Beliefs ........................................................................... 20
2.1.1.6. Brand Knowledge .................................................................... 21
4.1.1.7. Brand Emotions ....................................................................... 22
2.1.1.8. Brand Attitudes ........................................................................ 23
2.1.1.9. Brand Benefits ......................................................................... 24
2.1.2. Brand Meaning ..................................................................... 25
2.1.3. How Brands Acquire Meanings ............................................ 30
2.1.3.1. Brand Strategies’ Role in Brand Meaning Creation ................. 30
2.1.3.2. Consumers’ Role in Brand Meaning Creation ......................... 33
2.1.3.4. Societies’ and Other Stakeholders’ Role in Brand Meaning
Creation ................................................................................................ 36
2.2. THE CONCEPT OF CULTURE ........................................................... 38
2.2.1. Culture, Cultural Differences and Boundaries of Culture .... 38
2.2.2. Cultural Dimensions ............................................................. 41
3. PREVIOUS RESEARCH AND HYPOTHETICAL MODEL ....... 51
3.1. THE INFLUENCE OF CULTURE ON BRAND MEANING ...................... 51
3.1.1. The Influence of Cultural on Brand Communication............ 52
3.1.2. The Influence of Cultural on Brand Communication
Interpretation .................................................................................. 53
3.1.3. Cultural Dimensions as an Antecedent to Brand Meaning... 57
3.2. FROM CULTURALLY DETERMINED NEEDS TO BRAND MEANINGS . 59
3.2.1. Consumer Needs ................................................................... 60
3.2.2. Brand Benefits....................................................................... 63
3.2.3. The Relationship between Culture, Consumer Needs, Brand
Benefits and Brand Meanings......................................................... 65
3.3. THE EFFECT OF BRAND USAGE AND BRAND AFFECTION ............... 74

I
4. RESEARCH METHODOLOGY...................................................... 77
4.1. RESEARCH APPROACH ................................................................... 77
4.2. RESEARCH DESIGN ........................................................................ 78
4.2.1. Research Methods................................................................. 78
4.2.2. Units of Analysis ................................................................... 82
4.2.3. Population and Sampling...................................................... 87
4.2.4. The Studied Countries........................................................... 88
4.3. VARIABLE OPERATIONALISATION ................................................. 89
4.3.1. The Construct of Culture ...................................................... 89
4.3.2. The Construct of Brand Meaning.......................................... 90
4.3.3. The Construct of Type of User .............................................. 93
5. EMPIRICAL RESEARCH ON BRAND MEANING..................... 94
5.1. PILOT RESEARCH ........................................................................... 94
5.1.1. Research Procedure.............................................................. 94
5.1. 2. Results.................................................................................. 95
5.2. MAIN PHASE RESEARCH ................................................................ 96
5.2.1. Research Procedure.............................................................. 96
5.2.1.1. Data gathering...........................................................................96
5.2.1.2. Data Analysis Procedure...........................................................97
5.2.3. Results................................................................................... 99
5.2.3.1. The Meanings of Coca-Cola ...................................................101
5.2.3.2. The Meanings of Google ........................................................103
5.2.3.3. The Meanings of Nike ............................................................105
5.2.3.4. The Meanings of Nokia ..........................................................107
5.2.3.4. The Meanings of Toyota.........................................................109
5.2.4. Discussion........................................................................... 111
5.2.4.1. Defining Brand Meaning ........................................................112
5.2.4.2. Differences in the Brand Meanings across Cultures ...............123
6. EMPIRICAL RESEARCH ON CROSS-CULTURAL
DIFFERENCES IN BRAND MEANINGS ........................................ 127
6.1. PILOT RESEARCH ......................................................................... 127
6.1.1. Research Procedure............................................................ 127
6.1.1.1. Questionnaire Design..............................................................127
6.1.1.2. Questionnaire Distribution......................................................129
6.1.2. Results................................................................................. 130
6.2. MAIN PHASE RESEARCH .............................................................. 136
6.2.1. Research Procedure............................................................ 136
6.2.2. Data Analysis Procedures................................................... 138
6.2.2.1. Data reduction method............................................................138
6.2.2.2. Analysis of Variance...............................................................140

II
 6.2.3. Results................................................................................. 141
6.2.3.1. Sample Characteristics........................................................... 141
6.2.3.2. Brand Meaning....................................................................... 143
6.2.3.3. Types of Users ....................................................................... 157
6.2.3.4. The Brand Meanings across the Cultures and across the Types of
Users ................................................................................................... 159
6.2.3.5. The Moderating Effect of Country......................................... 180
6.2.4. Discussion........................................................................... 185
6.2.4.1. Hypothesis 1 .......................................................................... 185
6.2.4.2. Hypothesis 2 .......................................................................... 185
6.2.4.3. Hypothesis 3 .......................................................................... 186
6.2.4.4. Hypothesis 4 .......................................................................... 188
7. CONCLUSION................................................................................. 189
7.1. SUMMARY OF THE FINDINGS........................................................ 189
7.2. MANAGERIAL IMPLICATIONS ....................................................... 191
7.2.1. Brand Meaning is a set of Cognitive and Affective
Experience-Based Brand Associations ......................................... 191
7.2.2. Consumer Understanding of Brands is Culturally Determined
...................................................................................................... 192
7.2.3. Negative Brand Meanings Relate to Brand Usage ............. 192
7.3. SCIENTIFIC IMPLICATIONS ........................................................... 193
7.3.1. Major Scientific Insights of the Research ........................... 193
7.3.1.1. Contribution to Brand Management Field.............................. 194
7.3.1.2. Contribution to the Field of Cross-cultural Research ............. 195
7.3.1.3. Contribution to Methodology in Consumer Research ............ 195
7.3.2. Limitations of the Research................................................. 196
7.3.3. Future Research.................................................................. 197
REFERENCES ..................................................................................... 200
APPENDICES .......................................................................................... I
APPENDIX 1. THE IMAGES COLLECTED IN THE INTERVIEWS ..................II
APPENDIX 2. THE PRODUCT CATEGORIES AND IMAGES USED IN THE
PILOT PHASE SURVEY .............................................................XXV
APPENDIX 3.THE QUESTIONNAIRE ………………..……………....XXVI
APPENDIX 4. THE STATISTICS ………………………..………...…XXXI

III
 List of Figures
FIGURE 1-1. STRUCTURE OF THE THESIS ................................................... 10
FIGURE 2-1. THE BRAND IDENTITY PRISM ................................................ 13
FIGURE 2-2. THE BRAND IDENTITY SYSTEM ............................................. 14
FIGURE 2-3. THE SYSTEMS MODEL OF BRAND ANTECEDENTS AND
CONSEQUENCES............................................................................... 15
FIGURE 2-4. TRADITIONAL AND CONTEMPORARY BRAND IMAGE CREATION
........................................................................................................ 31
FIGURE 2-5. THE STABILISING OF CULTURE PATTERNS ............................ 41
FIGURE 2-6. CULTURAL DIFFERENCES DISTRIBUTION .............................. 49
FIGURE 3-1. MAKING MEANING: COMMUNICATION, KNOWLEDGE AND
MEANING ........................................................................................ 55
FIGURE 3-2. CONSUMER NEEDS AND BRAND BENEFITS IN CREATING
BRAND MEANINGS .......................................................................... 60
FIGURE 3-3. A HYPOTHETICAL MODEL ..................................................... 76
THE FOLLOWING CHAPTER WILL PROVIDE AN OVERVIEW OF THE
METHODOLOGY USED IN THIS THESIS............................................... 76
FIGURE 4-1. A RESEARCH SEQUENCE ....................................................... 80

List of Tables
TABLE 2-1. AN OVERVIEW OF THE EXISTING BRAND MEANING
DEFINITIONS .................................................................................... 29
TABLE 2-2. AN OVERVIEW OF SOCIETAL DILEMMAS AND CULTURAL
DIMENSIONS .................................................................................... 44
TABLE 3-1. THE RELATIONSHIP BETWEEN CULTURAL DIMENSIONS,
CONSUMER NEEDS, BRAND BENEFITS AND BRAND MEANINGS ...... 73
TABLE 4-1. METHODOLOGICAL APPROACHES........................................... 79
TABLE 4-2. THE STUDIED BRANDS ........................................................... 84
TABLE 4-3. THE STUDIED COUNTRIES’ CULTURAL DIMENSIONS .............. 90
TABLE 5-1. THE PRIMARY MEANINGS FOR COCA-COLA PER COUNTRY . 103
TABLE 5-2. THE PRIMARY MEANINGS FOR GOOGLE PER COUNTRY ....... 105
TABLE 5-3. THE PRIMARY MEANINGS FOR NIKE PER COUNTRY ............. 107
TABLE 5-4. THE PRIMARY MEANINGS FOR NOKIA PER COUNTRY ........... 109
TABLE 5-5. THE PRIMARY MEANINGS FOR TOYOTA PER COUNTRY ........ 111
TABLE 5-6. THE TYPES OF RESPONSES FOR COCA-COLA ........................ 114
TABLE 5-7. THE TYPES OF RESPONSES FOR GOOGLE .............................. 116
TABLE 5-8. THE TYPES OF RESPONSES FOR NIKE .................................... 119

IV
TABLE 5-9. THE TYPES OF RESPONSES FOR NOKIA ................................. 120
TABLE 5-10. THE TYPES OF RESPONSES FOR TOYOTA ............................ 122
TABLE 5-11. BRAND MEANING IN RELATION TO THE OTHER BRAND
CONCEPTS ..................................................................................... 123
TABLE 5-12. THE IMPORTANCE OF COCA-COLA, GOOGLE AND NIKE ..... 125
TABLE 6-1. COCA-COLA’S GROUPS OF MEANINGS AND THEIR
HYPOTHESISED RELATIONSHIPS TO THE CULTURAL DIMENSIONS . 132
TABLE 6-2. GOOGLE’S GROUPS OF MEANINGS AND THEIR HYPOTHESISED
RELATIONSHIPS TO THE CULTURAL DIMENSIONS .......................... 134
TABLE 6-3. NIKE’S GROUPS OF MEANINGS AND THEIR HYPOTHESISED
RELATIONSHIPS TO THE CULTURAL DIMENSIONS .......................... 135
TABLE 6-4. THE SAMPLE CHARACTERISTICS........................................... 142
TABLE 6-5. THE MEDIANS FOR COCA-COLA’S MEANINGS IN THE OVERALL
SAMPLE ......................................................................................... 144
TABLE 6-6. THE PATTERN MATRIX FOR COCA-COLA ............................. 145
TABLE 6-7. THE RELIABILITY TESTS FOR THE TOP ITEMS IN COCA-COLA’S
COMPONENTS ............................................................................... 147
TABLE 6-8. THE PEARSON’S CORRELATION MATRIX FOR COCA-COLA’S
COMPONENTS ................................................................................ 147
TABLE 6-9. THE MEDIANS FOR GOOGLE’S MEANINGS IN THE OVERALL
SAMPLE ......................................................................................... 149
TABLE 6-10. THE PATTERN MATRIX FOR GOOGLE.................................. 150
TABLE 6-11. THE RELIABILITY TESTS FOR THE TOP ITEMS IN GOOGLE’S
COMPONENTS ................................................................................ 152
TABLE 6-12. THE PEARSON’S CORRELATION MATRIX FOR GOOGLE’S
COMPONENTS ................................................................................ 152
TABLE 6-13. THE MEDIANS FOR NIKE’S MEANINGS IN THE OVERALL
SAMPLE ......................................................................................... 154
TABLE 6-14. THE PATTERN MATRIX FOR NIKE ....................................... 155
TABLE 6-15. THE RELIABILITY TESTS FOR THE TOP ITEMS IN NIKE’S
COMPONENTS ................................................................................ 156
TABLE 6-16. THE PEARSON’S CORRELATION MATRIX FOR NIKE’S
COMPONENTS ................................................................................ 157
TABLE 6-17. THE DISTRIBUTION OF THE RESPONDENTS BASED ON THEIR
USAGE AND FEELINGS FOR THE BRANDS ....................................... 158
TABLE 6-18. THE EFFECTS OF CULTURE ON THE COCA-COLA MEANINGS
...................................................................................................... 160
TABLE 6-19. THE EFFECTS OF TYPE OF COCA-COLA USER ON THE COCA-
COLA MEANINGS ........................................................................... 166
TABLE 6-20. THE EFFECTS OF CULTURE ON THE GOOGLE MEANINGS .... 169

V
TABLE 6-21. THE EFFECTS OF TYPE OF GOOGLE USER ON THE GOOGLE
MEANINGS ..................................................................................... 173
TABLE 6-22. THE EFFECTS OF CULTURE ON THE NIKE MEANINGS.......... 176
TABLE 6-23. THE EFFECTS OF TYPE OF NIKE USER ON THE NIKE MEANINGS
...................................................................................................... 178
TABLE 6-24. THE PARTIAL CORRELATIONS BETWEEN THE COCA-COLA
MEANINGS AND AFFECTION TOWARDS COCA-COLA IN EACH OF THE
COUNTRIES .................................................................................... 181
TABLE 6-25. THE PARTIAL CORRELATIONS FOR AFFECTION TOWARDS
GOOGLE......................................................................................... 183
TABLE 6-26. THE PARTIAL CORRELATIONS FOR AFFECTION TOWARDS NIKE
...................................................................................................... 184
TABLE 6-27. THE OVERALL EFFECTS OF THE CULTURAL DIMENSIONS ON
THE BRAND MEANINGS ................................................................. 187

List of Graphs
GRAPH 5-1. THE PERCENTAGES OF SYMBOLIC IMAGES .......................... 100
GRAPH 5-2. THE MEANINGS OF COCA-COLA .......................................... 102
GRAPH 5-3. THE MEANINGS OF GOOGLE ................................................ 104
GRAPH 5-4. THE MEANINGS OF NIKE ...................................................... 106
GRAPH 5-5. THE MEANINGS OF NOKIA ................................................... 108
GRAPH 5-6. THE MEANINGS OF TOYOTA ................................................ 110
GRAPH 6-1. THE INTERACTION EFFECT ON THE GOOGLE INFORMATIVE
MEANING ....................................................................................... 174

VI
Abstract
In a quest to identify what brands represent to consumers, why consumers
like them and what benefits they provide them with, practitioners,
business researchers and consultants have investigated and established
many detailed concepts. Despite the rich body of literature on this issue,
the question of what brands actually mean to consumers still remains
unanswered. Simultaneously, another important line of research reveals
that despite tremendous globalisation, the cultures of different countries
are not (yet) homogeneous; and, because consumers across cultures differ,
another interesting question arises: is it possible for a brand to acquire a
globally unique meaning?
In light of these two lines of research, this thesis aims to discover, firstly,
what brands actually mean to consumers and consequently to identify the
brand meaning concept; and secondly, to investigate whether brand
meaning is consistent across cultures or affected by culturally determined
values, needs and underlying assumptions.
A series of in-depth interviews with a panel of international consumers
provided data to conceptualise brand meaning as a set of cognitive and
affective experience-based brand associations. A survey distributed to a
cross-cultural sample of consumers provided support for the hypothesis
that brand meanings differ across cultures, and in addition provided
substantial evidence that differences can be predicted by the cultural
dimensions of the given country.
The findings of the research have several scientific and managerial
implications that should serve to boost and direct future research, as well
as to improve managerial practices in terms of cross-cultural branding.

VII
Zusammenfassung
Seit Jahrzehnten haben sich Manager, Betriebswirte und Firmenberater
ausführlich mit der Frage des Markenmanagements beschäftigt, um
herauszufinden was Marken für Konsumenten bedeuten, warum
Konsumenten Marken mögen und was für Vorteile diese den
Konsumenten vermitteln. Aus diesem Grunde wurden vielerlei detaillierte
Konzepte entwickelt und untersucht. Nichtsdestotrotz die übergeordnete
Frage was Marken eigentlich für Konsumenten bedeuten ist immer noch
unbeantwortet.
Eine weitere Forschungslinie die für diese Studie von Relevanz ist zeigt
auf, dass trotz bedeutender Globalisierung Kulturen (noch) nicht
einheitlich sind. Konsumenten in unterschiedlichen Kulturen
unterscheiden sich immer noch und daher ist es aufschlussreich die Frage
zu untersuchen, ob eine weltweit einheitliche Markenwahrnehmung
überhaupt möglich ist.
Angesichts dieser beiden Forschungslinien hat diese Dissertation die
Absicht erstens, die Frage zu erläutern was Marken eigentlich für
Konsumenten bedeuten, und demzufolge das Konzept der
Markenbedeutung zu erörtern. Zweitens ist es die Absicht, zu untersuchen
ob Markenbedeutung kulturübergreifend einheitlich ist oder von
kulturellen Werten, Bedürfnissen und Annahmen beeinflusst wird.
Ausführliche Interviews mit einem Forum internationaler Konsumenten
stellen die Grundlage dar um Markenbedeutung als eine
Zusammenstellung kognitiver und emotionaler, erlebnisbasierter
Markenassoziationen zu konzeptionalisieren. Eine Umfrage, die an eine
Stichprobe von Konsumenten diverser Kulturen verteilt wurde, unterstützt
die Hypothese dass Markenbedeutung kulturell unterschiedlich ist und
gibt bedeutende Anzeichen, dass Unterschiedliche Markenauffassungen
vorhergesagt werden können aufgrund der kulturellen Dimensionen des
Landes.
Die Erkenntnisse dieser Forschungsarbeit haben mehrere
wissenschaftliche und betriebswirtschaftliche Implikationen, die dazu
dienen sollten zukünftige Forschung zu stärken und zu leiten, sowie die
Ausübung kulturübergreifenden Markenmanagements zu verbessern.

VIII
 1. Introduction

1.1. Setting the Scene

It is interesting to realise that some of the strongest brands belong to

products that are nothing special: syrup with soda – Coca-Cola; rolled
tobacco - Marlboro: or a clear spirit - Absolut (Hatch and Rubin 2006). It
is even more intriguing that consumers are willing to pay a 37% price
premium on a branded product over the same quality private label product.
At the same time, they are not willing to pay a price premium on a private
label product even if it is of a higher quality than a branded product
(Apelbaum Gerstner, and Naik 2003). Does this mean that consumers
behave irrationally?

On the other hand, does it mean that by creating strong brands companies
are no longer devoted to what has traditionally been thought of as a
marketing concern, namely, building long-term value and profitability by
satisfying customers? (Kay 2006). Have corporations forgotten what their
mission is? Are they the ones behaving irrationally?

The answer lies in understanding that consumers consume the brand as

much as the product itself (Schroeder 2005). They no longer consume
products for their utility, but because of their symbolic meaning
(Veloutsou and Moutinho 2008); they have needs and wants that will not
only be satisfied by product consumption, but also by the consumption of
a brand itself.

1
1.2. Problem Definition and Purpose of the Study

As a noun, brand refers to entities such as people, places, things, and

ideas; as a verb, it refers to those processes included in a firm’s endeavour
to make a product meaningful (Calder and Reagan 2001), such as naming
the product, targeting and positioning it, and communicating its benefits.

For consumers to enjoy the whole range of brand benefits, it is necessary

for brand managers to create and sustain strong brands. However, despite
widespread agreement that strong brands are a necessity for corporate
success, there is no consensus on the definition of a strong brand. Kapferer
(2004, p. 17) considers aided and unaided brand awareness to be strong
contributors to brand strength, Krishnan (1996) considers that the number
of associations evoked by a brand characterise the brand’s strength, and
Broniarczyk and Alba (1994) further consider brand unique propositions
to be the key to a strong image and consequently a strong brand equity.

Kay (2006) reasons that the literature is increasingly suggesting that the
strength of a brand is not based on creating a difference in consumer
perceptions, but is due to the meaning that the brand creates, and therefore
he points out that managing the brand meaning is an essential task in the
process towards achieving a successful, strong brand. He further stresses
that differentiation should not be disregarded, but it is not a sufficient
condition for a meaningful brand. Only relevant differentiation results in
added value and meaning for consumers. Such a view is also incorporated
in the brand strength dimension of the Brand Asset Valuator, one of the
most popular measures of brand market value (Aaker 1996, p. 304).

The meaning of a brand is not predetermined in some brand identity

booklet. Brand meaning is constructed in consumers’ minds. People
experience brands in different settings, situations and psychological states,

2
and therefore their understanding of brands varies. Even the same person
may not always see and use a brand in the same way or the same situation.
In other words, brands have multiple meanings for consumers. Those
brands that have the highest number of positive meanings will be the most
valuable and will provide the most benefit for consumers (Krishnan 1996).

Obviously, creating multiple brand meanings questions whether brand

consistency is really a necessary condition for a strong brand creation.
Aaker (1996) includes consistency over time in his ten guidelines for
building strong brands; whereas Keller (2003b) propagates the importance
of the consistency of a marketing programme in order to provide support
for strong brands. However, a deeper analysis of the ideas of these
branding gurus reveals that even they only have a conditional
understanding of consistency. Keller (2003b, p. 637), for example, speaks
of a consistent strategic direction and not necessarily of consistency in the
tactics employed. Similarly, Kapferer (2004, p. 222) speaks of the
necessity for a brand to evolve over time and to reposition itself, adjusting
to changes in consumers’ needs (although keeping the brand identity
unchanged). In line with the need for a longitudinal evolution of brand
meaning, as well as of cross-sectional differences in consumers’ needs,
Berthon, Pitt, and Campbell (2008) consider consistency appealing, but
stress that brands need to stay relevant in a dynamic marketplace and not
necessarily strive to mean the same thing when that meaning is no longer
relevant or is relevant in a different way to another group of consumers.

Some other recent studies provide new evidence and further challenge the
brand consistency concept. Lange and Dahlen (2003) prove that some
brands should not only gradually evolve into representing something else,
but should instead shock consumers by purposefully breaking the
consistency by using “strange” ads that are incongruent with the pictures
that consumers have of the known brand. In his innovative cultural

3
branding approach, Holt (2004) went one step further and provided
examples of how some of the strongest brands (Volkswagen, Harley-
Davidson, etc.) have needed several drastic changes in their positioning
during their lifetimes to portray a brand myth diametrically opposite to the
previous one, which in turn has created more identity value for the brand
and its consumers. However, Holt admits the risk of this unconventional
strategy and underlines that not all brands are predisposed to such actions.

Besides doubting the necessity of sustaining consistency, it is also relevant

to acknowledge how difficult it is to achieve it. This is particularly valid
for global brands. Ever since Levitt’s (1983) influential paper on the
globalisation of markets in which he states that “ancient differences in
national tastes or modes of doing business disappear”, scholars have
continuously discussed and analysed to which extent his proposition holds
true. Different authors have studied different aspects of globalisation and
there is a whole range of different conclusions as to whether globalisation
is desired (Goldsmith 2004; Luo, Sivakumar, and Liu 2005), whether it is
possible (Gram 2007), and what the consequences are for global markets
(Schuh 2007) and strategic and operational decision-making in
multinational companies (Szymanski, Bharadwaj, and Varadarajan 1993).

In all of the mentioned papers, globalisation is considered a fact, but the

authors stress that even after the remarkable extent of globalisation,
consumers from different cultures have different perceptions, needs,
tastes, beliefs, attitudes, preferences and values. As a result, a consumer
from a particular country responds in a style consistent with that particular
culture’s norms and values (Zhang and Neelankavil 1997; Banerjee 2008).
Therefore, brands introduced to consumers of new markets, with their
concomitant new cultural contexts, acquire reinterpreted and changed
brand meanings (Kay 2006).

4
One of the first and most easily noticeable causes of different brand
interpretation is language. It has even caused some of the most banal
international brand failures. Perhaps the most famous one was
Electrolux’s ambiguous slogan in the US: “Nothing sucks like Electrolux”
(Haig 2003). Images are assumed to be somewhat more universal and
easier to standardise (Moriarty and Duncan 1990). However, they cause
more subtle and therefore potentially more serious problems. This is
because it is even more difficult to understand how pictures are
understood by consumers in different cultures. For example, Hampden-
Turner and Trompenaars (1996) explain that consumers belonging to
integrated cultures have richer association networks that are wider in their
meaning than consumers in specific cultures.

The sub-surface meanings of symbols used in brand communication

fluctuate considerably according to variations in life experiences, product
knowledge, brand/advertisement history, and other factors (Bulmer and
Buchanan-Oliver 2006). In a cross-country investigation of visual
representations of beauty and health, Bjerke and Polegato (2006)
confirmed that different cultures do not interpret beauty and health
symbols equally. In another research work, the same authors (Polegato
and Bjerke 2006) studied the understanding and appeal of the
controversial Benetton advertisements across cultures and again found
inconsistencies.

Hence, Bulmer and Buchanan-Oliver (2006) advise advertisers to craft

scenes and sequences to provide the viewer with a brand meaning,
keeping in mind that their interpretation is dependent on knowledge of
culturally constituted signs or symbols. Kates and Goh (2003) recognise
this phenomenon as “brand morphing” and define it as a way that brand
meaning, facilitated by practitioners’ efforts to accommodate, reinforce
and create diverse cultural meaning, changes among different groupings of

5
consumers. Brand morphing represents the third and the highest level of
brand adjustment to cultural differences (as opposed to a pure adaptation
of the executional elements of global ads and a change in brand
positioning, as the first two levels of brand adaptation).

To sum up, several researchers who have studied branding in cross-

cultural contexts discovered significant differences in the brand meanings
that consumers from different cultures attribute to the same brand. Most of
these studies share two characteristics. Firstly, they were designed to
analyse consumers’ interpretations of advertisements. Secondly, most of
these studies focused on the differences in meaning in terms of how the
particular message received gets decoded by consumers, and less in terms
of how the brand message was integrated into consumers’ memory
networks. Since advertising is not the single brand communicator, let
alone the single brand meaning creator, an investigation of the overall
meaning of the brand, as opposed to a mere advertising interpretation is
due. Furthermore, it is also necessary to understand how the brand
message gets integrated into consumers’ memory networks as this
indicates what the brand as such represents to consumers and what
benefits it can bring them. Understanding the integration of the message
into consumers’ memory networks is a much more suggestive indication
of what his or her purchasing decision would be.

The above-mentioned gaps serve as a platform for this research, which

aims to:

x Understand what brands mean to consumers and consequently

conceptualise brand meaning,

x Reveal how different cultural backgrounds affect global brands’

meanings among otherwise homogeneous consumers in terms of
demographic, professional and occupational characteristics.

6
There is also a set of sub-questions that will direct the research and that
the results of the research should supply answers to. These are:

x What do brands mean to consumers?

x How do brands acquire meaning?
x How are other brand concepts such as brand benefits, associations,
beliefs, and attributes related to brand meanings?
x What is culture and how do cultures differ?
x How do differences in cultural dimensions affect brand meanings?
x How are brand consumption frequency and brand affection related
to brand meanings across and within countries?
x What are the scientific and managerial implications of the findings?

The answers to these questions provide useful guidelines for multinational

corporations on how to avoid unintended positioning caused by
differences in consumers’ interpretation of brands. They also serve to
guide brand managers in managing their brands internationally in order to
provide rich brand meanings to consumers of different cultural
backgrounds, and gain their loyalty as a result.

A detailed explanation of the methodology applied in this thesis is

provided in the fourth chapter. However, at this stage, it is helpful to
address the main aspects of the methodology so as to better understand the
aim of the paper. Since the research aims to explore certain new concepts
and also test theoretically-based hypotheses, both an inductive and
deductive approach were applied - the former through a series of
structured in-depth interviews with business graduates, and the latter
through an online survey with business students. In both stages of the
research, several global brands were evaluated in several culturally
different countries.

7
1.3. Structure of the Thesis

This thesis is structured into seven chapters as presented in Figure 1-1.

The first introductory chapter sets the stage for the research. It briefly
provides an overview of the current state of scientific thought on the
concept of brands and, in particular, brand meaning. It also explicitly
states the purpose of the thesis articulated through the research questions,
and finally presents the structure of the paper.

The concepts of brand, brand meaning, culture and cultural dimensions are
introduced in the second chapter. This chapter begins with a short
synopsis of what a brand is, how it is understood in this thesis and which
other concepts are connected to its contemporary understanding. Then the
chapter introduces brand meaning as a central concept to the thesis and
discusses the source of brand meaning creation. In the second part of the
chapter, culture, its definition and role in social studies are introduced, as
well as cultural dimensions as described by previous cross-cultural
research.

The third chapter elaborates on previous research and builds the basis to
theoretically connect cultural differences to brand meaning. This is done
by introducing two latent constructs, i.e. consumer needs and brand
benefits. In addition, brand usage and brand affection are also introduced
as determinants of brand meaning. In this chapter, a hypothetical model is
gradually constructed and presented.

The fourth chapter details the considerations taken into account in

selecting an appropriate research methodology design. Since the study has
two separate, but tightly connected purposes, the study was conducted in
two stages. In order to define what a brand means to consumers, a
qualitative inductive study was conducted; whereas, to explore how brand

8
meanings vary across cultures, a quantitative deductive study was carried
out.

Chapter five presents the qualitative empirical research. It starts with the
research procedure and then continues by reporting the results of the
research. Finally, it concludes with a discussion of the findings. The
central outcome of the qualitative research was the conceptualisation of
brand meaning.

The quantitative empirical research is presented in the sixth chapter. It is

structured in a similar way to the fifth chapter. Firstly, the research
procedure is presented. Then, the central part focuses on presenting
differences in brand meanings among consumers in the four chosen
countries, and among different types of users. Finally, based on the
results, the hypotheses are discussed.

This thesis ends with a conclusion in the seventh chapter. This chapter
provides a summary of the research and findings, the implications for
managers, and the scientific implications: contribution to the theory,
limitations, and directions for future research.

9
 1. INTRODUCTION

2. DEFINITIONS AND THEORETICAL BACKGROUND

Dimensions

Meaning
Cultural

Brand
Consumer Brand
Needs Benefits

Type of
User

3. PREVIOUS RESEARCH AND HYPOTHETICAL

MODEL

4. RESEARCH METHODOLOGY

5. QUALITATIVE EMPIRICAL RESEARCH

DEFINING BRAND MEANING
Research Procedure
Results
Discussion

6. QUANTITATIVE EMPIRICAL RESEARCH

EXPLORING DIFFERENCES IN BRAND MEANINGS ACROSS CULTURES
Research Procedure
Results
Discussion

7. CONCLUSION

Summary of Managerial Scientific

the Findings Implications Implications

Figure 1-1. Structure of the Thesis

10
 2. Definitions and Theoretical Background

This chapter provides an introduction to two key concepts of this study,

namely brand and culture. Furthermore, it gives a short overview of how
both concepts evolved and indicates the current stage both of them have
reached in the scientific research.

2.1. The Concept of Brand

A brand does not exist by merely giving a product a name, a trademarked

logo, unique packaging, a recognisable colour and other possible design
features. As Holt (2004, p. 3) explains, the name, logo and design are the
material markers of a brand. If a product does not yet have a history, these
markers are empty. They have no meaning, so there is no brand. Brand
markers are essentially a blank screen on which to project a fantasy (Hatch
and Rubin 2006). Only over time, as ideas about the brand and
experiences with it accumulate and fill the brand markers with meaning,
does the brand come to life.

Achenreiner and John (2003) studied the evolution of a child’s

understanding of brands and realised that younger children are more
attuned to brands on a perceptual level (serving as a cue for products with
familiar names and perceptual features), and that only later at some point
of their personal development do conceptual meanings become salient
(serving to evaluate a brand as such and not as the name of a product, or
inferring the personality characteristics of the consumers owning it). The
learning process also takes place in adults, however here it is not an

11
adult’s conceptualisation that develops, but rather his experience with, and
knowledge of, a brand.

Only a decade ago, even brand managers observed brands as symbols and
not as carriers of identity, personality and benefits. Brands were primarily
seen as one time transaction facilitators, far away from the long-term
relationship approach. Brands were also considered as the producers’
property. It was implied that the producer is mostly responsible for the
communication and the activities developed in the long run of the brand’s
reputation. This is far from the contemporary view in which brands belong
to all the stakeholders, and all the stakeholders contribute to the creation
of its identity. Today, brands are thought of as complex entities and their
expression includes the perception of their product characteristics,
personality and values (Veloutsou 2008) 1.

The shallow view of brands is what Ambler and Styles (1997) called a
product-plus definition. They consider such a view to be outdated because
a brand is seen as an addition to the product, and branding as one of the
last decisions to be made in the product development process. In contrast,
a profound understanding of brands is what they call a holistic definition.
According to them, a holistic definition of a brand puts the focus on the
brand itself, which encompasses much more than just the product. Such a
definition views the brand as “a promise of the bundles of attributes that
someone buys and that provides satisfaction” (Ambler and Styles 1997, p.
1).

1
In that light, it is quite disturbing to realise that contemporary branding thoughts have not
been reflected in the definition of brand by, according to many, the most important academic
marketing organisation: the American Marketing Association. The dictionary of this
organisation still defines a brand as: a name, term, design, symbol, or any other feature that
identifies one seller's good or service as distinct from those of other sellers.
(http://www.marketingpower.com/_layouts/Dictionary.aspx?dLetter=B, 11th November
2009)

12
Given the scientific and managerial need for proper phrasing and
consistent terminology (Brown et al. 2006), and despite ongoing discourse
on what a brand is and many existing incongruities on how to define it
(Stern 2006), it is a fact that brand experts and the broader public have by
now accepted the holistic view of branding. Such a view propagates and
studies what has over the years been named brand identity.

PICTURE OF SENDER

Physique Personality
EXTERNALISATION

INERNALISATION
Relationship Culture

Reflection Self-image

PICTURE OF RECIPIENT

Figure 2-1. The Brand Identity Prism

Source: Kapferer (1998, p. 107)

For example, Kapferer (2004) developed a hexagonal identity prism

(Figure 2-1) as a model that defines what a brand really is and what it
stands for. Identity is what a brand wants to be perceived as, i.e. what it
“transmits”. On the other end of the communication channel, i.e. what
customers receive, is the brand’s image. Due to the market noise, the
brand image is not identical to the brand identity. Obviously, the
perceived image is what matters in the end; nevertheless, to manage a

13
perceived image, the inner self of the brand, i.e. its identity, is what should
really be managed.

Similarly, Aaker (1996), proposed a brand identity system (Figure 2-2)

that encompasses the whole complexity surrounding how a brand evolves.
His model emphasises the need to consider product-related associations,
organization-related associations, personality, and symbolic associations
in order to ensure that brand identity has texture and depth. He also
differentiates between core and extended brand identity, and later with
Joachimsthaler (Aaker and Joachimsthaler 2002) adds an even more
focussed dimension named “brand essence”, which might be useful for
brands that possess several compactly summarised core identities.

BRAND IDENTITY
Extended

Core

Brand as product:
- product scope
- product attributes Brand as Brand as Brand as
- quality/value organisation: person: symbol:
-uses - organisation - personality - visual image
-users atributes - customer / brand and metaphors
- country of origin - local vs. global relationship - brand heritage

VALUE PROPOSITION CREDIBILITY

-functional -emotional -self-esteem Support other brands
benefits benefits benefits

RELATIONSHIP

Figure 2-2. The Brand Identity System

Source: Aaker (1996, p. 79)

14
 Company actions

Strategy Programmes

Direction Quality Specifics: Quality

type, budget

Competitor's What customers think Partners' actions:

actions chanels,
and feel about a brand
employees
Industry /
environmental
conditions

Awareness Associations Attitude Attachment Activity

(Satisfaction)

What customers do
about a brand

Financial market
impact

Figure 2-3. The Systems Model of Brand Antecedents and Consequences

Source: Keller and Lehmann (2006)

A decade after the seminal work by Aaker (1996), Keller and Lehmann
(2006) proposed a new model that connects key concepts in the world of
brands. They suggest that although many brand dashboards capturing
multiple aspects of brand equity and performance have been developed by
firms, they are, rarely linked together. Therefore, they consider it
necessary to develop a comprehensive model (as depicted in figure 2-3) of
how brand equity operates, and to develop estimates of the various cause-

15
and-effect links within it. Thus, they recognise the model as a brand value
chain which develops from company actions, to what customers think and
feel about a brand, then further to what customers do about a brand, and
finally to how this transfers into financial value.

2.1.1. Branding Concepts

Throughout the years, various research activities in branding have

produced different kinds of concepts that may become linked to a brand.
According to Keller (2003b), these are: awareness, attributes, benefits,
images, thoughts, feelings, attitudes, and experiences. Though Keller has
offered a simple definition of these concepts (explained further on in this
chapter), he admits, even worriedly, that many more definitions exist for
these concepts and, what is worse, many of these concepts are being used
for different phenomena. Other brand concepts also frequently appear in
the literature: the already mentioned brand identity (Aaker 1996; Keller
2003b; Kapferer 2004), brand personality (Aaker 1996; Aaker 1997;
Kapferer 2004; LaPla and Parker 2002), brand beliefs (Orth and De
Marchi 2007), brand relationships (Fournier 1998; Sweeney and Chew
2002; Veloutsou 2007), and many more. The following section of this
thesis elaborates key brand concepts as well as their interrelationships. It
also points to some theoretical inconsistencies in both the concepts’
definitions and the cause-effect relationships among them.

2.1.1.1. Brand Awareness

The strength of a brand’s presence in a consumer’s mind is defined as the

level of awareness (Kapferer 2004, p. 17), while brand awareness is
defined as a precondition for the existence of a brand (Franzen and
Bouwman 2001, p. 171). According to Aaker (1996, p. 10), there are three
levels of awareness, where recognition is the weakest level, recall

16
somewhat stronger, and dominance the strongest. Keller (2003a, p. 67)
defines awareness more stringently and considers that association of the
product category and targeted needs to the brand are necessary conditions
for the existence of brand awareness. In this sense, mere recognition
would not be regarded as awareness.

Keller (2003a) further differentiates between the depth and breadth of

brand awareness. Depth is understood as the ease of brand recognition and
recall, and breadth refers to the number of purchase and consumption
situations for which the brand comes to mind. Kapferer (2004, p. 17)
differentiates between aided and unaided awareness, and claims that both
contribute to brand equity. According to Aaker and Joachimsthaler (2002,
p. 17), although awareness is often taken for granted in branding and is
considered an insufficient condition for measuring a brand’s strength, in
their opinion it strongly contributes to brand equity.

In summary, awareness can be defined as a multilevel understanding of

the brand, which is a prerequisite for brand existence, and at the same time
contributes to brand equity.

2.1.1.2. Brand Associations

Krishnan (1996) defines associations as a link between nodes in the

consumer's mind. In other words, memory is composed of knowledge that
is organised as a network of connections. The building blocks of this
network are mental nodes that represent any piece of information like a
brand, an attribute, a situation or similar, and connections between the
nodes that represent associations.

Dillon et al. (2001) refer to Aaker (1991), who views associations as one
of five components of brand equity, along with brand loyalty, awareness,
perceived quality, and proprietary brand assets; and Keller (2003), who

17
claims that consumers may have a brand node with a variety of
associations linked to that node, including attributes, usage occasions,
benefits, and attitudes. Kaynak, Salman, and Tatoglu (2008) further
elaborate the relationship between these concepts and propose that brand
benefits, brand attributes and brand attitudes provide ground for brand
associations in consumers’ minds.

Leo, Bennett, and Härtel (2008) propose that a brand be analysed on three
distinct levels. They adopt brand-specific associations and general brand
impressions from Dillon et al. (2001), but they add brand commitment.
They further divide each level into two constructs: brand specific
associations into emotional value and perceived quality; general brand
impressions into brand awareness and brand image, and brand
commitment into brand loyalty and purchase intention.

Overall, brand associations are understood as links in the consumer’s

memory between brand on one side and attributes, benefits, and attitudes
on the other.

2.1.1.3. Brand Attributes

Before the brands2, product attributes were the features that differentiated
the products of one producer from the products of another. Thus, product
attributes may be considered the basic descriptive features that
intrinsically or extrinsically characterise the product (Keller 2003b), are
tangible or intangible (Riesenbeck and Perrey 2007), and can be evaluated
by trial (Orth and De Marchi 2007).

Brand managers who focus on attributes to develop a brand’s identity are

paving a path to weak customer loyalty (Phau and Lau 2000). Such a

2
More precisely, in those times when brands were not so extensively present in consumers’
lives.

18
focus results in brand traits that are relatively easy for competitors to
copy. Strong, successful brands “move beyond attributes to a brand
identity based upon a brand personality and a relationship with customers”
(Aaker 1994, p.122). De Chernatony (2001, p. 209) views attributes as the
basis of a brand pyramid. According to that pyramid, brand attributes
provide brand benefits that further provide values, which finally create a
certain brand personality.

In summary, agreement exists that brand attributes are brand features

which are a necessary but not sufficient condition for creating durable
consumer relationships, and that brand attributes are just a step towards
brand personality.

2.1.1.4. Brand Personality

Aaker (1997) defines brand personality as the set of human characteristics

associated with a brand. This set of human characteristics is often the most
differentiating dimension between brands (Aperia and Back 2004).
According to Riesenbeck and Perrey (2007, p. 151), brand personality
accounts for 70% of a brand’s strength, and according to LePla and Parker
(2002, p. 86) it is the foundation for each customer’s emotional
relationship with a brand and its company as a whole.

In Kapferer’s (2004, p. 107) identity prism, among the six edges,

personality is the innermost edge of a brand. It is internal and it is seen
from the brand’s perspective. Personality explains the character of a brand
by answering the following question: ‘Were it a person, what kind of
person would it be’? Personality, as an expression of a brand’s identity,
has been experimented with in brand advertising as far back as the 1970s
(Czerniewski and Maloney 1999, p. 99). It was then considered that the
best way to communicate a brand’s traits was to have a celebrity endorse
it. By this, the celebrity transfers his or her own personality to the brand,

19
as elaborated by McCracken (1986). When people give a brand human
attributes such as humour or intelligence, they can relate to it far more
easily.

Czerniawski and Maloney (1999, p. 19) claimed that a strong brand is the
one that lives up to its clearly defined “Brand Positioning Statement”. This
consists of three simple sentences, but includes all of the important facets
of the brand’s identity, including brand character (i.e. brand personality).

Aaker (1997, p. 352) has probably given the most profound explanation
on how to measure a brand’s personality. She initially recognised 309
human personality traits, which she filtered down to 114 and finally to 5
stable and robust traits named the “Big Five” factors of brand personality
(sincerity, excitement, competence, sophistication and ruggedness). She
further described each of the “Big Five” with 2 - 4 facets to provide
“texture and descriptive insight”. Although Aaker’s 15 item scale has
often been criticised as not really measuring brand personality (Azoulay
and Kapferer 2003) and as being country-specific (Sung and Tinkham
2005), it is still a widely cited and appreciated marketing tool that has
resurfaced again in recent studies (e.g. Sweeney and Brandon 2006).

In summary, brand personality is an expression of brand identity, an

important contributor to brand strength, and a basis for the consumer
brand relationship.

2.1.1.5. Brand Beliefs

Brand beliefs underlie benefits (Keller 2003a), and they are a key reason
for the strength of certain brands, i.e. “high brand equity”, and favourable,
strong, and unique associations in consumers’ minds (Keller 1993). Orth
and De Marchi (2007) use the term brand beliefs interchangeably with
brand associations. They define brand beliefs as features, attributes or

20
benefits that consumers link to a certain brand and that help to
differentiate that brand from the competing ones. According to Kempf and
Smith (1998), brand beliefs are brand cognitions that are antecedent to
brand attitude and consumer purchase intentions (Orth and De Marchi
2007; Brown and Stayman 1992).

Overall, the most important characteristic of brand beliefs is its personal

and cognitive aspects. Researchers also claim brand beliefs are the reason
for brand strength.

2.1.1.6. Brand Knowledge

Brand knowledge is a network in a consumer’s mind consisting of nodes

and links between the nodes (associations). It can be defined in terms of
the personal meaning about a brand stored in a consumer’s memory, i.e.,
all descriptive and evaluative brand-related information (Keller 2003a). It
is characterised in terms of brand image and brand awareness. The former
is the other nodes that a brand becomes linked (associated) to, while the
latter is the strength of the node in terms of the number of links to it, with
the higher number characterising easier and more frequent access to a
brand node.

Peter and Olson (2001) relate consumer brand knowledge to the cognitive
representation of a brand. Cognitions, as opposed to affection, signify
personal responses to any brand related information (Keller 2003b). More
precisely, it is a process of integrating previous knowledge with
informational input to evaluate its relevance and importance for judgment
making (Jun, Cho, and Kwon 2008).

In summary, brand knowledge is a concept that is often related or even

equated to brand beliefs.

21
4.1.1.7. Brand Emotions

Moods and emotions are two common forms of affection, and affection,
emotions and feelings are used interchangeably in advertising literature.
They signify an appraisal of an object, person, or event as good or bad,
favourable or unfavourable, desirable or undesirable, or momentarily
pleasant or unpleasant (Jun, Cho, and Kwon 2008). Plassmann et al.
(2007) further claim that research in cognitive psychology provides
evidence that emotions play an important role in memory processes as
they help people to learn and remember cognitive processes.

Franzen and Bouwman (2001, p. 228) admit that brands invoke affective
reactions, but they question whether such emotions go much further than
positive sensations such as “appealing” or “likeable”, and whether such
sensations may be called feelings. They do not question the widespread
notion that successful brands communicate with consumers on an
emotional level (cf. Heath, Brandt, and Nairn 2006, Pawle and Cooper
2006; Thompson, Rindfleisch, and Arsel 2006), but rather they stress that
brands evoke mere affection and not deeply experienced emotions.

Based on a study by Homer and Yoon (1992), Orth, Koenig, and

Firbasova (2007) propose that emotions influence cognition, which further
influences attitude and finally the purchase intention. A similar model was
also proposed by Kim, Morris, and Swait (2008), in which they propose
and empirically confirm that brand credibility influences brand affection
and brand conviction, which then influences brand attitude strength.

More specifically, emotions act together with cognition to strengthen the

representations of brands in consumers’ minds.

22
2.1.1.8. Brand Attitudes

When a new brand is introduced, individuals form attitudes toward it

through changes in beliefs (Fishbein and Middlestadt 1995). Attitudes are
a result of a consumer’s experience with a brand (Franzen and Bouwman
2001, p. 277), and can be defined as summary judgments and overall
evaluations to any brand-related information (Keller 2003b), or a
generalised predisposition to behave in a certain way towards an object
(Park and MacInnis 2006). Jun, Cho, and Kwon (2008) referred to brand
attitude as brand preference, but Franzen and Bouwman (2001, p. 274)
disagree. According to them, preference is expressed in the purchasing
decision, which is, for the majority of products, based on awareness and
accessibility rather than likeability. Attitudes, on the other hand, do not
result in the product purchase decision (especially not in the case of fast
moving consumer goods). Along the same lines, Kaynak, Salman, and
Tatoglu (2008), differentiate between behavioural loyalty and attitudinal
loyalty.

Kumar, Lee and Kim (2008) studied the role of cognitive and affective
responses in the purchasing intention, and oppositely to the afore-
presented studies by Homer and Yoon (1992), Orth, Koenig, and
Firbasova (2007) and Kim, Morris, and Swait (2008), they propose a
model in which attitude towards a brand precedes and influences the
cognitive and affective response. Finally, Jun, Cho, and Kwon (2008) do
not relate affection and cognition on one side, and attitude on the other
side as a cause-effect relationship, but rather they speak of the cognitive
and affective aspects of attitude.

In summary, brand attitudes are summary judgments, and although they

are clearly related to both emotions and cognition, the research has not
reached common ground in understanding their relationship.

23
2.1.1.9. Brand Benefits

Brand benefits are personal values and meanings that consumers attach to
a brand's product attributes. Frequently, a major distinction is made
between three basic categories of benefits according to the underlying
motivations to which they relate – functional, experiential, and symbolic
benefits (Keller 2003a; Park, Jaworski, and Maclnnis 1986). Functional
benefits are the more intrinsic advantages of product consumption, and
they usually correspond to product attributes. These benefits are linked to
basic motivations, such as well-being and health. Experiential benefits
relate to what it feels like to use the product and they also usually
correspond to product attributes. These benefits satisfy experiential needs
such as sensory pleasure and cognitive stimulation. Symbolic benefits are
the more extrinsic advantages of product consumption. They usually
correspond to non-product-related attributes and relate to underlying needs
for social approval, personal expression, and outer-directed self-esteem
(Orth and De Marchi 2007).

In summary, the literature defines brand benefits as the personal value and
meaning attached to brand attributes which evoke affective and cognitive
responses.

Evidently, there are numerous interconnected branding concepts. As

shown at the beginning of chapter 2.1, many researchers have previously
engaged in creating models to clarify the cause and effect relationships
between the branding concepts. Consequently, a number of conceptualised
and even practically implemented models that explain these relationships
exist (The Brand Identity System (Aaker 1996, p. 79), The Brand Identity
Prism (Kapferer 2004, p. 107), and The Integrated Brand Model (LePla
and Parker 2002, p. 15)). However, the models and definitions of the

24
concepts are not uniform, and they are sometimes even contradictory. In
an attempt to clarify this confusion, this research (admittedly, like many
before) aims to define the key branding phenomenon and to provide an
explanation for why consumers are willing to pay the afore-mentioned
price premium of 37% for branded products. This research project argues
that the question of what brands really mean to consumers is crucial at this
stage of branding research. Hence, the thesis proceeds by focusing on the
desk research evidence of how the concept of brand meaning is
understood in the literature so far.

2.1.2. Brand Meaning

Brand meaning is an emerging concept in brand management, so it is still

not clearly defined in the literature. For example, Oakenfull et al. (2000)
present the results of research devoted to developing a measurement tool
for brand meaning. Although their paper focuses on the concept of brand
meaning and mentions it more than a dozen times, the paper is devoid of a
clear definition of what brand meaning is. Davis (2007, p.255) later
interprets Oakenfull et al.’s (2000) understanding of brand meaning as
“the most definitive or core attributes of the brand that the consumers
perceive”. Similarly, Park, Milberg, and Lawson (1991) also study the
relationship between attributes and meanings, but in their understanding
brand concepts are brand-unique abstract meanings that typically originate
from a product attributes and a firm's efforts to create meanings from these
arrangements.

In a study by Henderson et al. (2003), brand meaning is considered equal

to a brand association; that is, the participants in their study are asked to
list the first meaning or association that comes to mind when looking at
the brand’s logo. Moore and Homer (2008) also use the terms ‘brand
associations’ and ‘brand meanings’ as synonyms, and Feldwick (2002)

25
even defines brand meaning as the collective associations and beliefs that
a consumer has about a brand. Quester, Beverland, and Farrelly (2006)
studied how brand meaning differs for different members of a subculture.
Although brand meaning is a term used in the title of their paper, it is not
clear what they refer to when they mention it as a marginal term in the
paper. Raggio and Leone (2007) refer to brand meaning as well, but from
a company’s perspective, and they claim that value signifies what the
brand means to its company.

Another widely cited paper by McCracken (1986) focuses on how

meaning is manufactured and transferred from celebrities to brands and
further to consumers. In that process of cascaded meaning transitions,
consumers opt to acquire an image of their beloved celebrity by using the
same brands that the celebrities use. McCracken, however, also fails to
define what he understands by brand meaning. Since he speaks of
meaning being transferred from a celebrity to a brand, it can be assumed
that brand personality is what is referring to.

Escalas and Betmann (2005) also claim that meaning transfers from
brands to consumers. More precisely, they assert that consumers
appropriate brand meanings, emerging from associations of brands with
reference groups, to construct their self-concepts. Similarly, Hollenback,
Peters, and Zinkhan (2008) believe that the congruency between a brand’s
identity and a consumer’s identity is an important source of brand
meaning. They further believe that the meaning of a brand can be
expanded by strengthening the brand’s identity. Along those lines, Moore
and Homer (2008) believe that perceived brand meaning includes image
and brand personality.

Keller (2003b) suggest that providing a brand with meaning actually

implies explaining to consumers what a particular product can do for them

26
and why it is special and different from the others in the category. Thus,
he strongly relates meaning to brand benefits. Similarly, Veloutsou and
Moutinho (2008) suggest that consumers use products not only for their
utility, but because of their symbolic meanings.

Along the same lines, in their study on the VW Beetle, Brown, Kozinets,
and Sherry (2003) mention that the Beetle has rich symbolic meanings.
They found that social and cultural contexts animate brand meaning, so
that brands mean more than relatively fixed arrangements of associative
nodes and attributes. As they stress, brands are not only fixed cognitive
associations of meanings as implied by the strategic brand management
models of Keller (2003) and Aaker (1996), but also dynamic, expanding
social universes composed of stories. Despite their criticism of the
cognitive view of brands by others, their paper also remains vague on
what ‘brand meaning’ is. However, they introduce the ‘4A’ concept of
brand meaning, which includes aura (brand essence), allegory (brand
stories), arcadia (idealised community) and antinomy (contradiction).
Aura, allegory and arcadia are the character, plot, and setting of brand
meaning, while antinomy (contradiction) is an element that represents
brand paradox and therefore brings the cultural complexity necessary to
animate each of the other three elements.

Chang and Chieng (2006) put brand meaning into the context of other
brand terms and consider brand knowledge (or brand meaning) to always
be linked to brand awareness, brand association, perceived quality, brand
attitude, brand personality, and brand image. However, Martin, Stewart,
and Matta (2005) claim that knowledge and attitudes (“brand meaning”)
are brand associations belonging to a network of associations which
includes the brand name, concrete and abstract product attributes, and
usage occasions, among other things. Plassmann et al. (2007) do not

27
define what a brand is, but they assume that brand meaning consists of
cognitive and affective clues.

Finally, Franzen and Bouwman (2001) speak of mental links between

brand names, images and cognitions in a consumer's memory that cause
the brand to acquire meaning. They also differentiate between four layers
of brand meaning, depending on the level of socialisation at which the
meaning gets created - starting from the private meaning to the social one.
These layers are: the direct sensory or iconic impression (the colour,
shape, sound of a brand, and similar, which do not vary between
consumers); the idiosyncratic meaning (associations due to personal
experiences with an object which are unique to each consumer); the sub-
cultural associations (connected to an object or an idea by members of a
given subculture); and the cultural associations (connected to an object or
an idea by the members of a given culture). Finally, Franzen and
Bouwman (2001) also identify ten types of brand meanings: brand signs,
sub-brands, provenance, product-related brand meanings, situational
meanings, symbolic meanings, perceived quality, perceived price,
presentation and advertising, and other communications means.

As shown in subchapter 2.1.1, different concepts within the world of

brands are used interchangeably and for multiple manifests. Subchapter
2.1.2 has provided evidence that the concept brand meaning is not devoid
of complexity and ambiguity either. The above-mentioned definitions of
brand meaning are analysed in terms of their constituting elements and an
overview is presented in table 2-1. From the summarised definitions, it
can be seen that the terms most often equated or related to brand meaning
are brand associations and brand attributes.

28
Table 2-1. An Overview of the Existing Brand Meaning Definitions

Author(s) Brand meaning is Brand meaning originates from

Brown, Kozinets, and Sherry Aura, Allegory, Arcadia, Antinomy

2003
Chang and Chieng 2006 Knowledge Related to: Awareness, Association,
Perceived quality, Attitude, Personality,
and Image
Davis 2007 Attributes
Escalas and Betmann 2005 Personality Associations
Feldwick 2002 Associations, Beliefs
Franzen and Bouwman 2001 Impression, Mental links between brand names,
Associations images and cognitions
Hendersson et al. 2003 Associations
Hollenback, Peters, and Identity
Zinkhan 2008
Keller 2003b Benefits
Martin, Stewart, and Matta Knowledge and Belongs to a network of associations
2005 attitudes that also include brand name, concrete
and abstract attributes, and occasions
McCracken 1986 Personality , Image
Moore and Homer 2008 Associations, Image,
Personality
Oakenfull et al. 2000 Attributes
Park, Milberg, and Lawson Attributes
1991
Veloutsou and Moutinho 2008 Image

Before moving on to exploring how brands acquire meanings, a small

detour into the field of psychology and neuro-linguistic programming
(NLP) offers the following definition of meaning in general. Meaning is
related to the intention or significance of a message or experience. It is the
natural consequence of interpreting experience using inner representations
or experiences that are associated with external cues and events.

29
Therefore, the meanings that people construct and how these meanings are
constructed are connected with the richness and flexibility of each
person’s internal representations of the world. Fundamentally, meaning is
a product of one’s values and beliefs (Dilts and DeLozier 2000, p.703).

This definition serves as a good introduction to take a step forward in the

quest to understand what brands mean to consumers, and to trace the path
along which a brand acquires meaning. At this stage, it is particularly
important to analyse how brands come to acquire meaning, and who
influences brand meaning the most.

2.1.3. How Brands Acquire Meanings

It has long been believed that brand managers and advertising agencies are
the creators of brand meaning. Now it has become clear to researchers in
the brand field that brand managers and agencies only propose the brand
stories. In actual fact, brand meaning is not constructed in advertisements,
but rather in consumers' minds. This difference between the traditional
and contemporary view of brand image creation is depicted in figure 2-4.

The following section explains the role of each contributor in the brand
meaning creation process.

2.1.3.1. Brand Strategies’ Role in Brand Meaning Creation

Not many consumers are aware that their favourite branded shoes are not
manufactured by the company that owns the brand. Brand-owning
companies manufacture a brand, while the shoes are manufactured
elsewhere, preferably in the “Third World, where labour is dirt cheap,
laws are lax and tax breaks come by the bushel” (Klein 2000, p. 28). This
is because anyone can manufacture a product. It is a manual task. Brand
manufacturers, meanwhile, are free to focus on the real business of:

30
“creating a corporate mythology powerful enough to infuse meaning into
these raw objects just by signing in its name” (Klein 2000, p.28).

The Traditional View of Brand Meaning Creation

Brand identitiy
is communicated
Marketers to consumers via Brand image is
New product decide on vision, culture, absorbed and
is created desired brand positioning, accepted by
identity personality, consumers
relationshp and
presentation

“Pool of Reality” model of Brand Image Creation

New product is Brand manager

created instigates dialogue

Consumers enter
dialogue about
brand image

Central
pool of Creatives and
meanings agency enter
dialogue about
Brand consultant s communications
enter dialogue about strategy and
brand identity tactics

Figure 2-4. Traditional and Contemporary Brand Image Creation

Source: Blythe 2007.

Successful brand managers are aware of several strategies of how to best

create a brand meaning. They realise that lasting brand relationships are
created when brands communicate with both the consumer’s head and
heart; that is, when both rational and emotional benefits are offered by a
brand (Heath, Brandt, and Nairn 2006). Schmitt (1999) talked of
experiential marketing and categorised it into: sense marketing, feel

31
marketing, think marketing, act marketing, and relate marketing. Chang
and Chieng (2006) developed this model and categorised sense, feel, and
think experiences into individual experiences, and act and relate
experiences into shared experiences. This is an insightful categorisation
which needs to be taken into consideration when approaching culturally
different (individual or collectivist) international markets.

Emotional branding is defined as a consumer-centric, relational, and story-

driven approach to forging deep and enduring emotionally charged bonds
between consumers and brands (Holt 2004; Thompson, Rindfleisch, and
Arsel 2006). Such branding has provoked the interest of researchers as
some of the most unbelievable brand successes happen when emotional
branding is used, two good examples of which are Absolut vodka and
Marlboro cigarettes.

In the last decade, emotional branding has gained significant advantage

over rational branding among practitioners, but especially so in the
branding literature (Pawle and Cooper 2006). It is superior to rational
branding as the content of communication (i.e. the messages transferred in
the rational branding) fades more easily and vanishes over time, while the
more subtle messages evoked by the emotional meta-communication
endure. During meta-communication, consumers are engaged in automatic
learning and processing of messages without consciously participating in
the process (Heath, Brandt, and Nairn 2006). These messages are best
communicated when all of the consumer’s senses are addressed: touch,
taste, smell, sight and sound (Lindstrom 2005). According to Thompson,
Rindfleisch, and Arsel (2006), the conceptual division between emotional
branding and more traditional benefit-driven branding strategies is rapidly
narrowing as emotional-branding principles are being integrated into the
benefit-driven brand management paradigm.

32
A relatively recent approach to branding is called cultural branding. This
approach departs from both rational and emotional branding and is defined
as a set of axioms and strategic principles that guide the building of brands
into cultural branding. For such a technique, it is necessary to recognise a
contradiction in society and to position a brand in a gap between the
socially portrayed ideal lifestyle and the individually experienced average
person’s reality (Holt 2004, p. 10). Companies like Nike, Polo and
Tommy Hilfinger do not create a brand by adding value to a product. They
“thirstily soak up cultural ideas and iconography that their brands could
reflect by projecting these ideas and images back on the culture as
extensions of their brands. Culture, in other words, adds value to their
brands” (Klein 2000, p. 32).

The afore-mentioned argument supports the notion that companies do

their best to impress consumers and in doing so opt for the most varied
strategic options. However, only those companies that offer the benefits
that consumers recognise as valuable will succeed. Certainly, with the
overload of brands and brand stories, consumers understand that
“marketers promiscuously stitch stories and images to their brands that
may have nothing to do with the brands' real history and consumption”
(Holt 2002, p. 84). Hence, consumers are increasingly looking for
evidence suggesting their brand has earned its image and authenticity.
This introduces the consumer as an important co-creator of brand
meaning. The consumer’s role is elaborated in detail in the subsequent
subchapter.

2.1.3.2. Consumers’ Role in Brand Meaning Creation

When consumers are exposed to a company’s stories or brand

information, they do not passively absorb those messages. Instead, they
create their own meanings by mixing received information with their own

33
memories, other stimuli present at the moment, and the metaphors that
come to mind as they think about the firm’s message (Zaltman 2003).
Therefore, although brand meanings might be ascribed and communicated
to consumers by marketers, consumers in turn uncover and activate their
own brand meanings, which are then communicated back to marketers and
the associated brand community (Brown, Kozinets, and Sherry 2003).

What matters in the construction of the brand relationship is not simply

the idea that managers intend for them, but what consumers do with
brands to add meaning to their lives (Fournier 1998; Cova and Pace 2006).
As Vargo and Lusch (2004) conclude in their paper: The consumer is
always a co-producer and he determines the perceived value of a brand on
the basis of “value in use”. The firm can only make value propositions,
but the final meaning is the result of an interpretation process (Richins
1994). Thus, although marketers initially design brands, final brand
meanings reside in consumers’ minds (Keller 2003; De Chernatony 2001,
p. 19).

Brown, Kozinets, and Sherry (2003) stress that the consumer’s role as co-
creators in brand creation does not imply that brand management is
impossible in a world of consumer-mediated meanings, but rather that it is
more complex and that meaning is co-created rather than imposed by
managerial dictate. According to McCracken (1986), meaning ascription
is neither directed from brand managers to consumers, nor from
consumers to brand managers, but is rather a two-way flow from brand
managers to consumers and back. One of the most important aspects of a
brand, then, is that its value is highly individual (Berthon, Pitt, and
Campbell 2008). Even when a group of peers negotiate the meaning of
widely consumed brands, their understandings are not unique to all the
peers, but rather individualistic (Nairn et al. 2008).

34
It must also be noted that post-modern consumers value brands more if
they are offered not as cultural blueprints, but rather as cultural resources,
i.e. as useful ingredients to reproduce oneself as one chooses (Holt 2002).
For example, Paul Edwards, the Chief Strategy Officer of Publicis, the
French media and advertising conglomerate, speaks about customer
empowerment and how brand managers need to actively listen to their
customers by adopting a ‘receiver’ approach to communications as
opposed to the ‘transmitter’ approach they have used for many decades
(Christodoulides 2008). Some modern managers notice the consumers’
need to co-create brand meaning. Such managers give consumers space in
brand creation. Ferrero’s brand management, for example, decided to take
a step back in managing the Nutella website and offered the brand’s fans
more creative space to self-express and show off (Cova and Pace 2006).

The online world is the environment in which co-creation is the most

emphasised and the environment in which predominantly user-generated
brands such as Youtube and Wikipedia exist. It is thus not surprising that
of the top five global brands that most influence human lives, three are
online brands: Google, Youtube and Wikipedia (Zumpano 2007).
Obviously consumers no longer want to be passive. The act of taking
brand management away from the marketing professionals and giving it to
consumers allows consumers and other stakeholders to shape brand
meaning and endorse the brand. It is a way to establish true loyalty, as
opposed to mere retention. This phenomenon has also been called ‘brand
hijack’ (Wipperfürth 2006).

Despite the obvious advantages of consumer involvement, managers are

nonetheless warned that communication intensive environments such as
the Internet multiply the complexity of brand meanings and emphasise the
co-invention of brand interpretations (Ward and Ostrom 2003). Therefore,
managers of brands that are not “by definition” consumer-created should

35
be careful to maintain an equal level of understanding of their brands’
meanings as consumers (Berthon, Pitt, and Campbell 2008).

2.1.3.4. Societies’ and Other Stakeholders’ Role in Brand Meaning

Creation

Apart from brand managers and consumers, society also plays an

important role in brand meaning creation. This is primarily because, as
Thompson, Pollio, and Locander (1994) note, personal understandings are
situated within a network of culturally shared knowledge, beliefs, ideals,
and taken-for-granted assumptions about the nature of social life. Previous
researchers have proposed that brands are in fact the embodiment of the
product, the user, the producer, the marketer and the use situation (Coulter
and Zaltman 1994). Therefore, not only do manufacturers and consumers
co-create brand meaning, but there are also several other co-creators.
Social settings, rituals, mass media images, product symbolism, language,
cultural ideals, gender roles, religious and ethnic traditions are just a few
of the broad cultural factors that have systematic influences on the
experiences of individual consumers (Thompson, Pollio, and Locander
1994). Through ‘collective interpretation’, brands acquire meaning in the
social context and at particular moments in time as they are interpreted by
various stakeholders, all of whom contribute to brand meaning, some
more and some less than others (Hatch and Rubin 2006).

Gregory (2007) proposed a process of how stakeholders can contribute to

brand creation based on the concept of a ‘negotiated brand’. According to
him, the negotiated approach starts by a corporation identifying its core
values, which are a blend of management vision and internal stakeholder
views. Then the brand goes out to the world to be modelled by external
stakeholders and takes its final form. Gregory depicts this process as a
spiral development.

36
Generally, marketing literature is increasingly highlighting the role of
society in brand creation (Ligas and Cotte 1999; Vargo and Lush 2004;
LeBel and Cook 2008; Thompson 2004; Blythe 2007). Some recent cross-
cultural evidence even negates the individual self-concept, suggesting that
individuals’ mental representations of themselves may depend on the
social aspects of self, such as relationships with others and membership in
social groups (cf. Escalas and Bettman 2005); hence, the social influence
dominates the individual. In other words, neither managers nor consumers
completely control the branding processes. Instead, cultural codes
contribute to and constrain how brands work to produce meaning
(Schroeder 2005).

Along these lines, Brown, Kozinets, and Sherry (2003) consider brands as
social entities that are experienced, shaped, and changed by communities.
Kay (2006) explains that brands are social or cultural “property” to the
extent that consumers incorporate elements of “brand meaning” into their
lives, while O’Reilly (2005) considers brands as socially constructed texts
which mediate meanings between and amongst consumers and producers.

In conclusion, this chapter has outlined various branding concepts and

focused on establishing what brands mean to consumers. In the next
chapter, the focus switches to the second variable of interest, i.e. culture
and its role in the brand meaning creation process.

37
2.2. The Concept of Culture

Culture (and cultural dimensions) is the second field of research relevant

for the present study. This chapter provides an overview of what culture is
and how it differs among various groups of people.

2.2.1. Culture, Cultural Differences and Boundaries of Culture

Although the key psychological variable across different groups is their

culture (Pankhania, Lee and Hooley 2007), the word “culture” is used
very loosely to mean anything from patterns in eating norms to visual and
performing arts (Brannen et al. 2004). In 1952, two famous
anthropologists, Kroeber and Kluckhohn, collected 164 definitions of
what culture is (Bidney 1954). In their compilation, the oldest definition,
now more than a century old, was given by Taylor (1871). According to
him, culture is “a complex whole which includes knowledge, beliefs, art,
morals, laws, customs and any other capabilities and habits acquired by
individuals as members of society”.

With the growing opportunities for intercultural interaction since the

fifties, the world has seen a proliferation of modern studies on culture. The
consensus on what constitutes the core of culture is still not uniquely
understood by even the most prominent researchers in the field. In line
with the social science theories of the 1950s, Hofstede (2001), for
example, considers ideas and especially their attached values to be the
core of culture. He cites the definition by Kluckhohn (1951), who says
that “Culture consists of patterned ways of thinking, feeling and reacting,
acquired and transmitted mainly by symbols, constituting the distinctive
achievements of human groups, including their embodiments in artefacts;
the essential core of culture consists of traditional ideas and especially
their attached values” (Hofstede 2001, p. 9).

38
On the other hand, Trompenaars (1994), positions meaning and human
interpretation at the core of culture, in line with contemporary theory.
According to him, the different interpretations of (even possibly equal)
norms and values distinguish cultures. This school of thought refers to
culture as a set of social meaning-making processes (O’Reilly 2005) or as
a constituent of the world which supplies it with meaning (McCracken
1986).

Other researchers in the field of international management offer their own

definitions of culture. For example, culture is the way a group of people
lives (Romani 2004) and makes decisions (Leo, Bennett, and Härtel
2005). It is a combination of interdependent, gradually changing elements
(including assumptions, beliefs, values, practices, and institutions) and
links between these elements, which are distinctive to a particular society
(Brannen et al. 2004). It is a group-level phenomenon, but influences an
individual’s perceptions, values and behaviours, especially in a social
setting (Maznevski et al. 2002). It is learned and shared; it links
individuals to groups, but allows individual variability (Brannen et al.
2004).

Regardless of the definition, several common aspects are the key defining
concepts of a culture. Firstly, culture is not innate, but rather learned;
learning takes place in institutions starting from the family (usually being
the first institution a (lucky) individual encounters) and later on spreading
to institutions like church, school, clubs and similar; learning happens
informally and often also subconsciously; and finally, it is a shared
property of a group of people (Frith and Mueller 2007).

Furthermore, culture is a multilayered phenomenon that consists of

artefacts, values and underlying assumptions. Artefacts are the most easily
observed manifestation of culture, whereas underlying assumptions are the

39
most hidden and often the most taken for granted manifestation of culture
(Schein 1988). In this light, culture is often described as an iceberg. About
10% of the iceberg is observable and easy to spot. The remaining 90% of
the iceberg is hidden below the surface, and is the more troublesome part.
The hidden 90% of the iceberg is what sunk the Titanic. Though cultural
differences can be observed in the human behaviour of a society,
behaviour constitutes only 10% of the culture. The remaining portion is
values and norms, which are much more difficult to observe and yet cause
the most misunderstandings and failures in communication (Romani
2004).

Some research (cf. Pankhania, Lee and Hooley 2007) suggests that
marketers should take into account cultural diversity within countries as
well as between them, because culture has many boundaries. There may
be regional, ethnic, religious, generational, industry, occupational and
corporate culture, to name but a few (Brannen et al. 2004). Nevertheless,
nationality remains the most viable proxy for culture because the members
of a nation share an understanding of its institutional systems, a bond of
identity, and an experiential understanding of the world (Hofstede 2001;
Brannen et al. 2004). Prevalent value systems are a key component of
national culture and explain the differences between preferences for one
state of affairs over others (Hofstede 1985; Broderick 2007).

National culture has been defined as a society’s personality or as the glue

that binds people together (Watson et al. 2002; Pankhania, Lee and
Hooley 2007). Furthermore, it has a tendency to stay stable over the time.
This is because of its mechanism to resist change across many
generations. As Hofstede (2001) explains, the value systems shared by
major groups in the population are initially influenced by physical and
social factors (e.g. climate, geography, and demography). These value
systems are then expressed as societal norms that lead to the development

40
and maintenance of institutions with particular structures and ways of
functioning (e.g. family patterns, religion, and legal systems). Once the
institutions are established, they reinforce the societal norms and
conditions that led to their establishment (figure 2-5). For all of the afore-
mentioned reasons, the term nation is used as a proxy for culture in this
study as well.

Outside influence
Forces of nature Forces of man:
Trade
Domination
Scientific discovery

Origins Societal norms Consequences

Ecological factors: Value systems of Structure and functioning
Geography major groups of a of institutions:
History population Family patterns
Demography Role differentiation
Hygiene Social stratification
Nutrition Socialisation emphases
Economy Education systems
Technology Religion
Urbanisation Political systems
Legislation
Architecture
Theory development

Reinforcement

Figure 2-5. The Stabilising of Culture Patterns

Source: Hofstede 2001.

2.2.2. Cultural Dimensions

Despite the debate on the appropriate definition and understanding of

culture, most social scientists (e.g. Kluckhohn and Strodtbeck 1961;
Schwartz 1999; Hofstede 2001; Trompenaars 1994) have long recognised

41
the complexity and multidimensionality of the concept. For this reason,
most of them opted to describe cultures. Brannen et al. (2004) compare the
process of describing cultures as geographical map making, where a map
is a simplified version of reality. They note that a good map highlights the
important features of a space and tells the map reader the distance between
two points.

Each cross-cultural researcher depicts cultural difference by using

different cultural dimensions. Despite the differences, all of the above-
mentioned studies started their search for the dimensions by identifying
and focussing on more or less equal societal dilemmas which, although
framed a bit differently in each research, boil down to what Kluckhohn
and Strodtbeck (1961) recognise as: the nature of the individual,
relationships with other human beings, relationships with nature, time, and
the primary mode of activity (Brannen et al. 2004).

As Hampden-Turner and Trompenaars (2002) note, virtually all of the

members of the human race engaged in enterprise face the same problems
or dilemmas, but their responses to these vary widely. Table 2-2 presents
the key societal dilemmas and cultural dimensions that various researchers
have identified. According to them, these serve as a key to understand
how various cultures react when faced with such societal problems.

Hofstede (2001) greatly contributed to the field of international

management by being the first to operationalise variables into measurable
components – cultural dimensions. Before his study, which was conducted
during the 60s and 70s, culture was considered vague, intangible and soft
(Romani 2004). He conducted two survey rounds between 1967 and 1973
on 116,000 employees of IBM (at that time, but also today one of the
largest multinational corporations) in 72 countries. He then identified four
dimensions along which cultures differ. These are:

42
 a) Individualism vs. collectivism: the degree to which individuals
are supposed to look after themselves or remain integrated into
groups usually around the family3,

b) Power distance: the extent to which less powerful members of

organisations and institutions accept and expect power to be
distributed equally,

c) Uncertainty avoidance: the extent to which a culture programmes

its members to feel either uncomfortable or comfortable in
unstructured situations (situations that are novel, unknown,
surprising, or different from usual),

d) Masculinity vs. Femininity: the distribution of emotional roles

between the genders in which tough societies are considered
masculine, and tender societies are considered feminine.

In the 1980s, a continuation of cross-cultural study uncovered the

existence of a fifth dimension, at the time named Confucian dynamism
(Hofstede and Bond 1988) and later changed to Long-term vs. short-term
orientation. This dimension is defined as the extent to which a culture
programmes its members to accept delayed gratification of their material,
social, and emotional needs. This dimension is particularly important for
capturing some of the behaviours and values in South-East Asian culture.

Despite the importance and wide acceptance of Hofstede’s theory, it has

faced criticism due to the nature of his data, the number of values and
their universality, and similar (Pankhania, Lee, and Hooley 2007; Watson
et al. 2002). Even Hofstede (2001) acknowledges the criticism and tries to
offer some solutions in his consequent work. He also challenges others to

3
Some cultural dimensions are proved to be correlated to some environmental forces. The
strongest correlation is captured between individualism and national wealth. Hofstede (1997)
found that rich nations generally tend to be more individualistic.

43
offer new models that would be scientifically and practically more valid.
Nevertheless, even today, his characterisation of cultures remains the most
popular and widely used in cross-cultural business studies (Watson et al.
2002; Furrer, Liu, and Sudharsan 2000).

Table 2-2. An Overview of Societal Dilemmas and Cultural

Dimensions

Nature of the Relationship with Relationship to Time perception

Author(s)
individual other beings nature

Hofstede Individualism vs. Power distance Uncertainty Long-term vs.

2001 collectivism avoidance short-term
orientation
Masculinity vs.
Femininity

Trompenaars Universalism vs. Equality vs. Inner vs. outer Sequential time
1994 particularism hierarchy direction vs. synchronised
time
Individualism vs.
communitarism
Analysed specifics
vs. integrated wholes
Achievement vs.
ascription

Schwartz Conservatism vs. Hierarchy vs. Mastery vs.

1999 autonomy egalitarianism harmony

House et al. Assertiveness Power distance Uncertainty Future orientation

2004 avoidance
Performance Gender
orientation egalitarianism Humane
orientation
In group collectivism
Institutional
collectivism

Source: adapted from Brannen et al. 2004.

Like Hofstede, Trompenaars (1994), conducted a study in a business

setting, i.e. he collected responses from about 30,000 managers from 55
countries. His research was motivated by the idea that everyone in the

44
world seeks the same end, but employs different means to arrive at these
ends (Hampden-Turner and Trompenaars 1996). His study resulted in the
following dimensions:

a) Individualism vs. communitarism; according to which people

belonging to individualistic cultures focus upon the enhancement
of each individual’s rights, motivations, rewards and capacities,
as opposed to those belonging to communitarian cultures in
which attention is paid to the advancement of the community,

b) Achievement vs. ascription; societies more directed to

achievement are those that appreciate individuals for their
achievement and performance, while societies more directed to
ascription are those that appreciate individuals for who they are

c) Universalism vs. particularism; which differentiates between

societies in which rules should always be applied, from those in
which one should allow for exceptions,

d) Inner vs. outer direction; which defines whether a society is

predominantly populated by individuals directed to their own
inner feelings and judgments, or by individuals directed to
signals and trends in the outside world to which they must adjust,

e) Analysed specifics vs. integrated (diffuse) wholes; this defines

whether decision makers in society are more effective when
analysing parts, specifics, facts, numbers and units, or when all of
these are taken together to form integrated wholes in the wider
contexts,

f) Sequential time vs. synchronised time is the difference between

the ability to perform one activity as fast as possible before
moving on to the next, or to synchronise and coordinate more

45
 activities at the same time. Trompenaars refers to this difference
as “times as a race” vs. “time as a dance”,

g) Equality vs. hierarchy determines whether it is more important to

treat people equally so they are intrinsically motivated to give
their best, or to establish a hierarchical authority system and
judge and evaluate people to extrinsically motivate them to
achieve their best.

Another important study into cross-cultural dimensions was conducted by

Schwartz (1999). The three dimensions he discovered in a study of 49
countries are: embededness (the traditional order in which people are
embedded in their group) vs. intellectual and affective autonomy (the
desirability of individuals’ pursuing their own ideas, intellectual directions
and affective positive experiences independent of others); hierarchy (the
legitimacy of an unequal distribution of power, roles, and resources) vs.
egalitarianism (the transcendence of selfish interests in favour of
voluntary commitment to promoting the welfare of others); and mastery
(getting ahead through active self-assertion) vs. harmony (fitting
harmoniously into the environment).

Although some of the dimensions from Hofstede’s (2001) study are

repeated in the studies by Trompenaars (1994) and Schwartz (1999), both
Trompenars and Schwartz contribute with new dimensions. Trompenars’
research focuses in particular on the nature of the individual, and it
discovers four dimensions that deal with this question. Apart from
Hofstede’s dimensions of individualism, power distance and uncertainty
avoidance, those most often cited and used as antecedents in various
cross-cultural studies are Trompenaar’s specific vs. diffuse (cf. Jun and
Lee 2007) and achievement vs. ascription dimensions (cf. Gaál, Szabó,
and Kovács 2007).

46
Finally, the most recent comprehensive study was conducted between
1992 – 2004 by Robert J. House and his multinational team of researchers.
They conducted a study into 62 societies within the Global Leadership
and Organizational Behaviour Effectiveness Research Program (GLOBE)
with the aim of increasing the existing cross-cultural knowledge. Their
theory-driven study commenced in 1992 and the initial results were
published only a decade later, in 2002 (House et al. 2002), with the
complete results being published in 2004 (House et al. 2004). The
dimensions identified in this project are very similar to those of Hofstede
(2001).

However, masculinity is divided into assertiveness (representing the level

of assertiveness, confrontation, and aggressiveness in societal
relationships) and gender egalitarianism (representing the level of gender-
role differences and gender discrimination). The study was also enriched
by a measurement of human orientation (expressing whether a society
encourages its members to be fair, generous, altruistic, caring, and kind)
and performance orientation (expressing the extent to which a society
encourages performance, improvement and excellence). The first
dimension is derived from Kluckhohn and Strodtbeck’s (1961) study of
the societal problem of human nature. The second dimension is derived
from McClelland’s (1961) study on the need for achievement, and
resembles Trompenaars (1994) achievement vs. ascription dimension.

Moreover, individualism vs. collectivism is the most thoroughly studied

cultural dimension (which all the above-mentioned studies recognise as a
distinctive dimension.) Some further studies also focus on this dimension
and recognise that individualism is not an absence of collectivism, but
rather a phenomenon in its own right (Schimmack, Oishi, and Diener
2005). Gaines et al. (1997) define individualism as an orientation towards
the welfare of oneself, and collectivism as an orientation toward the

47
welfare of one's larger community. Furthermore, they provide evidence
that the two constructs are uncorrelated rather than negatively correlated.
In addition, they suggest a new construct, namely familism, which is
defined as an orientation toward the welfare of one's immediate and
extended family. Triandis and Gelfand (1998) consider individualism to
be a four dimensioned construct, differentiating between Vertical
individualists - who score high on competition and hedonism, Vertical
collectivists - who score high on family integrity, Horizontal individualists
- who score high on self-reliance, and Horizontal collectivists - who score
high on family integrity and sociability and low in emotional distance
from in-groups. House et al. (2004) also recognised the
multidimensionality of this scale. They explicitly differentiate institutional
collectivism (expressing whether an organisation rewards the collective
distribution of resources and collective action) and in-group collectivism
(expressing whether an individual sees him or herself as part of a group,
be it an organisation or a family).

Finally, the study by House et al. (2004) differs from all the previously
elaborated studies in one important aspect. Hofstede (2001) speaks of a
cultural “onion” – where values make up the innermost layer of the onion;
and practices consisting of rituals, heroes and symbols make up the outer
layers of the onion. According to him, values determine practices; hence,
he does not consider it useful to study the two manifestations separately.
House et al. (2004), on the other hand, did not hold this assumption to be
true, so they tested cultural values and cultural practices separately. This is
also in line with Kroeber and Kluckhohn’s view on the necessity to
differentiate between three classes of data: people’s notion of the way
things ought to be done (values); their concepts of the way the group
actually behaves (practices), and what actually occurs in reality (Bulmer
1953).

48
As a result of separately measuring the two manifestations, House’s team
found huge differences between them. In general, the pattern across most
cultural clusters repeats, so: individuals value performance orientation,
future orientation, gender egalitarianism, in-group collectivism,
institutional collectivism, human orientation, and uncertainty avoidance
more than they practise them; while at the same time, they value
assertiveness and power distance less than they practise them. Exceptions
are observable for the Nordic and Germanic Europe clusters in respect to
uncertainty avoidance, for Nordic Europe and Confucian Asia in respect
to institutional collectivism, and for Confucian and Southern Asia in
respect to assertiveness (House et al. 2004).

Finally, it must also be acknowledged that research on cultural dimensions

particularly stresses the need to avoid the trap of stereotyping that an
overly shallow application of the cutural dimension may lead to. In other
words, scores along cultural dimensions signify an average country score,
which denotes that some people, for example, in a highly assertive culture,
might be less assertive than an average person from a generally low
assertive country (Frith and Mueller 2007). This is presented in figure 2-6.

Figure 2-6. Cultural Differences Distribution

Source: Romani 2004.

49
In this chapter, the concepts of brand and culture were presented.
Additionally, brand meaning and cultural dimensions were introduced as
the particular points of interest of this research. The coming chapter will
explore the relationship between these two concepts.

50
 3. Previous Research and Hypothetical Model

This chapter takes a deeper look at previous research exploring a direct

interaction between brand and culture. Then two auxiliary concepts (brand
benefits and consumer needs) are introduced into the aforementioned
relationship. Finally, a hypothetical model of this research is developed.

3.1. The Influence of Culture on Brand Meaning

While some (e.g. Levitt 1983) argue that the globalisation of markets is
inevitable, substantial research has cautioned that the evidence of growing
standardisation is misleading and superficial (e.g. Gram 2007). The role of
culture in business affairs has primarily been studied in regards to
international management, and recently a considerable body of research
has developed in the fields of international consumer behaviour,
marketing and brand management.

In general, research in cross-cultural marketing has discovered that culture

might motivate corporate decision makers to deliberately choose a
different brand positioning in different countries. However, even if
identical positioning is applied across nations, the same brand might be
perceived differently in different cultures (Foscht et al. 2008). The
difference may be caused due to: the adaptation of advertising to better
adjust to cultural traditions and norms; and different understandings of
applied communication. After a short introduction to the first issue, the
focus of the subsequent chapters switches to the heart of this research, i.e.
culturally influenced brand perceptions.

51
3.1.1. The Influence of Cultural on Brand Communication

A survey of leading brand advertisers in 15 countries discovered major

differences in creative approaches among countries in terms of the appeals
and messages portrayed (Synodinos, Keown, and Jacobs 1989). For
example, while no Swedish advertisements refer to brand benefits, 77% of
the Singaporean ones do; at the same time, Swedish advertisements most
often use spokespersons and refer to product itself. The reason for creative
communication differences among cultures might be ascribed to different
marketing and advertising traditions; the practice of different advertising
reflecting different national cultures; the processes of persuasion in
general, and the patterns of emotions in particular (Orth, Koenig, and
Firbasova 2007).

However, differences in brand communication might also be ascribed to

the transfer of the image and personality of those in contact with a brand
(brand managers, creative agencies, sales representatives and the like) to
the brand (McCracken 1986). While McCracken’s seminal paper proposed
a conceptual model of how personality and images transfer from people to
brands and back, Blythe (2007) provided empirical evidence for the
model. He discovered that advertising creatives (i.e. persons creating
advertisements) impact the final brand personality as they tend to produce
a piece of work that reflects their own personalities.

Research into cross-cultural differences in advertising has gone a step

further and instead of only identifying the differences, it has also
endeavoured to explain the roots of the differences by using cultural
dimensions. For some, cultural dimensions served as a posterior
argumentation to explain the findings, and to others, they served as the a
priori building blocks of their hypothetical model.

52
For example, Alden, Steenkamp, and Batra (1999) investigated humour in
advertisements across cultures and found that differences in Hofstede’s
dimensions implied differences in the use of humour in advertising as
well. Zhang and Neelankavil (1997) suggested and, through their results,
confirmed that messages that emphasise individualistic values and benefits
are more effective in the United States, which is an individualistic country
in comparison to collectivistic China. Similarly, Cho et al. (1999) found
partial evidence that cultural differences lead to more of an individualistic
element within commercials in the US, in comparison to Korea, where the
commercials reflect the more community-oriented nature of the culture.
Jun and Lee (2007) also focus on the United States and Korea, but the
object of their research is the adjustability of corporate visual identity
(logos and taglines) to Trompenaars’ (1994) cultural dimension, specific
vs. diffuse. They found that Korean brands are generally more diffuse,
which was observable in their more abstract and symbolic creative
designs.

3.1.2. The Influence of Cultural on Brand Communication

Interpretation

As previously discussed, research over the years has shown that

consumers actively participate in the brand meaning creation (Cova and
Pace 2006; Richins 1994; Christodoulides 2008; Zumpano 2007). In line
with the personality transfer theory (McCracken 1986), consumers also
transfer their own personalities to the brands. This transfer takes place
partly in the consumption process, during which consumers tend to
perceive the brand personality as they would like it to be and see their
preferred personality traits in the brand (Phau and Lau 2000). Perhaps
even more importantly, the transfer also takes place during the word-of -
mouth process, when consumers speak about the brand to their friends and

53
families (De Matos and Rossi 2008). In other words, the personality
perception is influenced by the personality preference of the consumers.

A number of researchers have argued that national cultural differences are

in fact engraved in consumers’ perceptions (Watson et al. 2002).
Therefore, although metaphors are used by advertising creators to convey
brand meaning and enhance brand information processing, little is
understood about consumers' comprehension of intended meaning
(Morgan and Reichert 1999), and even less about consumers’
comprehension in different cultural settings. For consumers from different
socioeconomic backgrounds, it is likely that the same stimuli do not
necessarily build equivalent symbolic associations (Khalid and Helander
2004), but rather very different sets of metaphors, personal meanings, and
cultural traditions of meanings (Thompson, Pollio, and Locander 1994).

This is because the messages, events and experiences that consumers find
the most meaningful are those which are the most connected to their core
values (Dilts and DeLozier 2000). Due to differences in culturally based
traditions, religions, and histories, individuals in distinct cultures tend to
hold different sets of values and preferences (Aaker 2000). Hence, altering
beliefs and values can immediately change the meaning of transmitted
messages, events and experiences (Dilts and DeLozier 2000). The same
experience or incident takes on different meanings to different individuals
depending on their internal mind maps and embedded previous knowledge
(Berthon, Pitt, and Campbell 2008). To summarise, culture acts as a
stringent screener that greatly influences the meaning of advertising
messages (Jun and Lee 2007).

Berthon, Pitt, and Campbell (2008) speak of brand meaning and suggest it
is an outcome of brand communication and the knowledge base of the
recipient of the communication (figure 3-1). According to them, the same

54
communication to different audiences will result in shared meaning if, and
only if, the different groups share a common knowledge base.
Furthermore, when people within the same culture assign the same
meaning to a stimulus, clear meanings are deemed to exist and clear
meanings are preferred and recognised (Henderson et al. 2003).

R1 M1

S C R2 M2

R3 M3
KR1
S = sender of communication «C»
KR2 R = recipient of communication
K = knowledge base of recipient
M = meaning ascribed to
KR3
communication

Figure 3-1. Making Meaning: Communication, Knowledge and Meaning

Source: Berthon, Pitt, and Campbell 2008, p. 3.

The literature also provided evidence that people in different cultures do

not process information in the same way (Nisbett et al. 2001).
Specifically, scholars found that East Asians are more field dependent.
They pay great attention to the context and focus on the relationship
between objects and the field (Hall 1989).

In their study of the variations in consumer interpretation of visually

complex advertising, Bulmer and Buchanan-Oliver (2006) explain that the
interpretation of visual stimuli in advertising is linked to literary and
cultural traditions that differ from cultural group to cultural group. Monga

55
and John (2007) conducted a study in which they hypothesise and then
confirm through empirical research that consumers from Eastern cultures,
who tend to be holistic thinkers, perceive a higher brand extension fit and
evaluate brand extensions more favourably than their more analytical
Western counterparts do.

In another study, Costa and Pavia (1992) speak of numbers and their
meanings, and confirm with an experiment that brands consisting of
numbers have an extra meaning that is understood only by some societies
(in their study American), due to the “excess meaning” of certain numbers
in those cultures. Finally, Gram (2007) discovered that although East
Asians prefer advertisements portraying Eastern values for some products,
especially luxury goods, Western advertisements not adapted to Asian
tastes will be more successful. However, the success is not due to the
universal content of such an advertisement; but quite the opposite, for
being understood and interpreted as Western.

Thus, the studies discussed above present a solid basis to acknowledge

culturally embedded meaning creation and to introduce the first
hypothesis of this research:

H1: The meaning of certain brands differs for consumers in different

cultures.

The following subchapter further expands the topic by providing results

from previous research which acknowledges the effect of culture on
consumer brand perceptions and understandings, and contributes to the
field by providing evidence for the effect of particular cultural dimensions
in brand interpretations.

56
3.1.3. Cultural Dimensions as an Antecedent to Brand Meaning

Phau and Lau (2000) suggested that understanding cultural meaning

requires linking it to the cultural dimensions identified by Hofstede (2001)
and Trompenaars (1994). They propose that the meanings of brand
personality dimensions differ between members of individualistic and
collectivistic cultures. Brand personality traits that are consistent with
interdependent self-construals (e.g. dependent, peaceful and harmonious)
are emphasised by members of the collectivistic cultures, while in
contrast, those consistent with independent self-construals (e.g.
achievement, competition and independence) are emphasised by members
of individualistic cultures.

Foscht et al. (2008) empirically tested the proposition by Phau and Lau
(2000). They used Hofstede’s (2001) cultural dimensions and Aaker’s
(1997) brand personality dimensions to test whether brand perceptions are
similar in the six chosen countries. The results provide clear evidence that
the same brand is perceived differently in different cultures in spite of
identical positioning. Although they assumed that Hofstede’s dimensions
would play a role in the perception of brand personality, they did not test
the influence of the particular cultural dimension on the extent to which a
certain brand personality perception is emphasised.

Similarly, Pankhania, Lee, and Hooley (2007) incorporated the effect of

the cultural dimensions of power distance and collectivism in their study.
Their findings show that collectivist cultures assign more importance to
the social value of brands and at the same time perceive a certain brand to
be more socially-oriented than individualistic cultures do.

The afore-mentioned studies by Foscht et al. (2008) and Pankhania, Lee,

and Hooley (2007) provide clear evidence that cultural dimensions
influence brand perceptions by altering brand personality, brand attributes

57
and brand values. The most relevant findings for the purpose of creating a
link between culture and brand meaning are the studies by Watson et al.
(2002) and Quester, Beverland, and Farrelly (2006). Quester, Beverland,
and Farrelly (2006) studied differences in brand meanings for the
members of extreme sports subcultures and found that brand meanings
and relevance appear to shift in accordance with an individual’s relative
involvement in the subculture - the primary value that links one to the
subculture, and pressures experienced by the subculture as a whole, such
as mainstreaming.

Watson et al. (2002) studied why people possess certain things. Although
they focus on things and not brands, their findings are easily applicable to
brands as well. They discovered that the reason for the possession of
certain things is the private meanings that those things have for their
owners. The entirety of these meanings represents the totality of an
individual’s thoughts and feelings about the importance of an object
(Richins 1994), and characterises a person’s individual values. Thus, the
same physical object has a different symbolic meaning for people who
have different value systems. For instance, a person who values
“conservatism” may prize an old ink pen as a family treasure because it
symbolises family ties and heritage. Alternatively, a person who values
“mastery” may treasure the same ink pen because it symbolises prestige
and enhances his or her self-esteem. In other words, each person extracts
the meanings from an object or a brand that he or she values the most.

Thus, the results of the research presented above lead to the second
hypothesis of this research:

H2: Brand meaning is affected by cultural dimensions.

In order to be able to predict how cultural dimensions affect brand

meanings, a model that connects cultural dimensions to brand meanings

58
has to be built. Indicating the need for such a model, Watson et al. (2002)
argue that most cross-cultural studies are primarily based on comparisons
rather than theory. In other words, cultural differences in consumer
behaviours and attitudes are usually simply observed and described, rather
than a priori hypothesised based on theory, and then empirically tested.
This research accepts the challenge of developing a model predicting
brand meaning based on cultural dimensions. The model will be presented
in the next chapter.

3.2. From Culturally Determined Needs to Brand

Meanings

The connection between cultural dimensions and brand meaning is

predicted to be mediated by two concepts, namely consumer needs and
brand benefits. Cultural heritage defines needs that are important to the
consumers of a particular society (House et al. 2004; Hofstede 2001). On
the other hand, a brand provides a bundle of benefits (Kapferer 2004).
Consumers are predicted to mainly benefit from those brand benefits that
fulfil their needs. Apart from the main purpose of the brand experience
and extraction of the main brand benefit, a consumer is also predicted to
extract those additional benefits that satisfy her or his emphasised needs.
Experience (i.e. the benefits for consumers in an experience) establishes a
network of associations in the consumer’s mind (Krishnan 1996). Such a
network is referred to as brand meaning.

The process of brand meaning creation in the consumer’s mind is depicted

in figure 3-2. As the figure shows, apart from key brand benefits
(represented by a white ellipse), a brand holds those meanings (grey

59
shapes) that stem from a customer’s culturally emphasised needs
(represented by the black triangle and the rectangle).

Culturally determined Brand meanings Brand benefits

consumer needs

Legend

Emphasised Emphasised Key brand

consumer brand benefit
needs meanings

Figure 3-2. Consumer Needs and Brand Benefits in Creating Brand

Meanings

The following paragraphs will introduce the concept of consumer needs

and brand benefits, and then explain in which way they mediate the
relationship between cultural dimensions and brand meanings.

3.2.1. Consumer Needs

It is the job of marketers to discover and satisfy consumer needs. Needs

are defined as those desires which motivate people to behave in a certain
way to achieve satisfaction (Maslow 1943; Raiklin and Uyar 1996).
Traditionally, marketing experts have done their job by supplying
consumers with products that provide functional benefits. Over time
however, as they discovered that consumers were developing needs linked

60
to particular brands (as opposed to needs directed to generic products),
marketers started to provide added value through brands. These needs,
again, evolved with time. As Pringle and Thompson (1999, p. I)
recognised, consumers mainly needed rational brand benefits such as
quality guarantees in the 50s; two decades later, more consumers needed
emotional benefits such as a feeling of happiness or belonging; whereas in
the 90s, some pioneering individuals opted for spiritual and ethical
benefits such as acknowledgement of their ethical behaviour. Brands that
could satisfy these emerging consumer needs established long-lasting and
meaningful brand relationships.

The evolution of these needs resembles Maslow’s (1943) hierarchy of

needs, which classifies needs from the most essential physiological needs
to the most advanced self-actualisation needs. He suggests that at the most
basic level, the most essential and primitive needs, which are usually
taken as the starting point for motivation theory, are the so-called
physiological needs. They include hunger, thirst, sleep and the like. The
next level of needs represents safety needs, which are expressed as the
need for shelter from any kind of real or perceived threat. The third level
of needs is represented by love needs as they were originally named by
Maslow, but later rephrased into love and belonging needs or social needs.
The fourth level of Maslow’s needs refers to esteem needs, which are
defined as needs or desires for a stable, firmly based evaluation of oneself
based upon real capacity, achievement and respect from others. The
highest level of needs, self-actualisation, is characterised by people’s wish
to fulfil their potential and possibly contribute to their society’s overall
welfare for the sake of doing good and not for the sake of being
recognised and merited by that act.

It is important for this particular study to stress that later cross-cultural

research questioned and negated the hierarchical order of Maslow’s needs.

61
For example, in a study of Korean workers, Raymond, Mittelstaedt, and
Hopkins (2003) discovered that social needs are the strongest motivator,
followed by esteem needs, physiological needs (although somewhat
modified to the working environment conditions), safety needs and self-
actualisation needs. Similarly Hofstede (1980) convincingly argued that
the hierarchy that Maslow established was not universally applicable
across cultures due to variations in those cultures. So, instead of speaking
of a hierarchy, it might be more appropriate to speak of categories and
cultural determination of the dominant category of needs.

Apart from Maslow (1943), many other psychologists and anthropologists

have studied needs as drivers of human decision-making and behaviour.
One of the most important and widely cited works is that of Herzberg
(1968). He conducted the study in a working environment and as a result
discovered satisfactory needs and dissatisfactory needs. The former
motivate people and can enforce the level of satisfaction, but their non-
fulfilment will not result in dissatisfaction. The later, which are also called
hygiene needs, result in dissatisfaction when non-fulfilled, whereas their
fulfilment does not significantly affect the level of satisfaction.

In another study, Scitovsky (1968) classifies needs into a more

parsimonious version of Maslow’s needs. He differentiates between
physiological, social and intellectual needs. Raiklin and Uyar (1996)
identified biological and socio-cultural needs, while Ryan and Deci (2000)
suggest competence needs, autonomy needs and relatedness needs,
especially in a job environment.

The categorisation of needs was also studied from a marketers’ point of

view, i.e in terms of consumer needs (e.g. Khalid and Helander 2004).
However, such studies mostly deal with very specific situational and
product related needs. Because this study is embedded in a social context,

62
i.e. it studies culturally determined consumer differences in brand
consumption, it is essential to recognise that consumer needs stem from
human needs. Consumer needs are the application of human needs in the
consumption process. Therefore, to understand consumer behaviour in the
brand consumption process, the underlying needs (i.e. basic human needs)
have to be studied. For this reason, Maslow’s needs serve as the building
blocks of the model in this research.

3.2.2. Brand Benefits

It was long assumed that consumers consumed a particular product in the

consistent manner no matter what the occasion was. However, during the
90s, it became obvious that the same object can be used in different ways
(Holt 1995) and for different purposes (Sheth, Newman, and Gross 1991).
This is because a product is a bundle of features and attributes that
provides a bundle of benefits. Initially it was believed that solely
functional characteristics of products provide values and benefits to
consumers; however, the seminal article by Srinivasan (1979) greatly
influenced the discourse and was the start of understanding the concept of
a brand as a value provider as well. The terms ‘value’ and ‘benefit’ are
often used interchangeably in the discourse on brand benefits. However,
brand value is also used to denominate the financial value that a brand
provides to a company (cf. Interbrand and Business Week 2008; Kapferer
2004), whereas the focus of brand benefit is more on value for consumers.
The term brand benefit is thus considered more appropriate for this study
and is used henceforth.

Benefits are the personal values that consumers attach to the product or
service attributes. In other words, benefits are what consumers think the
product or service can do for them (Keller 1993). Brand benefits, as most
of the concepts in marketing, have been categorised by several authors.

63
One of the most important categorisations is that of Park, Jaworski, and
Maclnnis (1986). They recognise functional benefits, experiential benefits
and symbolic benefits.

Another categorisation is provided by Sheth, Newman, and Gross (1991),

who differentiate between functional value (functional, utilitarian, or
physical performance); emotional value (the capacity to arouse feelings or
affective states); social value (the association with one or more specific
social groups); conditional value (the utility acquired as the result of a
specific situation or a set of circumstances facing the choice maker); and
epistemic value (the capacity to arouse curiosity, provide novelty, and/or
satisfy a desire for knowledge). Later, Sweeney and Soutar (2001)
proposed a reduced categorisation of benefits that consisted of only four
values: functional value in terms of performance and quality, functional
value in terms of value for money, emotional value, and social value.

In their attempt to classify benefits, Kim and Mauborgne (2000) made use
of terminology that was more brand specific than that of their
predecessors. According to them, brands provide the following benefits:
customer productivity (helping consumers to do things better, faster or
differently); simplicity (straightforward usage and easier understanding
than existing offerings), convenience (availability, ease of consumption
and purchase), risk reduction (safety of the product/service consumption
process, safety of the investment), fun and image (amusement or
enjoyment arising from the use or purchase of the brand) and positive
image portrayal (the character, reputation, mental representation, idea or
conception of the brand as perceived by the consumer), and environmental
friendliness (the capacity of the service to improve or reduce risk to the
environment).

64
Finally, Kapferer (2004, p. 23) also proposed his own types of brand
benefits. According to him, a brand can provide its consumers with eight
different types of benefits. Starting from the more simple and basic brand
benefits, to those consumer specific benefits, they are: identification
(providing quick identification of the sought-after products), practicality
(providing time and energy efficiency), guarantee (providing a constant
standard of quality), optimisation (ensuring customers buy the best value
for money), badge (providing confirmation of one’s self-image or image
presented to others) continuity (providing satisfaction through a
relationship with the brand), hedonism (enabling self enjoyment linked to
the attractiveness of the brand), ethics (ensuring the recognition of socially
responsible behaviour by using brands that propagate such values).

It is clear that although most of the authors refer to brand benefits, they
clearly attach product benefits as well as brand benefits to the concept.
Even Keller (1993) does this in his earlier research. Some of the brand
benefits he refers to are actually benefits that a product or a service
provides (and not the brand itself). The term brand benefit has come to be
more specifically recognised only recently, and increasingly, researchers
are splitting products from brands (Miliopoulou 2007) and speaking solely
of brand benefits.

Because of their brand specific context, Kapferer’s types of benefits will

be used in the model of this study.

3.2.3. The Relationship between Culture, Consumer Needs, Brand

Benefits and Brand Meanings

After having delved into the two supporting concepts of the conceptual
model, this subchapter explores previous research on the relationship

65
between the concepts, and it suggests the sub-hypotheses of hypothesis 2
that will be tested in the empirical phase of this study.

Roth (1995) discovers that cultures with high collectivism express more
attraction towards social values as a result of their higher social needs. On
the other hand Schimmack, Oishi, and Diener (2005) recognise that highly
individualistic cultures value freedom, independence, and individuality. At
the same time, the need for group approval is lacking in such cultures
(Leo, Bennett, and Härtel 2005).

In order for a brand to be able to satisfy a consumer’s social needs,

agreement about its meaning is necessary on: its physical make-up, its
functional characteristics and its characterisation - i.e., personality (Ligas
and Cotte 1999). When these conditions are fulfilled, brands can provide
the benefit of identification and social bonding to one’s family,
community, and/or cultural group (Muniz and O’Guinn 2001). Even
identification to a brand-centred group like a brand community,
subculture, consumer tribe or brand tribe (Cova and Pace 2006; Veloutsou
and Moutinho 2008), be it observable or psychological (Carlson, Suter,
and Brown 2007), can often be achieved.

Numerous brands even base their existence on providing content for brand
communities (Thompson, Rindfleisch, and Arsel 2006), some of the most
prominent among them being Harley Davidson (McAlexander, Schouten,
and Koenig 2002) and Apple Macintosh (Muniz and Schau 2005).

The important role that brands play in social life is perhaps best captured
in the comment by Renzo Rosso, Diesel’s executive who once said: "We
don't sell a product; we sell a style of life... The Diesel concept is
everything. It's the way to live, it's the way to wear, it's the way to do
something" (Klein 2000, p. 29). Diesel’s positioning statement most
closely represents the badge type of benefit in the way that collectivists

66
will use Diesel to express their similarities and belonging to the reference
group; whereas individualists on the other hand might utilise the same
brand to express their uniqueness (Phau and Lau 2000) and differentiate
themselves from the mass by creating their self identities (Escalas and
Bettman 2005). Hence, the following might be hypothesised:

H2a: For consumers from cultures displaying higher collectivism,

brand meanings will be more related to group activities, belonging,
network, sacrifices for others, and common goals than for consumers
from cultures expressing lower collectivism.

H2b: For consumers from cultures displaying lower collectivism,

brand meanings will be more related to themes like individualism,
autonomy, freedom, responsibility, self-sustainability than for
consumers from cultures expressing higher collectivism.

Hofstede (2001) claims that uncertainty avoidance assumes the need for
security. This need is often fulfilled by rules. Barr and Glynn (2004) tested
and confirmed this assumption. Specifically, they found that the more
cultures avoid uncertainty, the more they associate controllability with
opportunity and the lack of it with threat. This indicates their need for
safety. Similarly, Verhage, Yavas, and Green (1991) discovered that the
degree to which perceived risk influences decision-making may vary
between countries. The mean perceived risk scores for the products they
studied were significantly lower in low uncertainty avoiding cultures (i.e.
Turkey and Thailand) as opposed to high uncertainty avoiding cultures
(i.e. the Netherlands). Therefore, it can be concluded that consumers from
cultures displaying higher uncertainty avoidance will have more of a need
for safety than consumers from cultures displaying lower uncertainty
avoidance.

67
Safety needs can be satisfied by brands as well. They are, to a large extent,
the reason why brands were introduced in the market. Rather than
engaging in a detailed search for information when deciding between
competing brands, consumers use brands as clues to indicate product
performance (Lim and O’Cass 2001). In such situations, brands provide
the benefits of quality guarantee, identification, optimisation, value for
money, and search cost reduction (Keller 2003b, p. 9).

Furrer, Liu, and Sudharsan (2000) have shown that the relative importance
of the service quality dimensions varies from one culture to another, and is
predictable by the scores on Hofstede’s (2001) dimensions. Obviously,
consumers who have higher safety needs extract more quality guarantee
benefits from brands than consumers who have lower safety needs. For
example, according to Erdem, Swait, and Valenzuela (2006), credible
brands provide more value to high uncertainty-avoidance consumers
because such brands have a lower perceived risk and lower information
costs. Continuing with this line of thought, it can be hypothesised:

H2c: For consumers from cultures displaying higher uncertainty

avoidance, brand meanings will be more related to high quality, safety,
value for money, standards, reliability, order, definitions, and
consistency than for consumers from cultures displaying lower
uncertainty avoidance.

H2d: For consumers from cultures displaying lower uncertainty

avoidance, brand meanings will be more related to risk, adventure,
innovation, excitement, unpredictability than for consumers from
cultures displaying higher uncertainty avoidance.

Cultures that display high power distance also display a need for status
and prestige (Leo, Bennett, and Härtel 2005). Schimmack, Oishi, and
Diener (2005) refer to power distance as vertical individualism, which

68
they claim is represented by the need for status and hierarchical
organisation of society. Individuals in high power distance societies are
thus considered to place more emphasis on esteem needs and tend to
observe brands as demonstrators of power. Thus, the corresponding
hypotheses are as follows:

H2e: For consumers from cultures displaying higher power distance,

brand meanings will be more related to power, admiration, class
difference, honour, and being better than others than for consumers
from cultures displaying lower power distance.

H2f: For consumers from cultures displaying lower power distance,

brand meanings will be more related to being approachable, equal,
humble, and having a low profile than for consumers from cultures
displaying higher power distance.

House et al. (2004) elaborate that countries low in assertiveness focus on

the quality of life rather than performance; and on relationships between
people rather than money and things. On the other hand, consumers in
cultures displaying higher assertiveness will have a greater need for
esteem and self-actualisation. At the same time, consumers in countries
high in performance orientation will also value development, good
performance, individual achievement and competition.

Since global brands, in particular, are powerful symbols to portray cultural

meanings such as prestige, progress and modernity (Zhou, Teng, and Poon
2008; Kinra 2006, Kumar, Lee, and Kim 2008), they are assumed to be
better liked by highly assertive and highly performance-oriented cultures.
Similarly, Bodkin, Amato, and Peters (2008) suggest that spectators as
consumers of a sporting event tend to integrate and identify themselves
with an archetype of a sportsman. Since sportsmen as an archetype are
usually successful, achievers, winners and the like, a brand being endorsed

69
by a sportsman is very often an appropriate source of identity construal for
consumers from cultures high in assertiveness and performance
orientation.

Park, Milberg, and Lawson (1991) discovered that Rolex is associated

with status, wealth, luxury, fashion, etc. These are clearly the expressions
of the self image of a consumer from a culture high in assertiveness and
performance orientation. Therefore, such consumers are more likely than
others to use brands mainly because brands help them to express their
personal or business successes. However, this is not to say that the
symbolism only lies in some particular brands, but rather it is in a
consumer’s interpretation of brands in general. Proof of this can be found
in a study by Coulter and Zaltman (1994). In a brand mapping process,
they discovered that consumers connect the strong and dependable Tide
detergent to self-confidence. They further claim that such associations to
Tide help the brand’s consumers to be perceived as self-confident people.
Thus, if Tide as a non-personal brand can provide such a benefit for some
consumers, then it is plausible that any brand might have this ability.
Hence, it can be concluded that a brand’s success is not entirely an
objective category, but rather depends on subjective perception in a way
that:

H2g: For consumers from cultures displaying higher assertiveness,

brand meanings will be more related to assertiveness, advancement,
toughness, aggression, control, and competition than for consumers
from cultures displaying lower assertiveness.

H2h: For consumers from cultures displaying lower assertiveness,

brand meanings will be more related to care, weakness, helpl,
solidarity, tradition, life quality, and ambiguity than for consumers
from cultures displaying higher assertiveness.

70
 H2i: For consumers from cultures displaying higher performance
orientation, brand meanings will be more related to efficiency,
effectiveness, success, competitiveness, progress, and target orientation
than for consumers from cultures displaying lower performance
orientation.

Furthermore, cultures high in human orientation have an enhanced self-

actualisation need which is satisfied by caring for others, being altruistic,
protective, etc. (House et al. 2004). Consumers from such cultures opt for
spiritual, responsible behaviour by decisively choosing brands that they
consider socially, ethically and environmentally responsible. For example,
such brand is the retail chain The Body Shop. Its founder, Anita Roddick,
explained that her stores are not about which product they sell, but about a
grand idea - a political philosophy about women, the environment and
ethical business (Klein 2000). Such communal brands serve as a
foundation of group identification and social solidarity experiences that
consumers with high human orientation look for in brand consumption
(Thompson 2004). Therefore, it can be concluded that a socially
responsible brand will provide these consumers with more spiritual
satisfaction than consumers who have lower human orientation. Hence,
the hypotheses:

H2j: For consumers from cultures displaying higher human

orientation, brand meanings will be more centred on love, care, social
responsibility, and the environment than for consumers from cultures
displaying lower human orientation.

H2k: For consumers from cultures displaying lower human orientation,

brand meanings will be more centred on self-interest, material
pleasures, parties, and hedonism than for consumers from cultures
displaying higher human orientation.

71
And finally, along similar lines, for the last two cultural dimensions of
future orientation and gender egalitarianism, the following hypotheses can
be constructed:

H2l: For consumers from cultures displaying higher future orientation,

brand meanings will be more related to investing, planning, saving,
and sacrifices for the future than for consumers from cultures
displaying lower future orientation.

H2m: For consumers from cultures displaying higher gender

egalitarianism, brand meanings will be more related to gender equality
and fairness than for consumers from cultures displaying lower gender
egalitarianism.

In conclusion, table 3-1 summarises the afore-hypothesised connections

between the four studied concepts: cultural dimensions (by House et al
2004), consumer needs (by Maslow 1943), brand benefits (by Kapferer
2004) and brand meanings. Most of the cultural dimensions are bipolar,
i.e. cultures that are high in a certain dimension display a highly specific
set of characteristics; whereas cultures that are low in that same dimension
do not display a low level of the same set of characteristics. Instead, they
display a specific set of their own characteristics. Therefore, for most
cultural dimensions two hypotheses are constructed.

Finally, it is essential to note that the sub- hypotheses do not assume that
different meanings will be ranked differently in either of the above-
mentioned sub-hypotheses (although this may actually be the case). It is
hypothesised that certain meanings will be more emphasised in one
culture over another across all the studied brands.

72
Table 3-1. The Relationship between Cultural Dimensions, Consumer
Needs, Brand Benefits and Brand Meanings

Cultural dimension a Consumer need b Brand Benefit Brand Meaning

H
(House et al. 2004) (Maslow 1943) (Kapferer 2004)
H2a Collectivism + Social Badge Belonging, network,
sacrifices for others,
common, group
H2b Collectivism - Esteem Hedonistic Individual, autonomy,
Continuity freedom, responsible,
self-sustainable
H2c Uncertainty avoidance Safety Recognition cue High quality, safety,
+ Quality guarantee value for money,
Optimisation standards, reliable,
Continuity order, definitions,
consistency
H2d Uncertainty avoidance Esteem Badge Risk, adventure,
- Hedonistic innovation, excitement,
unpredictable
H2e Power distance + Esteem Badge Power, admirable, class
Hedonistic difference, honourable,
better than others
H2f Power distance - Social Badge Approachable, equal,
humble, low profile
H2g Assertiveness + Esteem Badge Assertiveness, advanced,
tough, aggression, in
control, competition
H2h Assertiveness - Social Badge Helpful, care, weakness,
Ethical solidarity, tradition, life
quality, ambiguity
H2i Performance Esteem Practicality Efficiency, success
orientation + Identification effectiveness, progress
Badge competitiveness, target
H2j Human orientation + Social Ethical Loving, care, social
responsibility,
environment
H2k Human orientation - Hedonistic Self-interest, material
pleasure parties, tasty
H2l Future orientation + Safety Continuity Investing, planning,
saving, sacrifice for the
future
H2m Gender egalitarian.+ Social Ethical Gender equality, fair
a
the high pole of cultural dimension is indicated by +, whereas the low by –
b
The self-actualisation need is a very individually expressed type of need that might result in
very different types of benefit extractions and very different brand meanings. For this reason,
it is not included in table.

73
3.3. The Effect of Brand Usage and Brand Affection

Although the main focus of this research is the relationship between

culture and brand meaning, it is also necessary to examine the role that
brand usage has in this relationship. For any brand, there are at least two
broad segments of customers: users and non-users. When brand affection
is added, users as well as non-users might further be divided into those
who like and those who do not like the brand (Griffin 1997). The four new
groups might be named: Engaged users, Disengaged users, Engaged non-
users and Disengaged non-users. All these consumer segments are
important for future brand profitability in different ways. Understanding
the current users’ motives, thoughts and feelings contributes to
maintaining current sales; while understanding non-users provides an
insight into potential areas of improvement to increase the brand’s
customer base and potential sales.

Grier and Brumbaugh (1999) stress that the meanings constructed by

target and non-target consumers may differ. Although they do not
explicitly speak about users and non-users, their target and non-target
consumer classification closely resembles the user and non-user
classification in terms of how these two groups might understand the
brand.

Along the same lines, Phau and Lau (2000) note that previous research
found that consumer buying behaviour differs based on user type (user vs.
non-users). They also propose that the user type may also have an
influence on the concept of self-congruity, since it is one of the constructs
that defines consumer behaviour. In addition, Foscht et al. (2008) also
suggest that there is a relationship between the perception of brand
personalities and type of user. Almost all personality traits (being positive)
are perceived more strongly by users than non-users. In this thesis, brand

74
meaning is investigated, which unlike brand personality might in some
cases be negative. Taking into account the possible negative brand
meanings as well as further differentiating users into engaged and
disengaged users, and non-users, the following hypothesis may be
constructed:

H3: Brand users (especially engaged ones) will emphasise positive

brand meanings more than non-users, whereas non-users will
emphasise negative meanings more than brand users (especially
engaged ones).

Furthermore, if the relationship between culture and brand meaning is put

in the context of the type of user, the results can provide further important
insights and give more value and applicability to the overall findings. This
means that differences in brand meanings emphasis are predicted to vary
between users and non-users to a different extent in different cultures.
Similar findings were obtained in a study by Foscht et al. (2006), who
discovered that almost all personality traits are emphasised more by users
than non-users, but when the results are observed by country, differences
in some personality traits emphasis occur in certain countries, whereas
emphasis of some other traits occur in other countries. Based on this
elaboration, it can also be hypothesised that:

H4: Differences in brand meaning emphasis between different types of

users are not equal in each country.

Figure 3-3 presents a hypothetical model of this study. Because hypothesis

2 contains 13 sub-hypotheses, only the main one is depicted in the figure.

75
 Culture

Cultural H1
Dimensions

H2 Brand Meaning

H4
H3
Type of User

Figure 3-3. A Hypothetical Model

The following chapter will provide an overview of the methodology used

in this thesis.

76
 4. Research Methodology

This chapter portrays how the present research proceeded from a

hypothetical concept to the resulting empirical evidence. As the research
is deeply and essentially embedded in a cross-cultural context, apart from
employing the researcher’s due diligence, it was necessary to even further
sharpen the sense making process to avoid some of the traps that cross-
cultural studies conceal in all phases of the research process. Some of
these important issues and how they were addressed in this research are
discussed in this chapter.

4.1. Research Approach

Before choosing the appropriate research approach, it was crucial to firstly

determine whether the purpose of the research was of a descriptive,
explanatory, exploratory or predictive nature (Yin 1998; Snow and
Thomas 1994). This research had a twofold purpose. Firstly, to clarify
what brands really mean to consumers and what the meaning of a brand is
in conceptual terms; and secondly, to test whether brand meaning is
culturally and consumer-type determined. Therefore, two separate but
connected methodological approaches were needed. Defining the
construct brand meaning resembled descriptive theory building; while
testing the theoretically predicted influence of chosen cultural dimensions
and user types on brand meanings resembled explanatory theory testing
(Snow and Thomas 1994, p. 465).

One of the main decisions when designing research is whether it will use a
deductive approach, i.e. deduce hypotheses from existing theories and test

77
them on empirical data (Packer 1985); or an inductive approach, i.e.
induce theory from gathered empirical data by recognising patterns and
proposing relationships (Black 1999). An inductive model is generally
used in theory building, whereas a deductive model is more common in
theory testing. According to House et al. (2004), theory building is a
criterion reference approach, while theory testing is a construct oriented
approach. According to them, the main difference between the two
approaches is the point of time when the construct measured by a scale is
specified, i.e. before or after it has been measured. Langley (1999)
recognises inspiration as the third approach to research (besides induction
and deduction). Such an approach represents theory building that may be
stimulated by empirical data, general reading or intellectual exercise
(Weick 1995), but its roots are not traceable.

Given the purpose of the thesis and the nature of the research question, the
first phase of this research employed a mainly theory-free inductive
approach (Yin 1998). In turn, the second phase was predominantly
deductive, but stemed from inspiration (Langley 1999) and included some
inductive findings as well. According to Fine (1981), an iterative process
of inductive discovery and deductive testing is the most appropriate way
to generate theory and discover the truth.

4.2. Research Design

4.2.1. Research Methods

The most common practice for descriptive theory building is the use of in-
depth interviews, while explanatory theory testing calls for large sample
questionnaire surveys (Snow and Thomas 1994; Montgomery, Wernerfelt,

78
and Balakrishnan 1989). The research methods applied in this thesis and
the reasons for their choice are presented in table 4-1.

Table 4-1. Methodological Approaches

Stage Purpose Question Direction Methodological Method

approach

1 Define brand meaning What? Inductive Descriptive Interview

theory building
2 Test influence of culture Is it as Deductive Explanatory Questionnaire
on brand meaning predicted? theory testing
Source: Adapted from Snow and Thomas 1994.

Eisenhardt (1989) suggests that two-phase studies should be used to

neutralise some problems inherited by method choice and get more
accurate results. Recent studies embrace such an approach, even adding
more phases (Abou Aish, Ennew, and McKechnie 2003; Ross and
Harradine 2004). Both stages of this research were thus conducted in two
phases (figure 4-1). The first phase in each stage served to test the
measurement tool on a pilot sample, whereas the second served to apply
the fine-tuned measurement tool on the multicultural sample to reach the
targeted findings. In addition, the analysis of the output of the qualitative
stage was used in the subsequent quantitative stage of the research.

A hermeneutic research setting in the first part of the research provided

interviewees with an opportunity to reflect on their past experiences as
consumers, and to consider their personal significance and meanings of
brands (Thompson, Pollio, and Locander 1994). This produced an initial
understanding of the associations that consumers had with the chosen
brands (Coulter, Zaltman and Coulter 2001), and in turn enabled a more
accurate construction of the questionnaire for the second stage of the
research. For example, in their study on various automobile brands,
Pankhania, Lee and Hooley (2007) used such a two-stage approach to

79
“avoid placing any preconceived structure on the perceptions of each
responding group”. This allowed them to measure “the most pertinent
attributes rather than simply a list they created themselves”.

Pilot phase Main phase Output

Qualitative

In-depth interviews In-depth interviews on Brand

stage

with experts a cross-cultural panel meaning

definition

Experts’
questionnaire analysis
Quantitative
stage

Cross
cultural
Questionnaire on a Questionnaire on a differences
pilot sample cross-cultural sample in brand
meaning

Figure 4-1. A Research Sequence

A questionnaire survey is the usual method for testing proposed

hypotheses, so it was the method employed in the second stage of this
thesis as well. Geographically distant consumers were easily reached with
the Internet and various online survey tools.

There are two serious methodological challenges in terms of validity when

using a questionnaire survey. The first is a low response rate, which can
introduce bias and hinder generalisability (Snow and Thomas 1994). To
overcome this challenge and receive a sufficiently high response rate,
advice from Aaker, Kumar, and Day (2001) was followed. They proposed
a number of techniques such as reducing the length of questionnaires,
explaining the purpose of the survey to respondents, reminder notices etc.

80
The second challenge is the common method bias that may occur when
the same respondents evaluate both variables among which the
relationship is tested (Rindfleisch et al. 2008). Luckily, there are
numerous procedural and statistical techniques that can be used to control,
avoid and correct common method biases as well. According to Podsakoff
et al. (2003), it is essential to apply the following procedural remedies to
reduce measurement errors: obtaining measures of the predictor and
criterion variables from different sources, or using a temporal, proximal,
psychological, or methodological separation of measurement when the
former are not possible. In addition, they also propose protecting
respondents’ anonymity, reducing evaluation apprehension,
counterbalancing the question order and improving scale items.

In this research, the measure of cultural dimensions and brand meaning

were obtained from different sources. Cultural dimensions that acted as
predictors were obtained as secondary data from House et al. (2004),
whereas brand meanings were obtained from primary data. As for the
relationship between type of user and brand meaning, both were acquired
from the same sources, but the common method bias did not pose a major
problem because the type of user is a clearly defined interval variable and
therefore much less influenced by consumer bias. Furthermore, the
respondents’ anonymity was fully protected by using online
questionnaires sent to a large random sample of business students. The
nature of the research did not allow for questions to be shuffled between
the brands’ evaluations, but the order of brands was randomly generated
for each respondent. As for the ambiguity of the questionnaire items, tests
in the pilot phase enabled the reduction of ambiguities to a minimum.

81
4.2.2. Units of Analysis

In order to ensure the external validity of the findings, research should be

repeated on several units of analysis (Eisenhardt 1989). Therefore, this
research was based on five units of analysis that were chosen on several
criteria. Firstly, and most importantly, the chosen brands had to be truly
globally positioned. As pointed out by Steenkamp, Batra, and Alden
(2003), companies are increasingly moving towards global brand
positioning. According to their results, this is partly due to consumers
showing a greater preference for brands with a “global image” over local
brands, even when the quality and value offered by the global brand is not
objectively superior.

In this research, special attention was given to choosing brands that have
the least local market adaptation in terms of brand meaning and
positioning. Global positioning was critical because only stringent
conditioning could provide evidence that possible brand meaning
differences across cultures only stemmed from interpretation processes,
and not from positioning strategy differences. However, global positioning
should not be confused with globally standardised advertising. According
to Alden, Steenkamp, and Batra (1999) and Zhou, Teng and Poon (2008),
differences in communication translations and product adaptations based
on local needs are not equal to differences in terms of brand meaning and
positioning. Therefore, such differences did not represent a threat to this
research. To ensure that the chosen brands are globally consistent in their
brand positioning, a list of “Best Global Brands” (Interbrand and Business
Week 2008) was used. In addition, the output of the qualitative stage of
the research was investigated in detail to reveal any possible differences in
the positioning strategies of the initially chosen brands.

82
Secondly, brands endorsing different product categories were chosen. In
doing this, the goal was to choose those product categories that are widely
recognised as satisfying different consumer needs. This ensured control
for the influence of the product category on overall brand understanding,
and served to empirically challenge suggestions by Zhang and
Neelankavil (1997). Based on a cross-cultural study on individual vs.
collectivistic brand appeals, these authors claimed that choice in
advertising themes can be limited in the case of some products, such as a
toothbrush or razor that offer only personal benefits. They believed that
such products do not support collectivistic advertising motives because
they would be less successful regardless of the cultural context.

Finally, it was also considered interesting for the survey to choose one
brand that only exists in a virtual space. This was because of the
increasing importance that such brands have for consumers. In the 2006
survey on brands which had the most impact on consumers’ lives, Google
took the top spot. The video-sharing website, YouTube, came third, while
the online encyclopaedia Wikipedia came fourth. The other two spots
among the top five belonged to Apple and Starbucks (Zumpano 2007).
Zumpano (2007) himself stresses that newcomers YouTube and
Wikipedia indicate the growing impact of online brands. Additionally,
proof of the growing impact of online service brands is Google’s jump
from number 20 in Interbrand’s report in 2007 to number 10 in 2008
(Interbrand and Business Week 2008). Although ten-position jumps on the
list are not rare, it is considered a major jump for the brands situated on
top of the list, where absolute differences in brand values among positions
are huge.

In addition to these three major conditions, it was considered necessary to

think of the sample population when choosing appropriate brands for the
study. This means the respondents had to be very familiar with the brands

83
and, ideally, have personal experience with them to be sure that their
evaluations were accurate and stable (Dolnicar and Rossiter 2008). In this
research, apart from physical consumption, a cognitive experience was
also considered a mode of brand experience. More precisely, for a brand
to be chosen for this research it should have been a part of the evoked or
consideration set which Kapferer (2004, p. 17) defines as a shortlist of two
or three brands one would consider buying. It is the third level of the
consumer’s involvement with a brand; aided and unaided awareness being
the lower two levels, and consumption being the highest level.

Finally, it was considered beneficial to choose brands from different

countries of origin as that might have evoked some additional differences
in meaning. Taking all of these conditions and points into consideration,
five global brands were chosen. As table 4-2 shows, they are: Coca-Cola,
Nokia, Toyota, Google and Nike.

Table 4-2. The Studied Brands

Brand Country of origin Product category Rank a Targeted consumer need

Coca - cola United States Beverage 1 Physiological, (social)

Nokia Finland Consumer electronics 5 Social
Toyota Japan Automobile 6 Safety, (esteem)
Google United States Internet Services 10 Social, Self-actualisation,
(physiological)
Nike United Stated Sporting Goods 29 Self-actualisation (esteem)
a
based on brand value
Source: according to Best Global Brands 2008 survey (Interbrand and Business Week 2008).

The portfolio of the chosen brands was much diversified in terms of

product categories. Coca-Cola is a conventional product that consumers
invest little time and effort in during the purchasing process. Brand loyalty
in this category is the least, and the main criterion for consumer choice is
often availability. This is exactly Coca-Cola’s marketing strategy, i.e. to
be “within an arm’s reach of desire” (Riesenback and Perrey 2007, p. 31).

84
Despite the generally low consumer’ loyalty for conventional products,
Coca-Cola seems to have achieved a preference status. According to the
Brand loyalty barometer (Riesenback and Perrey 2007, p. 13), 52% of
German consumers would be loyal to Coca-Cola, whereas far fewer, i.e.
8% would be loyal to Pepsi-Cola.

Nokia and Nike belong to the retail products category in which brands are
among the most important cues in a consumer’s purchasing decision.
Nokia’s strategy focuses on tightly integrating operating businesses with
innovation, and as a result it achieves the right balance between
innovation and execution (Hickman and Raia 2002). The balanced
strategy led to Nokia’s favourable position over other brands, so that in
Germany for example, 48% of consumers consider Nokia the favourite
brand and only 14% prefer Samsung (Riesenback and Perrey 2007, p. 87).

To a great extent, Nike’s success is related to recognising the trend

towards sports shoes as fashion in the 1980s. For decades, Adidas (and
Puma) were the key players in the sports shoe industry, but both of them
failed to recognise that people had started wearing sports shoes on the
streets and during their leisure time (Peters 2008, p. 116). While they
ignored the trend, Nike used the opportunity to capture the market. By
responding to the changing consumer needs, Nike took the leading
position in the entire sports shoe market and holds it to the present day.

Toyota belongs to a category of special products which consumers tend to

buy only after having spent substantial time and thought in informing
themselves of the alternatives. Toyota’s success is based on its superior
production quality, which is further based on the total quality
management, the just-in-time delivery policy, and the six sigma business
management strategy (Henry 2007). Although it is a common business
case study in management, surprisingly it only plays a side role in the

85
branding literature despite its position as the sixth most valuable global
brand in the Interbrand and Business week list (2008).

Finally, the fifth brand explored in this study is Google. It is an online

service brand that is not purchased by end-consumers, but rather
consumed for free. For the internet users, Google is not a matter of choice.
Rather, it is considered as a taken-for-granted resource that is freely used.
From the perspective of this research, it is interesting to point out that this
brand is also not advertised to end-consumers, unlike all the other four
brands and the vast majority of the strongest global brands. In business
studies, it is often observed as a creative and lucrative business idea in the
topics related to e-commerce and online brands. For all these reasons, it is
particularly interesting to explore Google’s brand meanings and compare
it to other physical products’ brands.

The needs targeted by each of the chosen brands are also all different as is
visible in the last column of table 4-2. The needs that were addressed by
the invention of a particular brand are specified first. Then, the needs that
consumers also satisfy in the consumption process are specified in
brackets. The products’ ability to satisfy the needs in the brackets has
evolved over time, either as a result of corporate communication or
accumulated consumer experiences. In the case of Google, the
physiological need is also added in the brackets, although it is a slightly
exaggerated application of its meaning. Nevertheless, with a dose of
caution, it is not wrong to say that Google has actually grown to represent
almost a basic life necessity for young Internet users, who are the targeted
respondents of this research.

86
4.2.3. Population and Sampling

As mentioned earlier, the population for the study is comprised of

international consumers from several countries. There are numerous
sampling techniques suggested for such research, the choice of which
depends on the objective of the study (Aaker, Kumar, and Day 2001). The
objective of this study is to research the relationship among variables,
rather than generalise results to the population at large. For this reason, the
technique considered most appropriate was the matched samples
technique (Hofstede 1991; Verhage, Yavas, and Green 1991). It involved
identifying well-defined and homogeneous samples which differ in
nationality but are as similar as possible in as many other aspects, such as
age, sex, social class and rural-urban residency (cf. Orth, Koenig, and
Firbasova 2007). It was important to keep all but the studied variables
constant because, obstruction of that condition would impede the clear
relationship between the studied predictor (cultural differences) and
criterion (brand meaning). In other words, to increase the internal validity
of the studied relationship, the research was placed in an environment in
which variance in brand meaning could not have been be caused by
variables other than culture, (e.g. psychographic, demographic or other
differences among the consumers of the studied countries).

For the above reasons, business and business related4 student samples in
different countries were chosen. Across all the countries, the students are
generally aged between 18 and 25, are in the upper intellectual echelon of
their societies; have similar professional interests (business) and drivers
(education); and usually represent the middle to upper middle class of

4
These include, but are not limited to: students of graduate and undergraduate programmes
in business, management, marketing, business finance, commerce, retailing, economics and
business informatics.

87
society. Using student samples was also advantageous in terms of ease of
accessibility (via the student administration office).

4.2.4. The Studied Countries

The main criterion for choosing the surveyed countries was the fact that
they belong to different cultural clusters and different continents. An
additional issue that influenced the country choice was the language of
research. Because of the nature of the study, a huge dilemma was the
question of whether to administer the questionnaire in English or in the
respective native languages of the respondents’ countries. In cross-cultural
studies, it is common practice to translate the questionnaire to the native
language through the translate - back translate method. This method
reduces ambiguities as one bilingual speaker translates items from the
original language to another language, and then another native speaker
back-translates them to the original language (Craig and Douglas 2000).
However, despite the care and precision employed in the translation
process, some nuances in meanings may still be lost. For a study on
meanings, this might have caused a great problem. On the other hand, the
respondents’ non-understanding of nuances captured in foreign language
terms might have caused an even bigger problem (Dolnicar and Rossiter
2008). After extensive analysis and discussions with experts, it was
believed to be slightly more favourable to administer the questionnaire in
English in all the countries, but under the condition that the average
student’s English was good. According to the GLOBE study (House et al.
2004), the world is divided into ten cultural clusters. The United States
and India are the countries in which English is either native or official, so
they were the obvious choices for Southern Asian and Anglo clusters. As
for the European representatives, the United Kingdom was considered, but
being an Anglo-Saxon country it was evaluated as being too close to the

88
United States in cultural terms. Finally, Austria was chosen from the
Germanic European cluster and Finland was chosen from the Nordic
European cluster.

4.3. Variable Operationalisation

A construct is a conceptual term that describes a phenomenon of

theoretical interest, whereas its measure (often scale) is an observed score
(Edwards and Bagozzi 2000). The constructs of interest in this research
are Brand meaning, Culture and Type of user. Their measurements are
explained in the following subchapters.

4.3.1. The Construct of Culture

The construct of culture was measured by cultural dimensions. Secondary

data were used for this variable. The study by House et al. (2004) is the
most recent social study on cross-cultural differences and thus
incorporates previously accumulated knowledge on the topic, such as
findings by Hofstede (2002), Trompenaars (1994) and others. Table 4-3
presents the studied countries and their scores along the cultural
dimensions by House et al. (2004).

The table presents the scores for the practices and values for all the
studied countries. It was important to be careful when choosing the right
manifestation to be studied. Because consumers’ understanding of brand
meaning is an output of values and beliefs (Dilts and DeLozier 2000),
values are considered the more appropriate cultural manifestations for this
research.

89
Table 4-3. The Studied Countries’ Cultural Dimensions
Austria Finland India United States
Cultural dimension
practice value practice value practice value practice value
Performance orientation 4.44 6.10 3.81 6.11 4.25 6.05 4.49 6.14
Future orientation 4.46 5.11 4.24 5.07 4.19 5.60 4.15 5.31
Gender egalitarianism 3.09 4.83 3.35 4.24 2.90 4.51 3.34 5.06
Assertiveness 4.62 2.81 3.81 3.68 3.73 4.76 4.55 4.32
In-group collectivism 4.85 5.27 4.07 5.42 5.92 5.32 4.25 5.77
Institutional collectivism 4.30 4.73 4.63 4.11 4.38 4.71 4.20 4.17
Power distance 4.95 2.44 4.89 2.19 5.47 2.64 4.88 2.85
Human orientation 3.72 5.76 3.96 5.81 4.57 5.28 4.17 5.53
Uncertainty avoidance 5.16 3.66 5.02 3.85 4.15 4.73 4.15 4.00
Source: House et al. 2004.

4.3.2. The Construct of Brand Meaning

Consumers use brand associations to help process, organise, and retrieve

information in memory and to aid them in making purchase decisions
(Aaker 1991, p. 109-113). Although brand associations are not equal to
brand meanings in a conceptual sense, those techniques that are normally
used for association elicitation should be also used to evoke brand
meanings. Brand associations can be elicited by a variety of data
elicitation techniques, including free association and response (Krishnan
1996), laddering procedure (Reynolds and Gutman 1988), hierarchical
categorisation (Oakenfull et al. 2000), pair-wise similarity judgments
(Hauser and Koppelman 1979), Kelly’s repertory grid (Henderson 1998,
Henderson, Iacobucci, and Calder 2002), negotiation process (Nairn,
Griffin, and Wicks 2008) or ZMET (Coulter and Zaltman 1994).

The Zaltman Metaphor Elicitation Technique - ZMET (Coulter and

Zaltman 1994) is a multi step method that involves semi-structured, in-
depth personal interviews centred on visual images. According to this
method, each interviewee is asked to collect and submit images to the
interviewer that indicate what the brand means or does not mean to him or

90
her. The interviewee is then asked to describe those images and images he
or she could not have found, but wanted to submit. Then, the images are
categorised by the interviewee and each category is given a name, still
keeping in mind that the images represent the studied brand. Finally,
according to the ZMET procedure, the interviewee is asked to elicit those
sensory images he or she associates with the studied brand and those they
cannot associate with it. By doing this, some additional associations come
to mind that could not have been captured in the images. Two more final
steps: the interviewee creates a map or a causal model using the constructs
that have been elicited; and, with a technician’s assistance, creates a
summary image using digital imaging techniques.

Alber, Merunka, and Valette-Florence (2007) used a similar methodology

in which subjects select one image among 19 offered for each brand. The
chosen image symbolises their relationship with the brand. However,
Coulter, Zaltman and Coulter (2001) emphasise that ZMET method’s
strength lies in the fact that respondents, not researchers, supply the
stimuli (i.e., visual images such as pictures or photographs) for the
subsequent in-depth interviews. Being proactive in this task, the
respondents are able to better articulate their thoughts and feelings, and
identify issues that are important to them and potentially unknown to the
researcher.

Another very relevant path of research discovers meaning of a brand by

studying the product categories that brands could extend to. For example,
in their study on Rolex and Timex watch brands, Park, Milberg and
Lawson (1991) discovered that a brand could be extended to product
categories that offer the same functional or symbolic benefits, or both.
According to their results, the likelihood of the success of the extension
depends on whether the core association of the brand is functional or
symbolic. Similarly, Oakenfull et al. (2000) discovered that the Pennzoil

91
brand may not fit with clothing products in general, but Pennzoil work
clothes may be well received because tough remains a central attribute of
the brand.

To discover the meaning of the chosen brands in this study, the qualitative
phase of the research was, to a great extent, guided by the ZMET research
procedure. However, its final two steps were not included in this research
as they would significantly increase the complexity of the research (with
globally dispersed interviewees), and not have a significant effect on the
results.

The introduction of a new concept requires ensuring its validity in terms

of the construct itself and its operationalised measurement (Trochim and
Donelly 2007). The construct shows convergent validity if it is related to
other constructs with whom it has a theoretical connection. At the same
time, it shows divergent validity if it is not related to constructs with
whom it does not have a theoretical connection. Since brand meaning is
not defined a priori, but rather is a posteriori results-driven construct, its
correlations to other concepts are also not predetermined. Nevertheless, to
investigate whether brand meaning is a construct of its own value, an
analysis of similarities and differences with other similar concepts was
conducted.

Construct validity of the measurement implies that the measurement really

measures the construct (Denzin and Lincoln 1994). The face validity of
the measurement tool was examined by a panel of experts as suggested by
Trochim and Donelly (2007), while the convergent and divergent
validities were measured in a series of reliability and correlation tests on
the results of a second stage pilot study. Aspects of the measurement that
did not belong to a particular construct were excluded from the main
phase study, as is described in detail later on in the thesis.

92
4.3.3. The Construct of Type of User

The last construct studied in this research was the brand usage. It was
initially meant to be measured on a scale based on the frequency of
consumption, but eventually, based on common research practice, it was
turned into a dichotomous variable. Nevertheless, a five-point Likert-type
scale ranging from ‘I love the brand’ to ‘I hate the brand’ was added to
identify various types of users, namely: Engaged users, Disengaged users,
Engaged non-users, and Disengaged non-users.

93
 5. Empirical Research on Brand Meaning

This chapter of the thesis presents the results of the qualitative empirical
research. The aim of this part of the research was to identify what brands
mean to consumers. As explained in the previous chapter, firstly the pilot
research and then the main research are elaborated.

5.1. Pilot Research

In order to enable testing of the reliability of the findings, firstly the

research procedure is transparently described (Yin 1994). In that way,
subsequent researchers can repeat the research and arrive at the same
conclusions (Denzin and Lincoln 1994). Subsequently, the results will be
presented.

5.1.1. Research Procedure

Extensive interviews were carried out with five business research experts
in the pilot phase of the qualitative research. The experts’ comments
served to develop the correct structure for the interview, to fine-tune the
interview questions, and also to make the duration of the interview
acceptable.

In addition, the interviews served to test the online messenger service,

Skype, as a possible communication channel for the main phase research.
Two interviews were conducted via the online messenger service, Skype,
one via a series of e-mails, and two in person. The e-mail correspondence
was severely hindered and gave meagre results. Conversely, no difference

94
in response quality between the Skype communication and in-person
communication was discovered.

5.1. 2. Results

Each of the questions posed in the interviews with experts produced

different responses. For example, the images that indicated what a brand
means to consumers were all positive. This is in accordance with a
definition of high-equity brands as those for which consumers possess
substantial knowledge structures that often include readily accessible and
positive associations (Aaker 1996, p. 8; Roehm and Brady 2007). Images
that indicated what a brand does not mean to consumers often just had the
opposite meaning to the positive ones. Nevertheless, in some cases the
negative images gave some new meanings. Thus, both of the questions
(i.e. the positive and negative images) were retained in the main phase of
the qualitative research.

It was also discovered that the respondents had problems understanding

the particular question which asked for the sensory images they associate
with a brand. They had even bigger problems understanding the question
which asked for the sensory images they do not associate to a brand.
Consequently, exclusion of these questions was considered, but because
they evoked associations that would otherwise not be mentioned by
consumers, the positive version of the question was kept for the later stage
of the research.

Finally, questions which asked what fruit or animal the brand would be
were it a fruit or an animal were left out of the main phase research,
because the responses were irrelevant.

95
5.2. Main Phase Research

5.2.1. Research Procedure

This subchapter offers a detail description of data gathering and data

analysis procedures employed in the main phase of the qualitative
research.

5.2.1.1. Data gathering

The main phase of the qualitative stage of the research was conducted by
means of an in-depth interview with 19 interviewees (four from the United
States and five from Austria, Finland and India each)5. Three criteria were
defined for the interview candidates. They had to be younger than 33,
business-related graduates and not highly international. The last condition
was reworded to state that the respondents were supposed to have been
born in the studied country, and not to have lived abroad longer than six
months in the last 5 years or longer than a year ever. Finally, the interview
sample was also controlled for sexes so that both male and female
respondents were interviewed in all four countries.

These conditions resembled the conditions set for the quantitative

research, with one difference. The interviewees had to be business
graduates, whereas the questionnaire respondents were still students. It
was considered important to choose business graduates for the interviews
as the interviewees then acted not only as consumers, but also as experts
in the field during the interviews. In that way, they helped the researcher
to grasp culture-specific issues that might have occurred in the later stage
of the research.

5
In the validation studies of the ZMET procedure, Coulter, Zaltman, and Coulter (2001)
discovered that four to five focused in-depth interviews can provide up to 90% of the
information available from a larger set of interviews.

96
The interviews were conducted in the period between January and October
2008 and lasted between 60 and 120 minutes. The difference in duration
mainly occurred due to the difference in typing speed and less because of
the amount of information exchanged.

The structured interview consisted of seven main questions. Firstly, the

participants were asked to collect e-images that indicated what Coca-Cola,
Nike, Google, Nokia and Toyota mean to them, and e-images that
indicated what those brands do not mean to them. The interviewees were
not given any further explanation before the interviews. Most of the
candidates submitted one image for what each brand means to them and
one for what it does not mean. Another four questions were designed to
provoke further associations that the interviewees had with the brands.
Firstly, they were asked to explain why the submitted pictures indicated
the brands’ meanings to them; secondly, which adjectives they relate and
do not relate to the brands, thirdly which sensory images they associate
with the brands, and fourthly what the brand means for them in one word.
In addition, the respondents were asked for their feelings towards the
brand as well as the usage frequency.

5.2.1.2. Data Analysis Procedure

To analyse qualitative data gathered in the interviews, the hermeneutic

method was applied. Hermeneutics is the description and analysis of
meaningful human phenomena which seeks to highlight the often
unspoken background of socially shared meanings by which a person
interprets personal experiences (Thompson, Pollio, and Locander 1994). It
was originally a set of techniques for interpreting written texts, but then it
spread to interpreting images and also other types of communication and

97
human conduct (Packer 1985)6. Martin Heidegger later turned it into a
philosophy, so that it was no longer perceived as a methodological or
didactic aid for other disciplines, nor as a technique for understanding
linguistic communication. It became ontology, the study of the nature of
being, existence or reality in general (Ramber and Gjesdal 2009).

In the present research, however, hermeneutics was used as a theory of

interpretation. The interview transcripts were analysed to identify the most
common meanings of the chosen brands so that those could be used in the
closed-questions in the subsequent quantitative stage. Data from the
interviews also served to discover possible misunderstandings that were
corrected before quantitative research was administered. This was crucial
because mistakes in the quantitative data would have been very difficult or
even impossible to correct (Cavusgil and Das 1997).

The 150 transcript pages gathered during the interviews were screened for
off-topic comments and condensed to 91 pages of topic-related comments
(mainly 4-5 pages per interview). The interviewees’ responses were then
sorted in a three-step coding procedure taken from Lee and Rhee (2008).
First, meaning units composed of words, phrases and sentences were
extracted from the transcripts. Then, components of meanings were
identified from the meaning units and listed. Finally, the components were
sorted into groups of meanings. Similarly, Henderson et al. (2003) studied
meanings that certain brand logos evoked. For 195 logos they received
5,600 associations. They grouped similar associations in order to check
whether the logo evoked a clear meaning within a culture. Such a
procedure is in line with the basic guidelines of qualitative content

6
Initially hermeneutics was developed for the examination of biblical texts, carried out to
uncover and reconstruct the message from God that was believed the texts contained but hid.
The term itself refers to Hermes, messenger of the Greek gods, and also God of eloquence
and cunning as well as of roads and theft. Later, Hermeneutics became generalised to a
method of textual interpretation not restricted to religious works (Packer 1985).

98
analysis and the grounded theory approach. It requires data collection and
theory building to be performed simultaneously (Langley 1999).

The most frequently mentioned meanings (i.e. those mentioned by at least

three interviewees) were taken for further analysis and as inputs for the
later quantitative stage. It was important for all the respondents of the
quantitative stage (despite their country) to evaluate the items identified in
any of the four studied countries. This is because most differences in
meanings were expected to be found in the evaluation of the meanings
mentioned by only one country's respondents. Such meanings would not
have been included in the questionnaire if qualitative research had not
been conducted in all of the four countries.

5.2.3. Results

Altogether about 200 images (Appendix 1) and 150 pages of transcript

were collected. For Coca-Cola, most of the interviewees submitted items
depicting the Coca-Cola logo, Christmas associations, sugar and ice. In
the case of Google, most of the images frequently depicted the Google
logo, followed by the symbols representing ideas and searching activity.
The vast majority of Nike’s images represented sport or activity in
general. Nokia’s and Toyota’s images were dominated by phones and cars
respectively.

General analysis of the images revealed that interviewees offered two

main types of meanings represented by two types of images. Those in
which more direct meanings were depicted, i.e. either a product or a logo;
and those in which less direct meanings were depicted i.e. symbolic or
experiential associations. Graph 5-1 presents per country percentage of
images that were closely related to the brand as opposed to those more
figurative and symbolic.

99
Graph 5-1. The Percentages of Symbolic Images
100

80
The Percentages of Symbolic

60
Images

40

20

0
Coca-Cola Google Nike Nokia T oyota
Brand
Finland India US Austria

As the graph shows, across the brands most of the Indian respondents
submitted images that depicted neither the logo nor the name of the brand.
This is in line with the India’s high-context culture (Hall 1989). In such
cultures people create close connections over a long period of time and
many details are not made explicit because most members share a broad
common knowledge base. In such cultures, things carry many situational
meanings. On the contrary, in the low-context cultures, like the United
States and Austria, people tend to have connections of shorter duration or
for a specific reason. In these societies, most things need to be spelled out
explicitly. Therefore, the Austrian and the American images, as
anticipated, were more closely related to the essence of a particular brand.

Interestingly, being a Finnish brand, Nokia evoked very deep and varied
meanings for the Finnish respondents, none of whom submitted an image
of Nokia’s logo or a cell phone. On the other hand, Nokia occupies a very
marginal role in the United States market, so all of the American images
depicted a non-symbolic image of a cell phone. A more detailed

100
explanation of the images as well as the other meanings associated to the
studied brands follows in a separate subchapter for each brand.

5.2.3.1. The Meanings of Coca-Cola

From 93 groups of meanings for Coca-Cola, as graph 5-2 shows, Coca-

Cola was most often associated to refreshing (represented by the
manifested meanings of refreshing, cooling, fresh and freshness). Then
followed not healthy (with the manifested variables not healthy, no
nutrition value, not safe, harmful and unhealthy), cold (with the
manifested variables cold, cool, chilled, chilling, not hot), Christmas (with
the manifested variables Christmas, Christmas song, Santa Claus and
Christmas trucks) and sweet (with the manifested variables sweet, sugar
and not soar). Tiia7 from Finland summarised it all to:

Traditional…original…refreshing…dr. pepper…Santa Claus…warm summer

…refreshing drink8

Although most of the responses were very positive, a few respondents had
serious doubts about Coca-Cola, primarily for its non-natural ingredients
and obesity connotation. One of the most interesting answers in that
respect was given by Praveen. He said:

It's probably not as harmful as smoking, but somewhere in the vicinity. It's
about as addictive as smoking… This secret ingredient business, I'm not
happy with; caffeine carbonate we know, but for all I know they may be
sneaking baby hair into it, and we wouldn’t know… In college we tried
creating Coke with coffee, soda and sugar. It stunk and didn’t taste anything
like it. Why is it that every FMCG product has to have its ingredients written
on it, except Coke? And, of course the Cocaine origins of Coke are still
disturbing.

7
Names of all interviewees are changed to guarantee anonymity to respondents.
8
Responses are presented in their original form in terms of style (i.e. jargon and dialect), but
corrected for spelling mistakes.

101
Graph 5-2. The Meanings of Coca-Cola
Refreshing/cooling/fresh/freshness
Not healthy*
Cold/cool/chilled/chilling/not hot
Christmas/Santa Claus/Christmas
song/Christmas truck
Sweet/sugar/not soar *Not healthy/no
Global/not local nutrition value/not safe
/harmful/ unhealthy
Warm summer/hot day
American/American lifestyle **Fizzy/carbonated/aerated
Fizzy** /carbonation on tongue/
sound of bubbles/
Energetic***
sparkling
Big/large/huge
Exciting**** ***Energetic/drink it when
Recognisable/familiar/not tired/no sleeping/ keeps me
unknown/identifiable going/keeps me working
Opening bottle
Traditional/classic/not new/original *** Exciting/not boring /not
Bonding with friends/parties/fun serious/not uncool /not sad

Not natural
0 2 4 6 8 10 12 14
Finland India US Austria

As the graph also shows, in different countries, different meanings were

mentioned the most. Although refreshing was a very strong association in
Finland, India, and Austria, it was much weaker in the United States.
Similarly, the Austrians were very preoccupied with Coca-Cola not being
healthy and not being natural, while the other nations and especially the
Finnish respondents did not seem to be worried about it. For example,
Michaela from Austria expressed her feeling of guilt when consuming
Coca-Cola in the following comment:

I always drink Coke when I am tired and my body needs some sugar ;) and
whenever I drink Cola I feel a little bad because I know that I am consuming
a lot of sugar.

Christmas was one of the most often mentioned meanings for Coca-Cola,
which had even more significance due to the fact that the qualitative

102
research was conducted between January and October and not in the
Christmas season when Coca-Cola heavily advertises the Christmas
atmosphere. Naturally, the Indians, being mostly Hindus, did not associate
Coca-Cola with Christmas. Finally, the Indians mostly associated Coca-
Cola to bonding with friends, while the other nations did not report such
an association. A better overview of the primary meanings per country can
be seen in table 5-1.

Table 5-1. The Primary Meanings for Coca-Cola per Country a

Country Meaning 1 Meaning 2 Meaning 3 Meaning 4
Austria Not healthy Refreshing Christmas
Finland Refreshing Not healthy Christmas Global
India Refreshing Cold Bonding with friends
United States Cold Sweet
a
Only the first meaning is mentioned – it represents the whole group of meanings

5.2.3.2. The Meanings of Google

In the analysis of the responses for Google, among 59 groups of meanings,

the following were the most often mentioned: gives answers (with the
manifested meanings gives answers, provides data, all knowing,
informative, source of knowledge, key to everything, encyclopaedia,
Wikipedia, information) and fast (with the manifested variables fast, quick,
speed, not slow, efficient, timely, at any time) as graph 5-3 also shows. As
for the individual answers, Anisha from India sent an image of a smiley
face, and in the accompanying comment she explained that the image
indicated she was satisfied and further added:

Ne [eng. any] problem. ne issue. ne damn thing in the world; Google is there
for me, n [eng. and] it always keeps me happy.

Google, like Coca-cola, also evoked some country-specific associations.

For example, while the Finnish respondents considered it a life saver and
generally had very positive opinions towards it, almost all of the Austrian

103
respondents considered it to be a threat to their privacy. For example,
Peter said:

They have so many tools, too many, they are collecting too many personal
data… they are a little bit dangerous.

Graph 5-3. The Meanings of Google

Gives answers*
Fast/quick/high speed/not slow/efficient /timely/at any
time
Not old**
World/worldwide/global
Powerful***
Life saver/basic necessity/very important *Gives answers/provides data/ all
No privacy/no freedom/no independence/ knowing/ informative/source of
omnipresent/taking over our lives information/key to everything
Useful/helpful /encyclopedia/wikipedia/
Google logo information/knowledge

Happiness/satisfaction
**Not old/new/not antique/not
Huge/extensive/not small historic/not old fashioned/ not
Not reliable out of date/up to date/young
/innovative
Easy to use/convenient/no problem
Search ***Powerful/potent/almighty/t
Not stupid/not dumb/not ignorant/smart/savvy ook earth control/dominant/
multi billion industry/not poor
Reliable/trustful
Creative

0 2 4 6 8 10 12 14 16
Finland India US Austria

Interestingly, the respondents from the United States did not associate
Google with world and globe at all, and although some of them mentioned
gives answers, it was not one of the main associations. Finally, the feeling
of reliability seemed very bipolar within all the countries, so some of the
respondents mentioned its reliability, whereas the others mentioned its
unreliability. An explanation in that respect was given by Hanna from
Finland:

[Google] has changed the way kids see the world… Bart Simpson [cf.
appendix 1] comic shows the attitude that they have, "always trust Google"

104
 which is not necessarily a good thing but that's what people think these days,
"don't ask me, Google it"…kids just want simple answers from Google, ready
made answers through one keyword…[but] there is still a world or people
who you can't find with Google.

Table 5-2 gives an overview of the primary meanings of Google in each

studied country.

Table 5-2. The Primary Meanings for Google per Country a

Country Meaning 1 Meaning 2 Meaning 3 Meaning 4
Austria Gives answers Not old Fast No privacy
Finland Gives answers Fast World Life saver
India Gives answers Fast
United States Fast Powerful
a
Only the first meaning is mentioned – it represents the whole group of meanings

5.2.3.3. The Meanings of Nike

Nike is the brand to which 84 groups of meanings were elicited. Across all
the countries, the most strongly associated meaning, as graph 5-4 presents,
was by far sports (with the manifested variables sport, sportiness, sporty).
The next most often mentioned meanings were achiever (with the
manifested variables achiever, winner, victory, winning, not failure, not
losing), not formal (with the manifested variables not formal, casual and
be oneself), and sweat – all reported by much fewer, i.e. six respondents.
Most of these meanings were summarised in a response by Minna from
Finland:

[Image that indicate what Nike mean to me is] a Finnish tennis player who
works hard to get where he is today and yet his work is not over yet… same
as people, who have worked their bodies, and trained their skills to be the
best in their athletics…they have given up a few things which, for example, I
haven’t. It’s a way of life…they keep themselves in good shape and extremely
healthy… for me those would be sacrifices, but I’m not sure if they are for
them.

105
Graph 5-4. The Meanings of Nike

Sports/ sportiness/ sporty

Achiever/winner/victory/winning/not
failure/not losing
Not formal/casual/be myself
Sweat
Athletes/athletic
*Sweat shops/child labor/not
Cool/trendy
fairness/ exploitativeness/not
Expensive/pricey/not cheap socially responsible/not
Sweat shops* sustainable
Good quality/not low quality/durable
Activity**
Sport shoes **Activity/not lazy/active/not
Crowd cheering being immobile/not being
Cushion for feet*** still/movement
Hi-tech/innovative/innovation
Jogging/running
*** Cushion for feet/walking
Michael Jordan
on clouds/ soft-footed/air
Powerful
technology
Rough/not weepy/not weak
Self-assured/just do it/positive
thinking
Swoosh

0 2 4 6 8 10 12 14 16
Finland India US Austria

Nike’s most visible and most advertised symbols are its logo, Swoosh, its
tagline, Just do it, and its endorser, Michael Jordan. Swoosh is also one of
the most famous logos in the world and a rare example of a non-
personified logo with a name9. The “Just do it” marketing campaign
turned into the second best slogan of the century (AdAge 2008).
Moreover, in the branding literature, one of the most often cited examples
of a celebrity endorsing a brand is Michael Jordan endorsing Nike (Aaker
and Joachimsthaler 2002, p. 78; Keller 2008). It was thus surprising that
Robert from the United States was the only respondent to highlight the
importance of any of the three symbols by a comment:

9
Ironically, swoosh was invented just a few days before the launch of Nike and for as little
as 35 USD (Keller 2008, p. 127).

106
 Michael Jordan is certainly the most iconic athlete ever in the US. A large
part of his identity stems from his Nike commercials. When anybody says
Nike, the first thing people think about is Jordan. Nike symbolizes that perfect
athlete.

Country-wise, hi-tech is a group of meanings that only the American and

Indian respondents reported. Interestingly, Nike was not considered
expensive in India, whereas it was in the United States, although in terms
of purchasing power parity the prices were similar. The most frequently
mentioned meanings by country are presented in table 5-3. Unlike the
other studied brands, they demonstrate a very stable pattern.

Table 5-3. The Primary Meanings for Nike per Country a

Country Meaning 1 Meaning 2 Meaning 3
Austria Sports Sweat Sweat shops
Finland Sports
India Sports
United States Sports Expensive
a
Only the first meaning is mentioned – it represents the whole group of meanings

5.2.3.4. The Meanings of Nokia

When it comes to Nokia, unlike Nike, the meanings were more dispersed.
From a total of 84 groups of meanings, the meanings reliable (with the
manifested variables reliable, will not ditch you, feeling of security, handy
and there for you), hi-tech (with the manifested variables hi-tech,
technology and engineered) and safe choice (with the manifested variables
good, good investment, quality and safe choice) were the most frequently
mentioned as presented in graph 5-5. Himanshu from India, for example,
expressed his very positive feelings towards Nokia and explained why he
considered it reliable with this comment:

Wooooooowwwwwww, very positive feelings…Nokia always indicates in

India battery back up. Nokia means good battery that’s all.

107
Graph 5-5. The Meanings of Nokia
Reliable*
Hi-tech/technology/engineered
Good/good investment/quality/safe choice
Connecting people
Finnish
Easy to use /user friendly/functional/not complicated
Sturdy/strong/tough/hard
Communication *Reliable/will not ditch
Smooth keypads/sleek/clean design you/feeling of security/
Friends/family handy/ there for you
Not a camera/just phone/phone/not diverse
Modern/trendy/not old fashioned/not old
** Not cutting edge/
First cell phone
Innovative not up to date/not
Not cutting edge** innovator/not state of
Global player/not local the art
Not exciting/not cool/not exotic
Nokia original ring tone
Durable/break-proof/long lasting
0 1 2 3 4 5 6 7 8 9
Finland India US Austria

Another interesting observation is that although most of the respondents

strongly associated Nokia with their slogan connecting people, an Indian
respondent viewed Nokia’s role in human relations quite differently. He
said that:

Mobiles have turned us into people in our own bubbles. I love cell phones, but
I prefer f2f conversations... e.g. in a long journey on a train we have everyone
on their cells, in yearning for communication we become anti social.

Similarly a Finnish respondent admitted that Nokia represented:

…busy, hectic life. Few nights ago I dreamt losing my friend’s baby, because
I had to talk on the phone. I left the child in the store, and went out to talk. My
friend would never forgive me.

Some of the elicited meanings for Nokia (like in the case for Coca-Cola
and Google) are also country-specific. These are: sturdiness and durability
in India, innovation in Austria, and ordinariness and Finnish origin in

108
Finland. It is worth mentioning that each country associated different
primary meanings to Nokia as depicted in table 5-4.

Table 5-4. The Primary Meanings for Nokia per Country a

Country Meaning 1 Meaning 2 Meaning 3
Austria Innovative Reliable Communication
Finland Finnish Not exciting
India Sturdy Reliable Good
United States Hi-tech Smooth keypads
a
Only the first meaning is mentioned – it represents the whole group of meanings

5.2.3.4. The Meanings of Toyota

The last brand studied in the qualitative phase was Toyota. Out of 82
groups of meanings, similarly to Nokia, the reliability meaning (with the
manifested variables reliable, no surprises, dependable, not unreliable,
not undependable, long lasting) was evoked the most often. For example,
for Michaela from Austria, Toyota meant:

quality - my parents had once a Toyota and I learnt driving with it...it was a
very good car. It never broke down. It was easy and nice to drive... or high
quality ... It’s of high quality and I personally see no much difference to a
BMW...: maybe design...but we were very happy with the car and I loved
driving it ;)

The meaning of reliability was followed by the meaning average (with the
manifested variables average, good looking, good, practical, grey, safe
choice and rational) and several other meanings (as shown in graph 5-6).
An interesting comment relating to Toyota’s ordinariness was given by
Hanna from Finland:

I guess it is a typical, safe choice and many who don't want a good looking
nice car settle with Toyota. I might sound like a snob but after driving a
BMW, I wouldn't get a Toyota… there was a saying [in Finland] "take Toyota
if you can't get anything better”. It is statistically the most sold car in our
country.

109
Graph 5-6. The Meanings of Toyota
Reliable*
Average/ok/good/practical/grey/safe
choice/rational
Safe *Reliable/no surprises
Not luxury** /dependable/not unreliable/ not
Innovative/hybrid car undependable/long lasting

Japanese**** **Not luxury/not premium /not

Comfort/comfortable/relaxing exclusive/not high class/not
status symbol/not the highest
Expensive/not cheap/overpriced
brand value
Not exciting***
Family brand/children/family car ***Not exciting/not
adventurous/not passion/
High quality
boring/not extravagant/not
Not American/imported (for US) sexy/no emotions
Value for money/worth the money
Class/luxury/high-end/regal/grace ****Japanese hard working
Not for young/not for the next engineering team/JIT/
generation robots/Japanese car
Not expensive/quite cheap
0 1 2 3 4 5 6 7 8 9 10
Finland India US Austria

Reliability, as above mentioned, was the meaning most frequently

associated with Toyota; however, not at all in India. On the other hand, the
Indians associated Toyota with class, grace and luxury, which the other
countries’ respondents did not. Together with some Austrians, the Indians
also associated Toyota with elderly people and comfort.

Finally, unlike the Indians, the Austrians found Toyota to be a non-

luxurious car. Similarly, while the Indians found it expensive, the
Austrians found it affordable. When speaking of luxury and Toyota cars’
prices, it is also interesting that although the Indians considered it
luxurious, they did not ascribe it a high brand value which could
counterbalance its high prices. For example, Praveen from India said that
Toyota cars were:

…fuel guzzlers, and priced beyond my reach, and neither do they have a
brand value that I would go out of my way to acquire.

110
The Finnish did not find it exciting, but rather average. Some Finns even
mentioned it was a non-reliable car and a car not to show off with. Minna
gave the most picturesque explanation of how sceptical she was towards
Toyota:

I`m a car person… meaning REAL cars which are safe and to me Toyota is
not a car. It`s pile of rust. I don`t think it’s safe.

In conclusion, it could be said that Toyota has a very different overall

meaning across the four studied countries. In India it is a high class brand;
in the United States, a good quality, well-respected car; in Austria, a good,
safe choice, positive car; whereas in Finland, it is a reliable, but not highly
appreciated car. To summarise, when the Indian responses are put aside,
all the other three countries’ respondents disclose the same rationale, but
different feelings attached to it (as shown in table 5-5).

Table 5-5. The Primary Meanings for Toyota per Country a

Country Meaning 1 Meaning 2 Meaning 3
Austria Reliable Not luxury
Finland Reliable Average Not exciting
India Comfort Expensive Class
United States Reliable High quality
a
Only the first meaning is mentioned – it represents the whole group of meanings

5.2.4. Discussion

After having presented the results of the qualitative empirical stage and
provided initial comments, a deeper overall interpretation of the results is
needed. It is provided in the two following subchapters. The first one
serves to define the concept of brand meaning. The second one serves to
interpret the results in terms of whether the initially chosen brands are
truly global. More precisely, the differences in meanings of the chosen
brands were examined to decide whether they resulted from differences in
interpretation of the received message on the part of the consumers, or

111
rather from differences in positioning on the part of the companies owning
the brands.

5.2.4.1. Defining Brand Meaning

It is firstly necessary to classify the responses in terms of whether they

were elicited by a direct question on meaning or by other questions in the
interviews. This is because in the quantitative stage only those elicited in a
direct question were used. As tables 5-6 to 5-10 show, most of the evoked
meanings were mentioned in response to a direct-meaning question, i.e. a
question that investigates why submitted image(s) indicated what the
brand means to a respondent.

In the case of Coca-Cola, only when asking for what Coca-Cola does not
mean to the respondents were the items not healthy and not natural
evoked. The question on sensory images evoked associations of
carbonation and sound of bottle opening, whereas the question on
adjectives associated to Coca-Cola evoked the adjectives big, and huge. In
the case of Google, the question on what the meaning of Google was not
evoked only negative versions of the already mentioned positive ones
(such as, not old, not slow etc.). The question on adjectives further
generated the meanings of helpful, huge, reliable, not reliable, smart and
creative. Similarly to Google, Nike’s negative associations did not evoke
any new meanings. The adjectives question evoked rough, good quality,
expensive, cool and casual, whereas the sensory images question evoked
the meanings of sweat, crowd cheering and cushion for feet.

To reduce the number of meanings that were included in the questionnaire

in the quantitative stage, any redundant meanings evoked in the qualitative
stage of the research were excluded. There were two criteria for exclusion
of the items. Firstly, the number of items in groups of meanings - so that
some items were deleted from groups of meanings abundant in items.

112
Secondly, association stimuli - so that items elicited in a non-direct-
meaning questions were excluded, unless related to items elicited in a
direct-meaning question. Finally, in an attempt to rationalise the length of
the questionnaire even further, the brands’ logos were also excluded from
the quantitative research as their emphasis was predicted not to vary
across cultures. All the excluded items are indicated in tables 5-6 to 5-8,
together with the reasons for their exclusion.

In order to determine what brand meanings are and how they relate to the
other constructs of brand management, i.e. brand beliefs, brand benefits,
brand personality, brand attitudes and similar, the meanings were
compared to each of the constructs identified in table 2-1.

Analysing the answers in terms of their relationship to brand benefits

revealed that among the evoked key meanings for each of the brands, all
types of benefits, i.e. functional, symbolic and experimental benefits
(Park, Jaworski, and MacInnis, 1986) could be identified. However, not
all the responses could be equated to benefits. This is primarily because a
number of negative meanings were identified. Hence, it is concluded that
although brand benefits are rather similar and related to brand meanings,
these two are not the same construct, as brand meaning includes some
extra associations that cannot be defined as brand benefits.

Brand attributes are defined as features that characterise the product

(Keller 2003b), while brand personalities as a subset of brand attributes
related to human characteristics associated with a brand (Aaker 1997).
Among the studied brands, on average two thirds of meanings could be
described as brand attributes (either tangible or intangible), and even
fewer as brand personalities. Hence, the conclusion is reached that these
two brand concepts are also different from the concept of brand meaning.

113
Table 5-6. The Types of Responses for Coca-Cola

Brand
Attribute/
Meaning Stimulus Type of benefit meaning
Personality
category

Refreshing/cooling/fresh/freshness Meaning Functional/Experiential Attribute Product

- related
Not healthy/no nutrition value Negative - Attribute Product
a
/not safe a/harmful/unhealthy meaning related
Cold/cool/chilled a/chilling a/not Meaning Experiential/Functional Attribute Product
hot a related
Christmas/Santa Claus/Christmas Meaning Symbolic - Situational
song/Christmas truck a
Sweet/sugar/not soar b Meaning Experiential/Functional Attribute Product
related
Global/not local Meaning Symbolic Attribute Provenance
Warm summer/hot day Meaning Experiential - Situational
American/American lifestyle Meaning Symbolic - Provenance
b b
Fizzy /carbonated/aerated Sensory Experiential Attribute Product
/carbonation on tongue b/sound of image related
bubbles b/sparkling

Energetic/drink it when tired/no Meaning Functional/Experiential - Situational

sleeping/keeps me going/keeps me
working a

Big b /large b /huge Adjective Symbolic Attribute Provenance

Recognisable/familiar/not Meaning Symbolic Attribute Product
unknown/identifiable related
Provenance
Sound of opening a bottle b Sensory Experiential Attribute Product
image related
Traditional/classic/not new a Meaning Symbolic Attribute Provenance
/original
Bonding with friends/parties/fun Meaning Experiential /Symbolic - Situational
Exciting/not boring/not Meaning Symbolic/ Experiential Personality Symbolic
serious/cool
Not natural Negative - Attribute Product
meaning related
a
Item excluded from the quantitative stage to retain parsimony.
b
Item excluded from the quantitative stage as it is a response to a non-direct-meaning question.

114
Brand attitudes are defined as the psychological tendencies expressed by
evaluating a particular entity with some degree of favour or disfavour
(Jun, Cho, and Kwon 2008). Most, but not all of the evoked meanings
might serve as a basis for forming an attitude. However, only in very few
instances (e.g. average for Toyota, or good investment for Nokia) did an
evoked meaning in its form correspond to a definition of brand attitude.
Consequently, the conclusion is that brand meaning is a construct different
from brand attitude as well.

The respondents were also asked to indicate their feelings for the brands.
The answers to this question were neither versatile nor related to the other
answers presented in tables 5-6 to 5-10. Rather, the feelings were mainly
described in terms of liking or not liking the brand, and only sometimes in
terms of some of the issues already mentioned in a direct-meaning
question. The conclusion is that the question on feelings actually revealed
attitudes to a brand. This finding further provided evidence that brand
attitude is a different construct from brand meaning.

Similarly, the respondents were asked to define their relationships to the

brands. They mainly described brands as their friends, neighbours or
acquaintances. These answers have little value as they were given as
examples by the interviewer after discovering that most of the respondents
had problems understanding the question. The analysis of the relationships
and their comparison to the evoked feelings revealed that type of
relationship with a brand gave no extra value to the research neither in
terms of yielding new meanings, nor in terms of discovering new details
related to feelings towards the brands and frequency of their usage.
Moreover, the elicited types of relationships were quite related to the
elicited feelings. For example, the brand that the respondents loved was
considered a friend, best friend or an old friend, while the brand they did
not like was considered an acquaintance or a neighbour. The conclusion

115
was reached that brand relationship is also a concept entirely different to
the concept of brand meaning, so it was redundant in the research.

Table 5-7. The Types of Responses for Google

Attribute/ Brand meaning

Meaning Stimulus Benefit
personality category
Gives answers/provides data /all Meaning Functional Attribute Product related
knowing/ informative/ source of
information a /key to everything a
/encyclopaedia/ Wikipedia a/
information a /knowledge
Fast/quick a /high speed/not slow a Meaning Experiential Attribute Product related
/efficient /timely a /at any time/ /Functional
Not old/new/not antique a /not historic Meaning Symbolic Attribute Product related
a
/not old fashioned/ not out of date/up
to date a /young/ innovative
World/worldwide/global Meaning Symbolic Attribute Product related /
provenance
Powerful/potent a /almighty/ took Meaning Symbolic Personality Provenance
earth control /dominant /multi billion
industry/not poor a
Life saver/basic necessity/very Meaning Experiential Attribute Symbolic/ product
important related/
No privacy/no freedom/no Meaning Experiential - Product related
independence /omnipresent a /taking
over our lives a
Useful/helpful Adjective Functional Attribute, Product related
personality
Google logo c Meaning - - Sign
Happiness/satisfaction Meaning Experiential - Product related
Huge/extensive b /not small b Adjective Functional Attribute Product related/
provenance
Not reliable b Adjective - Personality Product related
Easy to use/convenient/no problem Meaning Experiential Attribute Product related
/Functional
Search a Meaning Functional - Product related
Not stupid/not dumb/not ignorant/ Adjective Symbolic Personality Symbolic
smart/savvy b
Reliable/trustful Adjective Experiential Personality Product related
Creative Adjective Experiential Personality Symbolic
a
Item excluded from the quantitative stage to retain parsimony.
b
Item excluded from the quantitative stage as it is a response to a non-direct-meaning question.
c
Item excluded from the quantitative stage as no difference across cultures was predicted.

116
Brand image is created by marketing programmes that link strong,
favourable and unique associations to the brand in memory (Keller 2003b,
p. 70). Band image is a brand identity transferred to consumers’ minds
(Kapferer 2004). Its main characteristic is that it is shared among
consumers. Since some of the brand meanings identified in the interviews
belonged to very private and particular associations, it is concluded that
brand meaning is also a different concept to brand image.

In relation to brand beliefs, as explained in chapter 2.1.1.5, the

terminology and the semantics are rather complex. According to Orth and
De Marchi (2007), beliefs are defined as features, attributes and benefits,
and sometimes as brand cognitions. As afore elaborated, brand meanings
include features, attributes and benefits, but are not limited to them.

When it comes to whether meanings are of a cognitive or an affective

nature, the affective responses can be observed from two points of view.
Firstly, whether reactions to the question were of an affective nature and
secondly whether the respondents relate certain emotions to a brand, but
did not experience emotion at the moment of responding to a question
(Franzen and Bouwman 2001).

With regards to the first type of affection, some of the respondents were
very passionate in their answers, and emotions could be clearly recognised
from what they said. For example, a feeling of pride was evident in
Hanna’s comment on Nokia: [Nokia] is our shiny and bright star, it is the
only product that we can take pride in on a global scale; nostalgia in
Michaela’s comment on Toyota: I learnt driving with it...It was a very
good car…I personally see no much difference to a BMW… we were very
happy with the car and I loved driving it ;),a feeling of joy and self-
esteem in Tiia’s comment on Nike: Freedom: Chance to be who I am and
do what I like; and a feeling of resentment towards Coca-Cola in Minna’s

117
comment: [Coca-Cola] is a brand to me, which I don`t buy. It`s too
American; It gives in my opinion a pompous picture: life is healthy, happy
people... no problems, no worries; it gives a fairytale side of life.

When the answers were analysed in terms of the second type of emotion,
as defined by Franzen and Bowman (2001), the responses were mostly of
a cognitive nature. However, there were a number of responses that were
of the affective nature, such as satisfaction and happiness for Google, and
fun for Coca-Cola. This argument provides evidence that brand meaning is
not equivalent to brand cognition. It also means that brand meaning is not
equivalent to brand beliefs and brand knowledge, because both of them
represent cognitive brand concepts (Peter and Olson 2001).

Franzen and Bouwman (2001) offered the best overview of what brand
meaning is. For this reason, their ten types of meanings were used in this
study (last column in tables 5-5 to 5-10). Although the data are of a
qualitative nature and have not yet been tested for their correlation,
generally it can be concluded that the most populous meaning category is
by far product related meanings, followed by symbolic meanings and
brand provenance.

However, although Franzen and Bouwman (2001) did not provide a

unique consolidated definition of brand meaning, they clearly characterise
it as a cognitive concept. This exploratory study provided evidence that
many emotions are included in the brand meanings as well. Thus, the
evidence from this study is more in line with the view of Richins (1994),
who claims that an object’s meanings represent the totality of an
individual’s thoughts and feelings about the importance of an object. He
further differentiates public meanings from private ones, so that public
meanings are likely to have an influence on shaping desire, while private
meanings on consumers' feelings about the things they already possess.

118
Table 5-8. The Types of Responses for Nike

Attribute / Brand meaning

Meaning Stimulus Benefit
Personality category
Sports/sportiness/sporty Meaning Functional/ Attribute Product related
Symbolic
Achiever/winner a Meaning Symbolic Personality Symbolic
/victory/winning/not failure/not
losing a
Not formal/casual/be myself b Adjective Symbolic Personality Symbolic
Sweat Sensory Experiential - Product related
image
Athletes/athletic Meaning Functional/ Attribute Symbolic
Symbolic
Cool/trendy Adjective Symbolic Personality Symbolic
Expensive/pricey/not cheap Adjective - Attribute Perceived price
Sweat shops/child labour/not fairness Meaning - - Provenance
a
/exploitativeness a /not socially
responsible/not sustainable
Good quality/not low quality/durable Adjective Functional Attribute Perceived quality
a
Not lazy/active/not being immobile Meaning Symbolic Attribute Symbolic
/not being still/activity/movement a
Sport shoes Meaning Functional - Product related
Crowd cheering Sensory Experiential - Product related/
image Symbolic
Cushion for feet/walking on Sensory Experiential/ Attribute Product related
clouds/soft-footed a /air technology image Functional
Hi-tech/innovative/innovation Meaning Functional Attribute Product related
Jogging/running Meaning Functional - Product related /
situational
Michael Jordan Meaning Symbolic - Advertising
Powerful Meaning Symbolic Personality Symbolic
Rough/not weepy b/not weak Adjective Symbolic Personality Symbolic
Self-assured/just do it/positive Meaning Symbolic Personality Symbolic
thinking
Swoosh c /Nike logo c Meaning - - Sign
a
Item excluded from the quantitative stage to retain parsimony.
b
Item excluded from the quantitative stage as it is a response to a non-direct-meaning
question.
c
Item excluded from the quantitative stage because no difference across cultures was
predicted.

119
Table 5-9. The Types of Responses for Nokia a

Attribute / Category of
Meaning Stimulus Benefit
Personality brand meaning
Reliable/will not ditch you/ feeling Meaning Experiential Personality Product related
of security/ handy/ there for you
Hi-tech/technology/ engineered Meaning Functional Attribute Product related

Good/good investment/quality/ Meaning Experiential Attribute Perceived quality

safe choice
Connecting people Meaning Functional Attribute Advertising
Finnish Meaning Symbolic Attribute Provenance
Easy to use /user friendly/ Meaning Experiential Attribute Product related
functional/not complicated
Sturdy/strong/tough/hard Meaning Experiential/F Attribute Product related
unctional
Communication Meaning Functional - Product related
Smooth keypads/sleek/clean Adjective Experiential Attribute Product related
design Sensory i.
Friends/family Meaning Symbolic - Situational
Not a camera/just phone/phone/ Meaning - - Product related
not diverse
Modern/trendy/not old fashioned/ Adjective Symbolic Personality Product related
not old
First cell phone Meaning Symbolic - Situational/
provenance
Innovative Meaning Functional Attribute Product related
Not cutting edge/ not up to date/ Negative - Attribute Product related
not innovator/not state of the art adjective
Global player/not local Meaning Symbolic Attribute Provenance
Not exciting/not cool/not exotic Negative - Personality Symbolic
adjective
Nokia original ring tone Meaning Experiential Attribute Brand sign
Durable/break-proof/long lasting Meaning Experiential/F Attribute Product related
unctional
a
The brand was not studied in the quantitative stage, so there was no need to identify the
items for exclusion.

120
Referring to the gathered field data, the comment that best (although
perhaps rather extremely) indicated the role of feelings in brand meaning
creation was Robert’s comment on Nokia:

That was the one brand I didn't really have anything for. Nokia is not a major
part of the cell phone market here and I have never used a Nokia phone. I can
quickly get some pictures, but I don't really have a feeling one way or another
towards Nokia. It’s not a matter of finding pictures.

Robert actually implied that he has certain associations to Nokia, but

because he did not have any experience with it, he has no feelings towards
it and thus Nokia does not have any meaning to him.

This elaboration provided grounds for differing brand meaning from mere
brand association as well. Kirshnan (1996) defines associations as links
between any two pieces of information. Similarly, Low and Lamb (2000)
comment that associations can be anything linked in memory to a brand.
Compared with general associations that could be based on mere
awareness of a brand’s existence, brand meanings are associations based
on brand experiences. Such associations are likely to be self-related
(Burnkrant and Unnava 1995), held with more certainty (Smith and
Swinyard 1983), and also represent the basis for autobiographical
memories (Baumgartner, Sujan, and Bettman 1992). In conclusion, brand
meaning is a different concept to brand associations as well.

121
Table 5-10. The Types of Responses for Toyota a

Attribute / Brand meaning

Meaning Stimulus Benefit
Personality category
Reliable/no surprises Meaning Functional Attribute Product related
/dependable/not unreliable/ not
undependable/long lasting
Average/ok/good/practical/ Meaning Functional Attribute Product related
grey/safe choice/rational
Safe Meaning Functional Attribute Product related
Not luxury/not premium /not Negative - Attribute Product related
exclusive/not high class/not status meaning
symbol/not the highest brand value
Innovative/hybrid car Meaning Functional Attribute Product related
Japanese hard working engineering Meaning Symbolic - Provenance
team/JIT/robots/ Japanese car
Comfort/comfortable/relaxing Meaning Functional/ Attribute Product related
Experiential
Expensive/not cheap/overpriced Meaning - Attribute Perceived price
Not exciting/not adventurous/ not Negative - Personality Symbolic
passion/ boring/not extravagant not meaning
sexy/no emotions
Family brand/children/family car Meaning Symbolic - Symbolic
High quality Meaning Functional Attribute Perceived quality
Not American/imported (for Adjective Symbolic Attribute Provenance
Americans)
Value for money/worth the money Adjective Functional/ Attribute Perceived price
Experiential
Class/luxury/high-end/regal/grace Meaning Symbolic Personality Symbolic
Not for young/not for the next Negative - Attribute Symbolic
generation meaning
Not expensive/quite cheap Negative Functional Attribute Perceived price
adjective
a
The brand was not studied in the quantitative stage, so there was no need to identify the
items for exclusion.

Obviously, there are grounds for discriminating brand meaning from the
other branding concept. An overview of how each of the brand concepts
differs from brand meaning is given in table 5-11.

122
Table 5-11. Brand Meaning in Relation to the Other Brand Concepts
Concept Difference from brand meaning
Brand belief/knowledge Does not include affective perspective
Brand attitude An entirely different concept
Brand benefit Does not include negative meanings
Brand attribute Includes limited array of meanings
Brand personality A very narrow brand related concept
Brand relationship Focuses solely on a relationship
Brand identity Focuses on brand relevant features from brand’s perspective
Brand association Does not demand experience with the brand

Defining a construct in terms of what it is not is generally an important,

but only a first step in defining what it is. The findings of this study
provide a solid ground for understanding the brand meaning concept as a
set of cognitive and affective experience-based brand associations.

5.2.4.2. Differences in the Brand Meanings across Cultures

As described, analysis of the interviews discovered initial differences in

meanings among cultures. Very importantly, some of these differences
were discovered not to be caused by consumer perceptions, but rather by
differences in the positioning strategies employed by the companies
themselves. Such differences were discovered for Nokia and Toyota.

Toyota is a brand endorsing a higher investment product, so among the

studied brands its perception is the most affected by the country’s
economic status. Being the richest of the studied countries, Finland had
the worst perception of the brand, while India being the poorest, had the
best. Moreover, in the Indian market, Toyota deliberately chose to move
away from positioning its cars as being middle-class and reliable (as they
are positioned in the other three countries). For the Indian market, Toyota
chose to capitalise on the image of being expensive and out of reach. It
only marketed a few of its models in India, all of which are bigger cars

123
targeted at higher class consumers. Praveen referred to this with a
comment:

… they have like 4 cars in India. All of which for a lower - upper class of
India is out of reach.

Nokia also does not have a unique positioning around the globe. It is not a
well known brand in the United States, where it occupies a very small
market share. Such a market position is also reflected in graph 5-5, in
which it is visible that the American respondents only submitted images of
Nokia that displayed either a logo or a cell phone. This is because
consumers need experience in order to be able to broaden their scope of
associations and meanings (Holt 2004), and obviously the respondents
from the United States lack experience with Nokia. The position that
Nokia has in consumers’ minds in the United States is also reflected in
Robert’s comment:

…Nokia is not a major part of the cell phone market here and I have never
used a Nokia phone… I don't really have a feeling one way or another
towards Nokia…We don't really use that brand in the US.

The perception of Nokia in Finland is, on the other hand, very biased.
Nokia is much and in many ways incorporated in the everyday lives of
Finns. Therefore, for Finns, it bears many meanings that the biggest
national employer and the most well-known national brand naturally
bears. Among the other meanings, the Finnish respondents reported high
national pride for Nokia as well as anger over the layoffs of many people
in Finland when the production facilities moved offshore to a cheaper
location. For example Hanna says:

Nokia is such a common phone brand in Finland - it is hard for me to

separate general mobile from Nokia. It is the only product that we can take
pride in on a global scale.

124
 In conclusion, as a result of the analysis of the reasons behind the
differences in the meanings of the five studied brands, only Coca-Cola,
Google and Nike were identified as being truly global brand (i.e. those
universally positioned). This is in line with Hankinson and Cowking
(1996), who claim that there are very few truly global brands, despite the
widespread application of the term “global brand”.

Table 5-12. The Importance of Coca-Cola, Google and Nike

Question Brand 1 Brand 2 Brand 3 Brand 4 Brand 5

What brand would you most like to sit next x x
to at a dinner party?
What brand, if sent back 100 years, would x x x
have the biggest impact on history?
What brand that no longer exists would you x x x x x
resurrect?
Which brand do you want to argue with? x x x x

Which brand inspires you the most? x x

If you were to describe yourself as being a x x x

brand, what brand would you be?
If you could re-brand any brand, what brand x x x x
would it be?

What brand can you not live without? x x x

What brand is most likely to revolutionize x x x

the branding industry in the next five years?
What brand do you think is truly (going) x x x x
"green"?
Source: Thompson 2008.

Interestingly, only brands originating from the United States seem to be

truly global. At this point, a spontaneous question arose: is the world
becoming globalised (Levitt 1983), Westernised (Gram 2007) or actually
just Americanised? Despite its appeal, the answer would have to be found
in some other study as it was outside the scope of the present one.

125
As further proof of the importance of Coca-Cola, Google and Nike,
Brandchanel’s 2008 Brandjunkie award (Thompson 2008) explored trends
in the world of brands. According to their results, branding professionals
and brand enthusiasts around the world chose these three brands as some
of the most fascinating brands. This is evident in table 5-12, as these three
brands (together with Apple) constantly reappear in all the investigated
topics.

For the above-mentioned reasons, Coca-Cola, Google and Nike were used
in the quantitative stage of this research. These brands were also
considered very appropriate for this study as they are well-known brands,
heavily consumed, and affordable for students in all the studied countries.

126
 6. Empirical Research on Cross-cultural
Differences in Brand Meanings

The second stage of the empirical research was conducted quantitatively

with the aim of testing the hypotheses and discovering the differences in
brand meanings among cultures.

6.1. Pilot Research

The pilot phase of the research was conducted with the intention of testing
the measurement instrument and understanding which of the identified
meanings form statistically sound entities. During this phase, it was also
the aim to concretely specify cultural dimensions that influence the
identified groups of meanings.

6.1.1. Research Procedure

6.1.1.1. Questionnaire Design

Dolnicar and Rossiter (2008) demonstrated that the low stability of

consumers' brand-attribute associations (temporary associations) can be
partly explained by the prevailing methods used in market research;
hence, they propose seven methodological factors that influence the
increased stability of brand-attribute associations. They are as follows:
using the respondent’s favourite brand, making the evaluation task easier,
directing the brand association relationship from the brand to the
association, and using brands and languages the respondent is familiar

127
with. This thesis applied all of the advice, with the exception of the
respondent’s favourite brand.

This advice could not have been applied because the nature of the study
demanded evaluation of the same brand by all the respondents.

For this study, self-relevant associations were crucial (cf. Burnkrant and
Unnava 1995), so the associated experience technique was applied. This
technique is a neuro-linguistic programming (NLP) technique that refers
to reviving the situation or event by experiencing it as if one was reliving
what happened. This is made possible by stimulating all the senses, so that
the respondents see through their own eyes, feel with their own body and
emotions, hear what they heard, and smell and taste what occurred at the
time of the revived experience (Dilts and DeLozier 2000). Therefore, the
evaluation of each brand was preceded by an introductory line that evoked
an associated experience. For example, for Coca-Cola, the following line
was used:

Coca-Cola is a drink we often encounter. Some of us like to drink it and some

don't. As you look at the following questions, please answer them while keeping
in mind how it feels to be in contact with Coca-Cola.

Before the questionnaire was distributed to the students, it was also given
to a panel of carefully chosen respondents that consisted of five marketing
and branding experts, three methodology experts and two marketing
students who were interviewed for their opinions on the quality of the
questionnaire - including items that might be difficult to understand for a
non-native English speaker. Subsequently, as the comments were accepted
and some changes applied, the questionnaire was distributed to a business
students sample at the University of Rijeka. This phase of the research
served to further test for ambiguous items that should be corrected or
excluded (Potsakoff et al. 2003), as well as to determine the reliability of
the measures (Ruigrok, Gibbert, and Kaes 2005).

128
The version of the questionnaire distributed to a student sample at the
University of Rijeka consisted of a set of demographic questions and three
sets of brand-related questions; one for Coca-Cola, one for Google and
one for Nike. To investigate the meanings of each brand, four questions
were posted within the set. An open-ended question asking for the
meaning of the brand was the first question in the set. It was important to
pose such unaided questions at the beginning to encourage the respondents
to think of the brand (envisage it and the situation they use or used it in)
without being primed to give particular answers (cf. Park, Milberg, and
Lawson 1991; Henderson et al. 2003).

In the second question of the set, the respondents were asked to specify to
which extent (5 point Likert-type scale) the brand-specific items presented
in the previous chapter indicated what the brand means to them. The third
question offered some of the images gathered in the interview stage, and
the fourth question some of the product categories that the studied brands
have some meaning in common with. In both the third and fourth
questions, the respondents were asked to rate the goodness of fit between
the brand on one side and the image and the product category on the other.
Appendix 2 presents the images and product categories used for each of
the studied brands.

In addition to these four questions aimed at discovering brand meanings,

the respondents also rated the extent to which they liked or disliked the
brand in question, and whether they consumed it.

6.1.1.2. Questionnaire Distribution

Based on the responses from the previous phase, the online questionnaire
was constructed using the online application Survey Monkey10. An online

10
www.surveymonkey.com

129
survey ensured: that only those students that were interested in the
research topic replied, that they replied at the time most convenient for
them, and that they replied in full anonymity. All of these conditions
increased the reliability of their answers.

The link to the questionnaire was distributed via e-mail to 260 business
students at the University of Rijeka and also posted on a student web
portal of the Faculty of Economics in Rijeka. A Total of 145 students
began the survey, but only 105 completed it. In the circular e-mail sent
consequently, the students were asked why they had abandoned the
questionnaire before its completion. The reasons included termination of
the connection with the server, problems with English, and duration of the
questionnaire. Analysis of the duration of the questionnaire completion
revealed that only 51 questionnaires were completed within 30 minutes,
and altogether 74 within 45 minutes. As many as 17 questionnaires were
completed in two phases (with a break in the middle). These results
indicated that the length of the questionnaire needed to be reduced and
that the more complex words should be translated.

6.1.2. Results

Cronbach’s alphas were calculated for the initially identified groups of

meanings. Some of the alphas were low, so in order to get a better
structure of the items, an explorative factor analysis was due. However,
because the ratio of items to observations was less than the recommended
minimum of 1:5, (Costello and Osborne 2005), subsets of correlated items
were identified and factor analyses were conducted on these subsets.
Based on the results of the factor analyses, new groups of items were
created. Cronbach’s alphas were recalculated, and in some cases several
items were deleted to further improve reliability.

130
The images and product categories across the meanings and across the
brands were generally the weakest items. Although the presented images
were those that were submitted in the qualitative stage of the research as
representations of particular meanings by some interviewees, the
interpretation of an image is individual and dependent on the individual’s
previous knowledge (Berthon, Pitt, and Campbell 2008). Thus, the images
did not evoke the same meanings for all the evaluators in the quantitative
stage. As a result, all the images were deleted from the survey in the main
phase. For similar reasons, as carriers of multiple meanings, the product
categories were excluded as well.

Apart from the image and product categories, some other items were
selected for removal for three reasons: reducing alpha scores and at the
same time not logically or statistically belonging to some other group of
meanings; having been the result of non-direct meaning questions in the
interview, and at the same time not highly correlating to another group of
meanings; and belonging to a meaning categories with many items.
Although some groups of meanings reported alphas lower than the
recommended cut-off point of Į = .70 (Nunnally 1978) even after
elaborated manipulations, they were kept for the subsequent research in
order not to lose some of the identified meanings. This is because the
correlation tests revealed that some of the items may belong to some other
groups of meanings. Their appropriate group of meanings were
determined in the factor analysis in the main phase of the research.

Finally, there were a few cases of items that were mentioned twice in the
same country in the interview stage. These cases were not included in the
pilot research as they were below the cut-off point of three mentions per
item. However, additional consideration of the issue revealed that they
should not be ignored as they might have captured a country-specific
meaning which could have been of special importance to this study. In

131
 addition, there were also several additional meanings that were often
mentioned in the analysis of the answers to the open-ended question on
meaning posted in the pilot research itself. Consequently, such meanings
were added in the main phase of the quantitative research. In the following
text, each brand’s groups of meanings are presented separately.

Table 6-1. Coca-Cola’s Groups of Meanings and Their Hypothesised

Relationships to the Cultural Dimensions

Group of Cron. Excluded Cron. Cultural

Meanings
meanings Į items Į dimension
Refreshing Refreshing/cooling/fresh/freshness .871 Image of ice, .878 No effect
/cold /warm summer/hot day/ fresh
image of ice
Unhealthy Not healthy/harmful .531 Harmful .676 Assertiveness –
/unhealthy/not natural Performance
Sweet/sugar/image of sugar .703 Image of sugar a orientation -
Christmas Christmas/Santa Claus/Christmas .900 Image of Santa .894 Human
song/image of Santa Claus Claus orientation +
Global Global/not local/huge/image of .122 Not local, huge, - Power distance
globe image of globe b +
American American/American .740 Image of .684 Collectivism +
lifestyle/image of American flag American flag
Carbonated Carbonated/sparkling .519 Carbonated, - -
sparkling
Energizing Energetic/drink it when tired/no .728 - .728 No effect
sleeping/keeps me going
Recognisable Recognisable/familiar/not .428 - .428 Uncertainty
unknown/identifiable avoidance+
Classic Traditional/classic/original .569 - .569 Uncertainty
avoidance+
Assertiveness-
Bonding with Bonding with friends/parties/fun/ .624 Parties, image .707 Collectivism -
friends image of friends of friends
Cool Exciting/not boring/not .550 Not serious .650 Assertive +
serious/cool
a
obesity, lot of calories added from Indian responses.
b
great brand, worldwide and multinational added from the pilot survey.

132
Table 6-1 shows the Cronbach’s alphas for the meanings belonging to
Coca-Cola. Some groups of meanings were statistically very robust:
Christmas, American and energizing. The groups of meanings cold,
refreshing and warm summer were discovered to be highly correlated to
each other, and hence also formed one very robust group of meanings. The
Sweet group of meanings was reliable, but it was also statistically strongly
correlated to the not healthy group. Consequently, these two groups of
meanings were summarised in one common group of meanings. The
newly introduced items obesity and lot of calories were predicted to
correlate to the sweet group of meanings.

The items of the global group were not very coherent. Most of these items
were removed and were replaced by worldwide, multinational and great
brand in the main phase of the research.

The Cronbach’s alphas for Google’s groups of meanings are presented in

table 6-2. Unlike in the case of Coca-Cola, Google’s groups of meanings,
with the exclusion of the intrusive group of meanings, were all robust and
showed alphas Į > .66. The Intrusive meaning was probably not coherent
because Google was viewed by many respondents as a “playground” that
allows free, unrestrained surfing across borders. However, there was a
group of respondents who observed it as a sort of “Big Brother” that is
very subtly acquiring a lot of information on users’ profiles and habits.
This contradicts the basic premises of freedom and privacy. However,
despite the removal of the meaning freedom, the Cronbach’s alpha did not
improve greatly, which might indicate that the other two items also have
ambiguous meanings.

Meanings often mentioned in the pilot questionnaire were observes

people’s habits, fun, entertainment, connecting to the world. The groups
of meanings they belonged to are marked in table 6-3.

133
 Table 6-2. Google’s Groups of Meanings and Their Hypothesised
Relationships to the Cultural Dimensions

Group of Cron. Cron. Cultural

Meanings Excluded items
meanings Į Į dimension
Informative Gives answers/provides .797 Provides data, .713 Uncertainty
data/all knowing/ informative/ gives answers avoidance +
encyclopaedia/knowledge
Not dumb/ not .650 Not dumb, not
stupid/smart/not ignorant stupid, not ignorant
Efficient Fast/high .633 Image of speed .677 Performance
speed/efficient/timely/at any orientation +
time/easy to
use/convenient/image of speed
Inspired Not old/new/not old fashioned/ .666 Not out of date/ not .662 Uncertainty
not out of date/up to old fashion avoidance –
date/young/innovative/creative Assertiveness +
Performance
orientation +
Almighty World/worldwide/global/huge/ .476 Image of globe a .705 Power distance
image of globe +
Powerful/almighty/took earth .596 Took earth control
control/dominant /multi billion image of dominant
industry/image of dominant
Fun Happiness/satisfaction/no .676 Image of smiley b .730 Human
problem/image of smiley orientation -
Helpful Life saver/basic necessity/very .574 .742 Human
important orientation +
Reliable/trustful .680 Assertiveness -
Useful/helpful .581
Intrusive No privacy/no freedom/no -.821 Image of freedom c .466 Collectivism -
independence/image of
freedom
a
connecting to the world added from the pilot survey
b
fun, entertainment added from the pilot survey
c
observes people’s habits added from the pilot survey

In the case of Nike, several separate groups of meanings were strongly

correlated, so for example sport, jogging and sport shoes formed one
common group; achiever, self-assured and crowd cheering the second
group; and pricey, good quality, cool and innovative the third group. All

134
 of these extended groups had high Cronbach’s alphas as table 6-3 reports.
The items not weak and not failure loaded to another group than the
predicted one, so they were reassigned to a different group of meanings as
shown in the table.

Table 6-3. Nike’s Groups of Meanings and their Hypothesised

Relationships to the Cultural Dimensions

Group of Cron. Cron. Cultural

Meanings Excluded items
meanings Į Į dimension
Sports Sports/sportiness/sporty/athletes/ .855 Sportiness, sport .842 No effect
athletic/jogging/running/sport shoes shoes a
Activity Not lazy/active/not being .374 Active, activity, not .598 Assertiveness
(Renamed to still/activity being still (Added not
Not weak) weak, not failure)
Sweat Sweat n.a. sweat - -
Achiever Achiever/victory/winning/not .678 Image of victory; not .815 Performance
failure/image of victory failure orientation +
Self-assured/just do it/positive .558 Image of self-assured Assertiveness +
thinking/ image of self-
assured/crowd cheering
b
Michael Jordan n.a.
Powerful Powerful/image of powerful .281 Image of powerful - Power distance
+
Rough Rough, not weak .113 Rough, not weak - -
Casual Not formal/casual -.212 Not formal, casual - -
Innovative Good quality/not low quality/durable .609 - .746 Performance
orientation +
Expensive/pricey/not cheap .669 Not cheap
Cool/trendy .557 -
Hi-tech/innovative/innovation/ .545 High tech, image of
image of high tech high tech
Not CSR Sweat shops/child labour/not socially .612 Not Sustainable .662 Assertiveness –
responsible/not sustainable
Human
orientation +
Cushioning Cushion for feet/walking on .345 Cushion for feet, n.a. No effect
technology clouds/air technology walking on clouds c
a
added fitness from the Indian respondents’ interviews
b
added Tiger Woods from the American respondents’ interviews, whereas life and freedom from the pilot
survey and young from the Austrian respondents’ interviews
c
added comfortable from the pilot survey

135
Some of the groups that were neither robust nor identified in the direct
meaning question were still kept for the main phase research because of
the new items that were added to the research as a result of their
appearance in the pilot survey. To be exact, the items life and freedom
discovered among the pilot survey respondents were assumed to belong to
the group of meanings achiever, whereas comfortable was assumed to
belong to the group of meanings cushion for the feet. In addition, Tiger
Woods, fitness and young were mentioned by the respondents from the
same country, so they were added to the main phase research.

At this stage of the research, since the meanings were reduced to rather
coherent groups, it was already possible to determine which group of
meanings would be influenced by which of the cultural dimensions. In the
last column of the tables 6-1, 6-2, and 6-3, the cultural dimension which is
hypothesised to influence a particular group of meanings is presented. In
addition, the direction of the influence is also indicated next to the
dimension, where + indicates a positive direction, and - indicates a
negative one.

6.2. Main Phase Research

6.2.1. Research Procedure

Based on the results of all the preceding phases, a new questionnaire was
constructed (Appendix 3). The literature claims that both five-point and
seven-point scales are equally often used (Jamieson 2004). Although a
five-point scale was used in the pilot research, the decision was made to
change to a seven-point scale for the main research phase. This is because
as a general rule it is better to use as wide a scale as possible, however the

136
seven-point scale reaches the upper limits of the scale’s reliability (Allen
and Seaman 2007).

The questionnaire was distributed to business students in the four chosen

countries. In each country, the six biggest cities were identified and in
each a business department of the local university was contacted. Apart
from the official contacts obtained on the universities’ web pages, several
faculty members were contacted directly for their help in distributing the
questionnaire. Finally, direct contact with student organisations and clubs
(in particular AIESEC) proved to be the most fruitful.

The questionnaire was available online for 18 days in December 2008. It

consisted of three sets of questions for the three studied brands and an
introductory demographic set of questions. Three demographic questions
were crucial to filter out the respondents who had lived abroad for longer
than a year and longer than six months at a stretch during the last five
years; those older than 31; and those not studying business or in a business
related field. Such a filtering procedure ensured that the samples were as
homogenous as possible.

The online environment allowed for a split questionnaire design (Adigüzel

and Wedel 2008). The questionnaire was broken into four pages: the first
page consisted of demographic questions and the latter three consisted of
the branding questions – one page for each brand. The participants could
exit the questionnaire after each page, but were forced to complete all of
the questions within one page. In the case they left a page halfway
through, their responses for the current page were not registered. Such a
procedure increased the response rate (Aaker, Kumar, and Day 2001), and
improved the accuracy of the responses (Dolnicar and Rossiter 2008).
This is because the participants were not forced to sit through all of the
sets of questions, which reduced their fatigue and boredom (Adigüzel and

137
Wedel 2008). The brands were displayed for evaluation in an arbitrary
order, i.e. pages two to four of the questionnaire were iterated randomly so
that each respondent started with a different brand. This ensured that each
brand received about the same number of evaluations and that no
distortions caused by late placement in the questionnaire occurred (Herzog
and Bachman 1981).

6.2.2. Data Analysis Procedures

The quantitative data was analysed using SPSS statistical software. Shook
et al. (2003) claimed that data analytical tools such as data reduction and
tests of means are among the most commonly used in business studies.
Besides general descriptive and correlation statistics, the above analyses
were used in this thesis as well. General considerations regarding data
reduction analysis, as well as analysis of variance and its non-parametric
alternatives are given in the following subchapters.

6.2.2.1. Data reduction method

There were 39 items representing 39 Coca-Cola’s meanings, 42 items

representing Google’s meanings and 37 items representing Nike’s
meanings. Such high numbers of variables are usually reduced to
manageable number of factors or components. Barlett’s tests of sphericity
were statistically significant (=.000 for all three brands) and Kaiser-
Meyer-Olkin measures of sample adequacy were high (>.80 for all three
brands), which indicated that items of each of the three brands correlated
and were suitable for data reduction analyses. Because data reduction
analysis is a highly influenced statistical method (i.e. it involves many
analytical decisions by a researcher), it requires rigorous due diligence
(Fabrigar et al. 1999).

138
Although, most of the discussion on factor and principal component
analysis is directed towards the method of extraction and rotation of
factors/components, (Velicer and Jackson 1990, Schönemann 1990), it is
far more critical to obtain large enough samples (Costello and Osbourne
2005) and define the correct number of factors/components to be extracted
(Velicer and Jackson 1990).

According to Velicer and Jackson (1990), the correct number of

extractions will reduce differences among different sub-methods.
Furthermore, factor and principal component analyses are generally large
sample methods that become increasingly stable with an increasing
number of observations. However, the proposed ratios between the
number of observed variables and the number of observations is just a rule
of thumb and should not be used as a rule (Costello and Osborne 2005). In
other words, when at least three or four measured variables represent each
common factor or component and the communalities are an average of .70
or higher, even 100 observations suffice; whereas samples of even 800 are
necessary with weaker results (Fabrigar et al. 1999).

Usually, the number of extracted factors or components corresponds to the

number of Kaisers’ eigenvalues higher than one. This method to
determine the number of factors or components is probably the most
commonly used, because it is a default method in most statistical
programmes. However, it is among the least accurate methods and often
leads to over-extraction (Velicer and Jackson 1990, Fabrigar et al. 1999;
Costello and Osborne 2005). The Scree test procedure proposed by Cattell
(1966) is considered more appropriate (Costello and Osborne 2005),
although it has some inherited subjectivity problems as well (Fabrigar et
al. 1999). According to this method, the most appropriate model is chosen
from a series of alternative models that differ in their complexity (i.e., the
number of factors or components) (Fabrigar et al. 1999). Firstly, a test

139
should be run on the a priori projected number of differentiated variables,
then at the number of points above the inflection point on a scree plot, and
then at numbers above and below those numbers. The results should then
be compared and the one offering the cleanest factors/components (items
loadings above .30; few items loading on more factors, no factors with
fewer than three items) should be retained (Costello and Osborne 2005).

As for the extraction method, a general consensus is that the principal

component analysis is a data reduction method, whereas factor analysis
(e.g. maximum likelihood and principal axis components) is applied to
identify the structure of the items (Costello and Osborne 2005). Factor
analysis is thus used when another study follows or when items are
summed up or averaged according to the structure, and as such taken in
the further analysis. Since component scores not only reduce data, but also
normalise them (unlike summed or averaged solutions), principal
component analysis was considered more appropriate for this study.

Finally, when it comes to the method of rotation, the two main categories
are orthogonal and oblique. While orthogonal rotations do not allow
components to be correlated, oblique rotations permit their correlations. If
the solution with the best simple structure involves orthogonal
components, oblique rotation will provide estimates of the correlations
among components close to zero and produce a solution that is quite
similar to that produced by a successful orthogonal rotation (Fabrigar et al.
1999). Hence, it is by far the better option to use oblique rotations in any
case.

6.2.2.2. Analysis of Variance

In order to investigate whether different meanings have different

importance in different countries and for different types of users, a two-
way analysis of variance (ANOVA) for independent samples was

140
conducted for each meaning. Preconditions for analysis of variance
include homogeneity of variance and normality of data (Wilcox 2002).
Normality of data is tested by Kolmogorov-Smirnov tests, while
homogeneity of variance is tested by Levene’s tests. If these two tests are
statistically significant (p < .05), the between-subject effects have to be
tested with a non-parametric alternative to a one-way ANOVA (Corain
and Salmaso 2007); i.e. Kruskal-Wallis test (Allen and Seaman 2007) with
Mann-Whitney alternative to post-hoc tests (and manually performed
Bonferroni correction)11. If the results of the non-parametric test differ
from the parametric test, the non-parametric results are reported. In the
opposite case, the parametric results (applying Dunnett's C post-hoc tests)
are reported. Effect size for analysis of variance is measured by Ș2. Partial
Ș2 is used for more than one independent variable as in such way test
controls for the effect of the other variable in the model. Effect size is
considered small if Ș2 < .06, medium if .06 < Ș2 < .14 and large if Ș2 > .14
(Sawyer and Ball 1981).

6.2.3. Results

6.2.3.1. Sample Characteristics

The average duration of the questionnaire completion varied among the

countries. On average, the most efficient were the United States
respondents and the least efficient were the Indian respondents. The
discrepancy might in part be ascribed to cultural differences in time and
performance orientation (Hofstede 2001; House et al. 2004), but also to
the technical performance of the computers in the respective countries (a

11
The Bonferroni correction takes into account the total number of pair-wise comparisons
and divides the chosen level of p with the total number of comparisons. Non-parametric tests
(unlike parametric tests) do not include the Bonferroni correction, so the corrected p has to
be used. For three groups p < .017 and for four groups, p < .008 is the cut-off point for the
overall p to be <.05 (Homack 2001).

141
professor in India commented that pages loaded slowly). For some
participants, it took longer than an hour to complete the questionnaire; but,
depending on the country, between 70% and 85% of the respondents
completed the questionnaire within 20 minutes.

Altogether, 676 students participated in the research (205 Austrian, 148

Finnish, 178 Indian and 145 the United States respondents), out of which
59% evaluated all three brands. This percentage varied from 52% for
India, 59% for Austria and the United States, to 68% for Finland. The total
numbers of responses per each brand per country are presented in table 6-
4.

Overall, 159 respondents were removed from the sample for distorting
homogeneity of the sample. Those respondents were breaking either one
or more predetermined conditions: (1) being older than 31, (2) being
internationally exposed, and (3) not studying business or business-related
field.

Table 6-4. The Sample Characteristics

Initial sample Excluded a Final sample
Coca- All Inter. Non- Coca- All
Country Total Google Nike 31+ Total Google Nike
Cola brands exp. bus. Cola brands

Austria 205 142 130 129 120 30 11 7 162 111 101 101 94
Finland 148 125 109 107 101 16 5 2 126 106 90 90 86
India 178 118 105 100 92 16 3 26 137 85 73 66 63
US 145 112 94 88 86 42 2 20 92 64 53 49 47
Total 676 497 438 424 399 104 21 55 517 366 317 306 290
a
Some of the respondents did not fulfil more than one condition, so they occur in two or
three shadowed columns.

Among the other characteristics of the final sample, the most relevant is
that 8% of the respondents were pursuing a BSc programme, 20% a BA
programme, 19% an MSc programme, 34% an MBA programme and 20%

142
another programme. Furthermore, 71% of the respondents were aged
between 20 and 25, whereas 92% were aged between 18 and 27.

6.2.3.2. Brand Meaning

The Brand Meanings for Coca-Cola

In order to understand which meanings are most associated with Coca-

Cola, the collected responses were sorted in a descending order. Although
some experts are divided on the issue whether data obtained by a Likert-
type scale can be considered interval or just ordinal, a more stringent
approach demands that they be considered as ordinal values (Dittrich et al.
2007). Table 6-5 gives an overview of the initial 39 items (i.e. meanings)
sorted from the highest to the lowest median scores.

Items that best indicated what Coca-Cola means to the respondents were
worldwide, recognisable, not unknown, global, etc.. These items
represented a combination of symbolic and functional meanings. Items
that indicated the least what Coca-Cola means to the respondents were
keeps me going, no sleeping and bonding with friends. Most of these items
were functional, with the exception of a symbolic Bonding with friends
meaning. The results were similar in each country, except in India. There
the items the least associated with Coca-Cola were traditional and the
three Christmas-related items, whereas those the most associated with the
brand were great brand and multinational. Also, not healthy was only by
the American respondents mentioned among the top four Coca-Cola
associations.

In order to prepare data for the ANOVA tests (providing equality of

variance is met), a principal components analysis was conducted on 39
initially measured items. A total of 366 observations presented a ratio of
items to observations of 1:9. The Scree plot (figure ap4-1.) suggested

143
seven components. The initial principal component solution was relatively
good, except that one component contained only four relatively weakly
loaded items (all less than .60) with two cross-loadings (items that load
.32 or higher on two or more components (Tabachnick and Fidell 2001)).

Table 6-5. The Medians for Coca-Cola’s Meanings in the Overall

Sample a

Item Median b Countries c Item Median b Countries d

worldwide 7 US, F, A refreshing 5
recognisable 7 US, I, F, A obesity 5
not unknown 7 I, F, A cooling 5
global 7 A cool 5
identifiable 7 US, F traditional 5 I
not healthy 7 US freshness 5
multinational 6 I parties 5
American 6 warm summer 5
sugar 6 fun 5
great brand 6 I hot day 5
not natural 6 energetic 5
sweet 6 drink it when tired 4
American lifestyle 6 Christmas 4 I
unhealthy 6 Santa Claus 4 I
cold 6 Christmas song 4 US, I
original 6 exciting 4 F, A
classic 6 bonding with friends 3 US, F, A
familiar 6 no sleeping 3 US, F, A
calories 5,5 keeps me going 3 US, F, A
not boring 5
a
N = 366; all countries sample; data were sorted primarily on median and secondarily on mean.
b
Medians close to 1, show the items very poorly indicated, while those close to 7 show that
the items clearly indicated what each brand means to the respondents.
c
The Country initial indicates a country in which the item was evaluated as one of the top
four.
d
The Country initial indicates a country in which the item was evaluated as one of the
bottom four.

The number of components was then increased to eight and nine, but the
solutions were still not satisfactory as either many cross-loadings
appeared, or the items did not load above .32 on either of the components.

144
 In addition, it was difficult to interpret some of the components; hence,
several more solutions with seven, eight or nine components were
modelled each time excluding some of the weakest items. Finally, the best
solution was obtained after the exclusion of the problematic American
lifestyle and not boring items. The final solution included seven
components. They were named: Refreshing, Globally recognised,
Unhealthy, Christmas, Classic, Energising and Bonding with friends. The
pattern matrix for these seven components, which represent seven brand
meanings for Coca-Cola, and the loadings of their items are presented in
table 6-6.

Table 6-6. The Pattern Matrix for Coca-Cola

Component
Bonding
Globally
Item Refreshing Unhealthy ChristmasClassic Energising with
recognised
friends
cooling .766 -.029 -.034 .052 .027 .085 -.163
refreshing .742 -.029 -.126 -.039 .104 .250 .019
cold .701 .011 .091 .041 -.027 .047 -.119
cool .620 .075 -.053 -.055 -.039 .133 -.244
freshness .574 -.023 -.177 -.148 .087 .329 .043
great brand .384 .274 -.259 -.114 -.007 -.008 -.072
familiar .381 .137 -.091 .180 .299 -.133 -.143
global -.185 .829 .118 .024 .086 .133 -.139
multinational .026 .805 -.034 .013 -.101 .032 -.152
worldwide -.149 .731 .002 .050 .265 .059 .049
recognisable .209 .522 .023 -.085 .120 -.096 .207
not unknown .196 .379 -.009 .030 -.113 -.165 .219
identifiable .350 .374 -.026 -.176 .117 -.184 .120
calories -.020 -.011 .790 -.046 -.007 -.082 -.235
unhealthy -.015 -.021 .785 -.019 .060 -.001 .103
obesity -.145 -.003 .782 .086 .015 .053 -.151
sugar .013 -.001 .763 -.136 .060 -.041 .078
sweet .339 .154 .519 -.142 -.159 .117 .117
not healthy .164 -.043 .393 .248 -.034 -.235 .295
not natural -.079 .119 .361 .076 -.315 -.017 .268
Christmas song -.035 .014 .080 -.911 -.054 -.056 .021

145
Santa Claus .011 -.053 -.022 -.868 .022 .061 .101
Christmas -.136 -.042 .078 -.866 .082 -.063 -.109
traditional -.034 -.037 .020 -.093 .833 .069 .065
classic .114 .077 -.073 -.001 .772 -.055 .026
American -.158 .241 .324 .010 .479 .121 .139
original .095 .234 .009 -.255 .410 -.028 -.166
drink it when tired -.035 .041 .022 .059 .027 .807 -.069
keeps me going .108 .132 -.195 .016 -.056 .708 -.059
no sleeping -.022 -.028 .121 -.030 -.054 .654 -.036
energetic .281 -.042 -.042 -.059 .139 .633 .191
bonding with friends -.004 .034 -.111 -.085 -.094 .166 -.739
parties .155 .330 .017 -.002 -.223 .067 -.592
hot day .402 -.118 .123 -.022 .057 .051 -.584
warm summer .334 -.155 .116 .001 .245 .061 -.578
fun .364 -.005 -.079 -.140 .105 .089 -.493
exciting .233 -.001 -.185 -.296 -.041 .193 -.400
Extraction Method: Principal Component Analysis; Rotation Method: Oblimin with Kaiser
Normalization; Rotation converged in 21 iterations.

It was interesting to investigate whether the same items formed the same
components in each of the studied countries. Therefore, reliability tests
were conducted and the results are presented in table 6-7. Almost all of the
Cronbach’s alphas were above the critical point (Į. > .70), indicating
consistency among the items, and further indicating that all the items
measure a single, unidimensional latent construct (Nunally 1978).

The Cronbach’s alphas lower than the recommended critical point (Į. <
.70), but still relatively high (Į. > .60) occurred in five cases as can be
read from table 6-7. For the further analysis, principal components were
used. Since they take into account loadings of items, the somewhat lower
Cronbach’s alphas did not cause a problem for the interpretation of the
results. Additionally, the Cronbach’s alphas were relatively similar for
each single component in all the countries, which reflected uniformity of
consumer logic in all the countries.

146
Table 6-7. The Reliability Tests for the Top Items in Coca-Cola’s
Components a

Bonding
Globally
Country Refreshing Unhealthy ChristmasClassic Energising with
recognised
friends
Overall .85 .74 .79 .86 .70 .73 .85
Austria .83 .62 .83 .81 .64 .73 .84
Finland .79 .78 .68 .91 .67 .77 .77
India .89 .78 .80 .86 .60 .70 .84
United States .87 .74 .74 .88 .72 .71 .77
a
Only items whose loadings were the highest on a particular component (in comparison to
their loadings on other components) were included in each component’s reliability tests
(Cronbach’s alphas).

As shown in table ap4-2, the total variance explained by the seven

components was 59.3%. The Pearson’s correlation matrix is presented in
table 6-8 for these components.

Table 6-8. The Pearson’s Correlation Matrix for Coca-Cola’s

Components

Bonding
Globally
Component Refreshing Unhealthy Christmas a Classic Energising with
recognised
friends a
Refreshing 1
Globally .247** 1
recognised
Unhealthy -.165** .086 1
Christmas -.169** -.095 .009 1
Classic .135** .179** -.044 -.160** 1
Energising .188** -.027 -.098 -.240** .076 1
Bonding -.254** .047 .177** .143** -.054 -.309** 1
with friends
Minimum -4.80 -5.15 -3.57 -2.07 -3.88 -2.50 -3.04
Maximum 2.11 1.47 1.96 1.93 2.52 2.85 2.97
K-S z (sig.) 1.29 (.07) 2.18 (.00) .70 (.72) 1.46 (.03) 1.28 .74 (.64) .80 (.54)
* p < .05 ** p < .01
a
Smaller values indicate a higher emphasis

The Kolmogorov-Smirnov tests of normality revealed that Globally

recognised (p = .00) and Christmas (p = .03) were not normally

147
distributed; hence Spearman’s correlation coefficients were also computed
for these components. All the Spearman’s coefficients were of the same
direction and of a similar magnitude to the Pearson’s coefficients.

Most of the components’ correlations were statistically significant, but

relatively weak, with the strongest correlation existing between
Energising and Bonding with friends (r = .31**).

The Brand Meanings for Google

The medians for each meaning were calculated to identify which ones are
most widely associated with Google. Table 6-9 provides a list of the initial
meanings with their median values sorted from the highest to the lowest
score on a 7-point scale. The items that best indicate what Google means
to the respondents were useful, easy to use, worldwide and global. These
items represented a combination of experiential, symbolic and functional
meanings. The country-wise situation was similar, except that the items
provides data in the United States and power in Finland were placed much
higher than in the overall sample.

Items that indicated what Google means to the respondents the least were
no independency, no privacy, life saver and happiness. They all
represented experiential meanings. Country-wise, the items: observes
people’s habits in Finland and the United States, new in India, and
freedom in Austria were much less indicative of Google than in the overall
sample.

In order to obtain data that was suitable for parametric statistical tests, the
principal component analysis was conducted on 42 initially measured
items. A total of 317 observations presented a ratio of items to
observations of 1:8. The Scree plot was examined for inflections (figure
ap4-2.). The first well-visible inflection appeared at the fourth point and

148
suggested three components, while the second inflection appeared at the
seventh point and suggested six components. In line with such results, the
analysis was initially performed for three components.

Table 6-9. The Medians for Google’s Meanings in the Overall Samplea

Item Median b Countries c Item Median b Countries d

useful 7 US, I, F, A reliable 6
easy to use 7 US, I, F, A creative 6
worldwide 7 I, F, A dominant 6
global 7 I, A satisfaction 6
convenient 7 US smart 6
helpful 7 encyclopaedia 6
fast 7 not old 6
at any time 7 entertainment 5
connecting with the world 7 young 5
provides data 7 US observes people's habits 5 US, F
multibillion industry 7 no problem 5
huge 7 fun 5
high speed 6 all knowing 5
world 6 new 5 I
informative 6 freedom 5 A
efficient 6 trustful 5 A
powerful 6 F almighty 5 I
knowledge 6 happiness 5 US, I, A
very important 6 life saver 5 A
innovative 6 no privacy 4 US, F
basic necessity 6 no independence 4 US, I, F
a
N = 317; all countries sample; data were sorted primarily on median and secondarily on mean.
b
Medians close to 1, show the items very poorly indicated, while those close to 7 show that
the items clearly indicated what each brand means to the respondents.
c
The country initial indicates a country in which the item was evaluated as one of the top
four.
d
The country initial indicates a country in which the item was evaluated as one of the bottom
four.

The obtained components could not be interpreted well, and as many as 11

communalities were less than .40. This was a sufficient condition for
adding an additional component (Costello and Osborne 2005). Hence, the
number of components was increased in each consecutive analysis until a

149
 nine-component solution was examined. Each solution was analysed for
the number of communalities, number of cross-loadings, number of non-
loading items, number of items per component, number of iterations and
the interpretability of the components as suggested by Velicer and Jackson
(1990) and Costello and Osborne (2005).

The best results were obtained for a seven component solution, however,
these components needed to be further improved. Hence, items with low
communalities and/or cross-loaded were excluded one by one and analysis
was repeated again for exclusion of each of these items for solutions of
five, six and seven components. The best results were obtained when
powerful, at any time, new and reliable items were excluded and a seven-
component extraction was performed (as shown in table 6-10). This
solution offered at least three strongly loaded items per component, no
non-loaded items, all but one communality higher than .40 (table ap4-4)
and only five cross-loaded items. The new components were named:
Almighty, Helpful, Intrusive, Inspired, Fun, Global and Informative.

Table 6-10. The Pattern Matrix for Google

Component
Item Almighty Helpful Intrusive Inspired Fun Global Informative
very important .713 -.213 .031 -.018 .096 -.035 .130
almighty .610 .140 .207 .091 -.032 -.153 .062
trustful .531 -.110 -.308 .143 -.135 .060 .093
all knowing .510 .115 -.029 .059 -.079 -.223 .326
basic necessity .502 -.408 .116 -.085 -.181 .064 -.020
life saver .479 -.083 -.060 .063 -.345 .035 .075
no problem .428 -.058 -.056 .104 -.230 .025 .239
dominant .420 -.009 .266 .318 .022 -.177 -.384
satisfaction .386 -.157 -.071 .305 -.187 -.033 .120
high speed .342 -.232 -.049 .305 .222 -.186 .073
useful .021 -.848 .035 -.248 -.102 -.036 .024
helpful .131 -.757 -.053 .019 -.016 -.050 .054
easy to use -.115 -.717 -.030 .059 -.096 -.179 -.055
fast .029 -.655 -.011 .128 .053 -.146 -.092

150
convenient .064 -.637 -.054 .050 .050 -.103 .192
provides data .001 -.350 -.005 .143 .087 -.234 .184
no privacy -.264 -.014 .781 -.073 -.017 -.057 .005
observes people's habits .080 .040 .738 .149 .109 -.014 .063
no independency .091 -.030 .717 .008 -.034 .111 .048
innovative .065 .043 .059 .784 -.091 .009 .053
creative .012 -.030 .058 .769 -.110 .041 .057
young -.030 .244 -.008 .704 -.301 -.157 -.014
smart .089 -.394 .065 .501 -.165 .135 .083
multibillion industry .069 -.150 .168 .448 .003 -.293 -.290
efficient .218 -.252 -.028 .357 -.026 .011 .203
entertainment -.103 -.096 .055 .080 -.766 -.035 .067
happiness .162 .031 -.077 .121 -.710 -.119 .053
fun -.054 -.092 -.057 .241 -.661 .033 .080
freedom .317 -.088 -.043 .260 -.432 -.025 -.053
not old -.229 -.255 -.144 .315 .373 .060 .209
global .085 .049 -.057 .094 .129 -.787 .057
worldwide -.056 -.207 .018 -.026 .011 -.778 -.022
world -.107 -.023 .025 -.054 -.064 -.768 .143
huge .220 -.146 -.010 -.098 -.170 -.547 -.199
connecting with the world -.144 -.230 .085 .025 -.264 -.499 .174
encyclopaedia .168 .010 .190 .015 -.193 -.028 .642
knowledge .178 -.057 -.020 .057 -.035 -.255 .578
informative .153 -.152 .024 .070 .058 -.288 .561
Extraction Method: Principal Component Analysis; Rotation Method: Oblimin with Kaiser
Normalization; Rotation converged in 22 iterations.

To investigate whether the same items formed the same components in

each of the studied countries, reliability tests were conducted. As the
results in table 6-11 show, the Cronbach’s alphas were above the critical
point of Į > .70 for five components in all the countries, which indicates
consistency among their top items. For the Fun and Intrusive components,
most of the Cronbach’s alphas were less than Į < .70.

The reliability problem for the Fun component, as discovered in

subsequent correlation tests, was caused by its item not old, which
distorted the results across the countries and mostly in Finland (Į. = .51).
The meaning Intrusive had two rather low alphas for India (Į = .42) and

151
the United States (Į = .44). This indicated that the top items of the Fun
component were not correlated.

Table 6-11. The Reliability Tests for the Top Items in Google’s
Components a
Country Almighty Helpful Intrusive Inspired Fun Global Informative
Overall .86 .85 .63 .83 .73 .81 .72
Austria .82 .86 .65 .82 .67 .82 .63
Finland .80 .84 .77 .74 .51 .82 .63
India . 84 .83 .42 .80 .78 .76 .80
United States .87 .83 .44 .81 .66 .82 .79
a
Only items whose loadings were the highest on a particular component (in comparison to
their loadings on other components) were included in each component’s reliability tests
(Cronbach’s alphas).

The per country correlation tests among these top items indicated
statistically significant correlations (p = .00) only between: no privacy and
no independency in India, and no privacy and observes people's habits in
the United States. These results implied that the respondents in the
different countries had different mind maps and did not associate the same
meanings to all the items within this component. The results for the
meaning Intrusive, thus, had to be interpreted with caution.

Table 6-12. The Pearson’s Correlation Matrix for Google’s

Components

Component Almighty Helpful a Intrusive Inspired Fun a Global a Informative

Almighty 1
Helpful -.252** 1
Intrusive .088 .073 1
Inspired .332** -.275** .061 1
Fun -.342** .109 -.083 -.232** 1
Global -.251** .401** -.157** -.263** .164** 1
Informative .206** -.278** -.118* .218** -.117* -.159** 1
Minimum -3.85 -1.31 -3.25 -3.54 -2.21 -1.59 -3.79
Maximum 1.89 8.95 2.90 2.06 2.90 4.63 2.24
K-S z (sig.) 1.14 (.15) 2.52 (.00) .52 (.95) 1.06 (.21) .78 (.57) 2.00 (.00) 1.30 (.07)
* p < .05 ** p < .01
a
Smaller values indicate a higher emphasis

152
The total variance explained by seven components was 60.2%, as shown
in table ap4-5. The Pearson’s correlation matrix for Google’s Components
is presented in table 6-12. The Kolmogorov-Smirnov tests of normality
revealed that the Helpful and Global components were not normally
distributed (p = .00); hence, the Spearman’s correlation coefficients were
also computed for these components. All of the Spearman’s coefficients
were of the same direction and generally of a slightly smaller magnitude
than the Pearson’s ones. Most of the components were statistically
significantly, but rather weakly correlated. The strongest correlation
existed between the Almighty and Fun components (r = .34**).

The Brand Meanings for Nike

To uncover which meanings were the most associated to Nike, the

medians for each meaning were calculated and sorted from the highest to
the lowest in table 6-13.

The items that the best indicated what Nike means to the respondents were
sports and just do it, sporty, athletes and athletic, almost all representing
functional meanings. The items that indicated the least what Nike means
to the respondents were crowd cheering, child labour, Tiger Woods and
life. Most of them had symbolic meanings. Overall, the results were
similar in all the countries, except for the Finnish respondents’ low
emphasis of victory and winning, and the Indian respondents’ high
emphasis of expensive.

In order to reduce the data for Nike and obtain normalised data for further
analysis, a principal component analysis was conducted on the 37 initially
measured items. A total of 306 respondents evaluated Nike. This
generated a ratio of items to observations of 1:8. The Scree plot was
examined (figure ap4-3). Its two inflection points were at the fifth and
seventh point. A four-component solution was rejected as the

153
communalities of five items were less than .40. The five-component
solution was better; that is, two communalities were less than .40, but
there were nine cross-loadings and the interpretability of components was
not optimal. In order to improve the model, the item not low quality was
excluded.

Table 6-13. The medians for Nike’s Meanings in the Overall Sample a

Item Median b Countries c Item Median b Countries d

sports 7 US, I, F, A trendy 5
just do it 7 A achiever 5
sporty 6 US, I, F, A powerful 5
athletes 6 US, I, F innovative 5
athletic 6 A, F victory 5 F
running 6 A durable 5
fitness 6 winning 5 F
jogging 6 innovation 5
expensive 6 I not socially responsible 5 I
not low quality 6 young 5
good quality 6 sweat shop 5 I
not weak 6 self-assured 5 US
pricey 6 positive thinking 5 A
Michael Jordan 6 freedom 5 US, A
comfortable 5 life 5 US
not failure 5 Tiger Woods 4 I, F, A
not lazy 5 child labour 4 I
cool 5 crowd cheering 4 US, F, A
air technology 5
a
N = 306; all countries sample; data are sorted primarily on median and secondarily on mean.
b
Medians close to 1, show the items very poorly indicate, while those close to 7 show that
the items clearly indicate what each brand means to the respondents.
c
The country initial indicates a country in which item was evaluated as on of the top four.
d
The country initial indicates a country in which item was evaluated as one of the bottom
four.

154
Table 6-14. The Pattern Matrix for Nike

Component
Not socially
Item Achiever Not weak Pricey Sports
responsible
winning .779 .076 .075 -.043 -.084
positive thinking .773 .079 -.129 .134 -.018
victory .749 .083 .036 -.064 -.131
innovation .706 -.006 -.224 -.092 -.037
self-assured .704 -.005 -.026 -.030 -.021
crowd cheering .700 -.098 .103 -.097 -.013
life .672 -.127 -.099 .042 -.003
innovative .668 -.050 -.249 -.070 -.101
Tiger Woods .661 .117 .278 .066 .127
freedom .642 -.063 -.234 .067 -.125
powerful .621 -.018 .015 -.109 -.119
achiever .609 .053 -.029 -.051 -.167
air technology .601 -.086 .115 .143 -.139
cool .584 -.013 -.233 -.056 -.209
Michael Jordan .567 .080 .495 -.072 .039
young .545 .015 -.128 -.196 .021
trendy .450 -.074 -.256 -.315 -.129
just do it .445 .167 .248 .041 -.193
durable .439 .019 -.269 -.239 -.148
not failure .089 .782 -.164 .056 .031
not weak -.011 .743 -.042 -.076 -.089
not lazy -.074 .728 -.045 .062 -.055
child labour -.086 -.246 .743 -.042 -.126
sweat shop .008 -.284 .676 -.221 -.106
not socially responsible -.392 .318 .492 -.007 -.037
good quality .128 .261 -.376 -.322 -.261
expensive -.052 -.003 .112 -.922 .039
pricey -.009 -.041 .107 -.911 .061
running -.046 -.072 .021 .049 -.905
sporty .048 .075 -.025 .009 -.814
jogging -.041 -.154 .036 .026 -.791
fitness .166 .027 .068 .068 -.767
athletic .187 .179 .027 .006 -.633
athletes .141 .165 .051 -.090 -.560
sports -.085 .246 .076 -.121 -.522
comfortable .286 -.023 -.264 -.094 -.362
Extraction Method: Principal Component Analysis; Rotation Method: Oblimin with Kaiser
Normalization; Rotation converged in 12 iterations.

155
A new Scree plot was constructed. It also revealed two inflection points,
which proposed four and six components. Principal component analyses
were repeated for four, five and six components, and the five-component
solution was deemed the best. This solution’s communalities were almost
all above .40 (as presented in table ap4-7); there were only three cross-
loaded items and no components with fewer than three items (as seen in
table 6-14). The obtained components were named: Achiever, Not weak,
Not socially responsible, Pricey and Sports.

Table 6-15. The Reliability Tests for the Top Items in Nike’s
Components a

Not socially
Country Achiever Not weak Pricey Sports
responsible
Overall .94 .73 .67 .87 .88
Austria .93 .72 .65 .87 .88
Finland .93 .78 .80 .82 .83
India .94 .77 .07 .75 .90
United States .94 .58 .77 .91 .91
a
Only items whose loadings were the highest on a particular component (in comparison to
their loadings on other components) were included in each component’s reliability tests
s(Cronbach’s alphas).

There was a problem in terms of consistency for the Not socially

responsible component for the Indian data (table 6-15), for which the
Cronbach’s alpha reached as low as .07. Since further analysis was
conducted with normalised components, the low Cronbach’s alpha of the
Not socially responsible component did not pose a major problem;
nevertheless, more caution was necessary when interpreting the results for
this meaning in India.

The total variance explained with the five components was 57.8%, as
visible in table ap4-8. Table 6-16 presents the Pearson’s correlation matrix
for Nike’s Components. The Kolmogorov-Smirnov tests of normality
revealed that only Sports (p = .01) was not normally distributed, hence

156
Spearman’s correlation coefficients were also computed for this
component. These coefficients were of the same direction and a similar
magnitude to Pearson’s coefficients, hence Pearson’s were reported. Most
of the components were statistically significantly, but weakly correlated (r
< .30**), while only the Achiever and Sports components were moderately
correlated to each other (rs = .50**).

Table 6-16. The Pearson’s Correlation Matrix for Nike’s Components

Not socially
Component Achiever Not weak Pricey a Sports a
responsible
Achiever 1
Not weak .057 1
Not socially
-.130* -.034 1
responsible
Pricey -.223** .015 .021 1
Sports -.474** -.214** .049 .272** 1
Minimum -3.05 -3.83 -2.70 -1.77 -1.45
Maximum 2.16 2.18 2.53 2.87 3.92
K-S z (sig.) .45 (.99) .56 (.92) .62 (.83) 1.26 (.08) 1.69 (.01)
* p < .05 ** p < .01
a
Smaller values indicate higher emphasis

6.2.3.3. Types of Users

Apart from varying along cultural dimensions, the respondents in this

research also differed on whether they use and whether they like a given
brand. Although usage and level of affection were correlated variables
(rcocacola = .509, p < .01; rgoogle = .144, p < .01; rnike = .509, p < .01), the
correlation was not very strong. Usually, there are consumers who do not
like a brand, but they still use it (normally when there is no better
alternative). On the other side, there are consumers who like a brand but
they do not use it (normally when the brand in question is unavailable or
unaffordable) (cf. Griffin 1997). The distribution of the responses for the
different affection levels and users vs. non-users in this research is
displayed in table 6-17.

157
In order to diminish the number of groups in a logical manner so that
between group tests could be conducted, three subgroups were identified
for Nike and Coca-Cola, and two for Google. More precisely, in the case
of Nike and Coca-Cola, all their non-users comprised one group. There
were 105 such respondents for Nike and 74 for Coca-Cola. The users of
Nike and Coca-Cola were further divided into two subgroups: users who
like the brand, named Engaged users, and users who do not like or are
indifferent to the brand, named Disengaged users. Hence, there were 135
Engaged Nike users and 66 Disengaged Nike users; as well as 192
Engaged Coca-Cola users and 104 Disengaged Coca-Cola users.

Table 6-17. The Distribution of the Respondents based on their Usage

and Feelings for the Brands a
Feeling
Type of Indifferent/
Brand Love it Like it Not like it Hate it Total
user mixed feelings

Users 291 1061 612 42 12 201

Nike Non-users 23 153 633 223 33 105
Total 31 121 124 26 4 306

Users 1354 1595 215 0 0 315

Google Non-users 0 15 15 0 15 3
Total 135 160 22 0 1 318

Users 366 1566 817 207 37 296

Coca -cola Non-users 0 68 308 288 108 74
Total 36 162 111 48 13 370
a
The Indices next to the number of observations denominate the new subgroup that the
respondents belonged to.

Because the responses were collected online, it was not surprising that
almost all the respondents use Google and have very positive feelings
towards it. Hence, only two groups of users were created for Google. The
135 representatives who use and love it formed the group of Engaged

158
Google users, while all the other 183 respondents formed the group named
Disengaged Google users.

6.2.3.4. The Brand Meanings across the Cultures and across the Types
of Users

The tests of hypotheses 1-3 are presented in this chapter. For each brand’s
meanings, a separate two-way ANOVA was conducted. Because there are
19 dependent and two independent variables in total, a short comment on
each result follows immediately after the data are presented.

Differences in Coca-Cola’s Brand Meanings

In order to determine whether differences in Coca-Cola’s meanings

existed among the different countries and different types of users, a set of
4 (countries) x 3 (type of Coca-Cola user) two-way ANOVAs were
conducted. Each of the seven previously identified components
(henceforth meanings) represented a dependent variable in a separate two-
way ANOVA. Variances in three of the seven meanings (Refreshing,
Globally recognised and Classic) across groups were not homogeneous
(ap4-10), so non-parametric tests were applied. Country was a nominal
independent variable which differentiated between four groups of
respondents. Type of Coca-Cola user represented an ordinal numerical
variable with three levels and served as the second independent variable.

The results of the ANOVA tests for the independent variable, Country, are
displayed in table 6-18. Country had a statistically significant effect on the
meanings: Refreshing (F (3, 354) = 3.69; p < .05), Globally recognised (F
(3, 354) = 5.03; p < .01) Unhealthy (F (3, 354) = 8.71, p < .01), Christmas
(F (3, 354) = 9.17; p < .01), Classic (F (3, 354) = 30.67; p < .01) and
Bonding with friends (F (3, 354) = 30.35; p < .01). Size of effect was
considered weak for Refreshing (partial Ș2 = .03) and Globally recognised

159
 (partial Ș2 = .04), moderate for Unhealthy (partial Ș2 = .07) and Christmas
(partial Ș2 = .07), and strong for Classic (partial Ș2 = .21) and Bonding
with friends (partial Ș2 = .21), which indicated that between 3% and 21%
of the variance in the meanings was explained by the variable Country.

Table 6-18. The Effects of Culture on the Coca-Cola Meanings a

Brand Austria Finland India United States F Partial
Meaning (1) (2) (3) (4) (3, 354) Ș2
Refreshing b -.287 (1.088)3 .039 (.800)3 .353 (.910)1 2 -.036 (1.113) 3.69* .030
Globally
recognised c .161 (.892)4 .119 (.914) -.179 (1.155) -.239 (1.031)1 5.03** .041
Unhealthy .359 (1.063)2,3,4 .033 (.783)1,3 -369 (1.042)1,2 -.187 (.950)1 8.71** .069
Christmas d -.400 (.958)2,3,4 .190 (.979)1 .261 (.965)1 .033 (.958)1 9.17** .072
Classic -.346 (.932)2,4 .551 (.736)1,3 -.631 (1.000)2,4 .526 (.707)1,3 30.67** .206
Energising .033 (1.070) -.173 (.938) .257 (1.002) -.111 (.914) 2.17 .018
Bonding with
friends d .324 (.942)3 .410 (.789)3,4 -.968 (.745)1,2,4 .045 (.865)2,3 30.35** .205
a
The values outside the brackets represent the countries’ means, whereas the values in the
brackets represent the corresponding standard deviations. The indices indicate which groups
are significantly different in statistical terms (Bonferroni post-hoc test).
b
The difference between Finland and India were captured in a non-parametric Mann-Whitney
test with a manual Bonferroni correction (table ap4-12).
c
The difference between Austria and the United States were captured in a non-parametric
Mann-Whitney test with a manual Bonferroni correction (ap4-15).
d
The negative values represent high emphasis of the meaning.
* p < .05; ** p < .01

The post-hoc tests for the Refreshing dependent variable revealed a

statistically significant difference in the emphasis of the Refreshing
meaning of Coca-Cola, where the Indians (M = .35; SD = .91) emphasised
it more than the Finns (M = .04; SD = .80) and the Austrians (M = -.29;
SD = 1.09) did. On average, the respondents from the United States (M = -
.04; SD = 1.11) emphasised the Refreshing meaning slightly less that the
Finns did, but its emphasis was not statistically significantly different
from any other country.

The Refreshing meaning was not supposed to be influenced by any of the

cultural dimensions. Hence, it was not supposed to be differently

160
emphasised among the countries in this research. However, countries
differ not only in cultural terms but also due to other environmental forces
such as economy, politics, demographics, legal systems, and the like.
Although the brands were chosen in a way that the effects of other
environmental forces were controlled for, in this case climate differences
might have influenced the results. India is by far the country with the
warmest climate among the studied countries, which might explain why
the Refreshing meaning of Coca-Cola was strongly emphasised in India
and weakly in cold Austria and Finland.

The post-hoc tests for the Unhealthy dependent variable revealed that the
Austrians (M = .36; SD = 1.06) emphasised the Unhealthy meaning
significantly more than the respondents in all the other studied countries.
Then followed the Finnish (M = .03; SD = .78) and the American (M = -
.19; SD = .95) respondents, who did not differ in their emphasis of the
Unhealthy meaning. Finally the Indian respondents (M = -.37; SD = 1.04)
gave it a very low score which was statistically different to the highest
Austrian score and the second highest Finnish score.

The Unhealthy meaning was predicted to be affected by performance

orientation and assertiveness in terms that cultures low in these two
cultural dimensions stress life quality (House et al. 2004, p. 405). Since
healthy food is an aspect of life quality, it was predicted that cultures low
in assertiveness and performance orientation would put higher emphasis
on the meaning of Unhealthy. In addition, uncertainty avoidance was also
predicted to affect this meaning, as cultures high in uncertainty avoidance
tend not to expose themselves to risks (House et al. 2004, p. 618) and
would thus consider Coca-Cola unhealthier. Since the scores of the four
countries were very similar in terms of performance orientation, while the
opposite in terms of assertiveness and uncertainty avoidance, the effects
were neutralised to some extent. Nevertheless, because the countries’

161
assertiveness scores vary more than their scores on uncertainty avoidance,
the effect of assertiveness was assumed to be stronger. Overall, it can be
concluded that results were as predicted. The country that values
assertiveness the most, i.e. India (AI = 4.76), had the lowest score on the
Unhealthy meaning emphasis, while the one that values assertiveness the
least, i.e. Austria (AA = 2.81), had the highest emphasis of the Unhealthy
meaning.

The post-hoc tests for the Christmas dependent variable revealed that the
Austrians (M = -.40; SD = .96) emphasised the Christmas meaning of
Coca-Cola significantly more than all other studied countries: Finland (M
= .19; SD = .98), India (M = .26; SD = .97), and the United States (M =
.03; SD = .96). The differences among the latter three were not
statistically significant.

Christmas was in general predicted to be more emphasised primarily by

countries that are Christian. India is a non-Christian country, so its
emphasis of Christmas was overall the lowest, although statistically lower
only from the Austrian highest emphasis. Among Christian countries, it
was very difficult to predict which country would have the highest
emphasis of this meaning. In other words, the genuine value of Christmas
was predicted to be the meaning affected most by the human orientation
cultural dimension. This is because cultures high in human orientation
consider others (family, friends, strangers etc.) important and value
altruism, kindness, love and generosity (House et al. 2004, p. 570), which
are all values associated with Christmas. However, for some respondents
(and especially those high in human orientation), connecting Coca-Cola to
Christmas might have symbolised excessive consumerism – a
phenomenon contrary to what is propagated by Christmas. So, to express
their non-agreement with such a view of Christmas, such respondents

162
might have given the Christmas meaning of Coca-Cola a low score. For
example, this view was discovered in the in-depth interview with
Christina from Austria.

Overall, the results for Christmas were partially as expected. The countries
that value human orientation the most are Finland (HOF = 5.81) and
Austria (HOA = 5.76). The Austrian high score on Christmas emphasis
was aligned to the prediction, but the Finnish score was supposed to be
statistically significantly higher than the American score (HOUS = 5.53),
and it was not.

The post-hoc tests for the Classic meaning revealed that the Indians (M =
-.63; SD = 1.0) and the Austrians (M = -.35; SD = .93) emphasised the
Classic meaning of Coca-Cola statistically significantly less than the
Finns (M = .55; SD = .74) and the Americans (M = .53; SD = .71) did,
while the former two and the latter two showed no statistically significant
difference between each other.

The Classic meaning was supposed to be affected by assertiveness and

uncertainty avoidance. Highly assertive cultures value tradition, seniority
and experience (House et al. 2004, p. 405). Hence, they were predicted to
emphasise the Classic meaning of Coca-Cola. At the same time,
Uncertainty avoiding cultures show stronger resistance to change (House
et al. 2004, p. 618), so it was supposed they would also emphasise the
Classic meaning of Coca-Cola. These two cultural dimensions have the
opposite scores in the four countries, with India and Austria being at the
extreme ends on both of the dimensions. In conclusion, the emphasis of
the Classic meaning was only partly affected as predicted. In India (UAI =
4.73; AI = 4.76), only uncertainty avoidance seemed to have affected it,
and in Austria (UAA = 3.66; AA = 2.81) only assertiveness.

163
For the Bonding with friends meaning, the post-hoc tests uncovered that
the Finns (M = .41; SD = .79) emphasised this meaning the least and
statistically significantly less than the Indians (M = -.97; SD = .75) and the
Americans (M = .05; SD = .87). The Austrian second lowest place in the
strength of this meaning emphasis (M = .32; SD = .94) was significantly
different in statistical terms from the high Indian score; whereas the
American middle position was statistically significantly different from the
lowest Finnish and the highest Indian score.

The Bonding with friends meaning was predicted to be affected by in-

group collectivism since cultures high in this dimension demonstrate
relatedness with members of groups (House et al. 2004, p. 471). The
results for Austria, Finland and the United States were in line with the
predictions (CA = 5.27; CF = 5.42; CUS = 5.77); however, the very high
Indian score on Bonding with friends could not be explained with their
relatively low score on valuing in-group collectivism (CI = 5.32).
However, the Indian score could again be ascribed to its warm climate. In
other words, the meaning Bonding with friends included the items bonding
with friends, fun and parties, but also warm summer and hot day (cf. table
6-6), which might have introduced bias to the results.

The post-hoc tests for the Globally recognised dependent variable

revealed that the Americans (M = -.24; SD = 1.03) emphasised the
Globally recognised meaning less than the Austrians (M = .16; SD = .89)
did and that the difference in emphasis was statistically significant. On
average, the responses of the Finns (M = .12; SD = .91) and the Indians
(M = -.18; SD = 1.16) were situated between those of the Americans and
the Austrians, and were not statistically significantly different from the
former two.

164
The Globally recognised meaning was supposed to be correlated to the
cultural dimension of uncertainty avoidance. This is because cultures that
malfunction in uncertain or unspecified situations (House et al. 2004, p.
618) were predicted to appreciate Coca-Cola for being recognised
worldwide and standardised, which diminishes the choice risk. The results
were not as predicted. Due to its low uncertainty avoidance score (UAA =
3.66), Austria was supposed to have the lowest score on emphasising
Globally recognised meaning, whereas the United States (UAUS = 4.00)
and India (UAI = 4.73), being high uncertainty avoiding cultures, were
supposed to have the highest scores on this meaning.

Finally, the difference in the emphasis of the meaning Energising, as

predicted, was not statistically significant among the countries.

The results of the ANOVA tests for the independent variable, Type of
Coca-Cola user, were displayed in table 6-19. Type of Coca-Cola user
had a statistically significant effect on the following meanings: Refreshing
(F (2, 354) = 24.58; p < .01), Unhealthy (F (2, 354) = 28.20; p < .01),
Christmas (F (2, 354) = 4.83; p < .01), Classic (F (2, 354) = 6.97; p < .01),
and Energising (F (2, 354) = 10.63; p < .01). Sizes of the effects were
considered weak for the meanings Christmas (partial Ș2 = .03), Classic
(partial Ș2 = .04) and Energising (partial Ș2 = .06), and moderate for
Refreshing (partial Ș2 = .12), and Unhealthy (partial Ș2 = .14). These effect
sizes indicated that between 3% and 14% of the variance in the meanings
was explained by the variable Type of Coca-Cola user.

The post-hoc tests for the Refreshing dependent variable revealed that the
Engaged users emphasised Refreshing meaning statistically significantly
more (M = .38; SD = .77) than the Disengaged users (M = -.32; SD = .99)
and the Non-users (M = -.52; SD = 1.15) did; whereas the latter two
subgroups did not exhibit a statistically significant difference between

165
each other. Likewise, the post-hoc tests for the Energising dependent
variable revealed that the Engaged users (M = .20; SD = .93) and the
Disengaged users (M = -.02; SD = 1.06) emphasised the Energising
meaning statistically significantly more than the Non-users (M = -.49; SD
= .92); whereas the former two subgroups did not differ statistically to
each other.

Table 6-19. The Effects of Type of Coca-Cola User on the Coca-Cola

Meanings a
Engaged Disengaged
Non-users F Partial
Brand Meaning users users
(3) (2, 354) Ș2
(1) (2)
Refreshing .378 (.768)2 3 -.324 (.994)1 -.523 (1.145)1 24.58** .121
Globally recognised .088 (.898) -.089 (1.135) -.104 (1.043) 2.96 .016
Unhealthy -.401 (.946)2,3 .244 (.829)1,3 .693 (.863)1,2 28.20** .137
Christmas b -.138 (.954)3 .128 (1.013) .178 (1.058)1 4.83** .027
Classic .148 (.903)2 -.306 (1.087)1 .041 ( 1.031) 6.97** .038
Energising .202 (.929)3 -.020 (1.064)3 -.490 (.921)1,2 10.63** .057
Bonding with friends b -164 (1.006) .119 (.953) .256 (.982) 2.98 .017
a
The values outside the brackets represent the countries’ means, whereas the values in the
brackets represent the corresponding standard deviations. The indices indicate which groups
are significantly different in statistical terms (Bonferroni post-hoc test).
b
The negative values represent high emphasis of the meaning.
* p < .05; ** p < .01

Furthermore, Classic is a meaning for which the post-hoc tests revealed

that the Engaged users (M = .15; SD = .90) had a statistically significantly
higher result than the Disengaged users (M = .31; SD = 1.08). The Non-
users (M = .04; SD = 1.03) did not significantly differ from either of the
former two types of users. Finally, the results obtained in the post-hoc
tests for the Christmas dependent variable revealed that the Engaged users
emphasised the Christmas meaning statistically significantly more (M= -
.14; SD = .95) than the Non-users (M = .18; SD = 1.06) did; whereas the
Disengaged users (M = .13; SD = 1.01) occupied a place between the first
two groups and showed no statistical difference from either.

166
On the other hand, the post-hoc tests for the Unhealthy dependent variable
revealed that the Engaged Coca-Cola users emphasised the Unhealthy
meaning statistically significantly less (M = -.40; SD = .95) than the
Disengaged users (M = .24; SD = .83) and the Non-users (M = .69; SD =
.86) did. In addition, there was also a statistically significant difference
between the Disengaged users and the Non-users.

The elaborated five results all had a predicted direction. The positive
meanings were more emphasised by those respondents who favour Coca-
Cola, whereas the negative meanings were more emphasised by those who
do not favour Coca-Cola. However, Bonding with friends and Globally
recognised were supposed to be the meanings that were stronger
emphasised by the Engaged users than by the others, but the expected
difference was not obtained in the post-hoc tests.

To measure the interaction effects, (i.e. Country and Type of Coca-Cola

user) the sample was divided into twelve subgroups. The interaction
effects were not statistically significant for any of the dependent variables:
Refreshing (p = .31), Globally recognised (p = .53), Unhealthy (p = .06),
Christmas (p = .35), Classic (p = .23), Energising (p = .20) and Bonding
with friends (p = .35).

Differences in Google’s Brand Meanings

In order to determine whether differences in Google’s meanings existed

between the different countries and different types of Google users, 4
(countries) x 2 (type of Google user) two-way ANOVAs were conducted.
The seven previously identified meanings were measured in continuous
numerical values and each one represented a dependent variable in a
separate ANOVA. The variances in the five meanings (i.e. Helpful,
Intrusive, Inspired, Global, and Informative) were not homogeneous
across the groups, so non-parametric tests were also conducted for these

167
variables (ap4-20 to ap4.-33.). The non-parametric results did not differ
from the parametric ones for any of the dependent variables, so the
parametric tests results were reported. Country was a nominal independent
variable with four levels, while Type of Google user was an ordinal
numerical variable with two levels.

The results of the ANOVA tests for the independent variable, Country, are
displayed in table 6-20. Country had a statistically significant effect on the
following meanings: Almighty (F (3, 309) = 10.76; p < .01), Intrusive (F
(3, 309) = 3.67; p < .05), Inspired (F (3, 309) = 9.15, p < .01), Fun (F (3,
309) = 17.18, p < .01), Global (F (3, 309) = 4.83, p < .01) and Informative
(F (3, 309) = 5.43, p < .01). The effect size was considered weak for
Intrusive (partial Ș2 = .03), Global (partial Ș2 = .05), and Informative
(partial Ș2 = .05); moderate for Almighty (partial Ș2 = .10) and Inspired
(partial Ș2 = .08); and strong for Fun (partial Ș2 = .14). This indicated that
between 3% and 14% of the variance in the meanings was explained by
the variable Country.

The post-hoc tests for the Almighty dependent variable revealed that the
respondents from India (M = .61; SD = .74) and the United States (M =
.37; SD = .88) emphasised the Almighty meaning more than the Austrian
(M = -.46; SD = 1.01) and the Finnish respondents (M = -.20; SD = .92)
and that the difference was statistically significant. Neither the former
two, nor the latter two countries differed from one another in statistical
terms.

It was hypothesised that the Almighty meaning would be affected by

power distance. This is because high power-distance countries that view
power as providing social order, relational harmony and role stability
(House et al. 2004 p. 540) put more emphasis on this meaning. The results
were in line with the Indian (PDI = 2.64) and American (PDUS = 2.85)

168
 higher scores on power distance than the Austrian (UAA = 2.44) and
Finish (UAF = 2.19) lower scores on this dimension.

Table 6-20. The Effects of Culture on the Google Meanings a

Brand Austria Finland India United States F Partial

Meaning (1) (2) (3) (4) (3, 309) Ș2
Almighty -.462 (1.011)3 4 -.196 (.916)3 4 .612 (.743)1 2 .370 (.876)1 2 10.76** .095
Helpful b .188 (1.293) -.022 (.942) -.066 (.755) -.229 (.652) .596 .006
Intrusive .247 (1.032)2,4 -.156 (1.078)1 .135 (.837)4 -.392 (.852)1,3 3.674* .034
Inspired -.350 (1.080)3,4 -320 (.877)3,4 .475 (.695)1,2 .556 (.896)1,2 9.15** .082
Fun b .455 (1.007)3,4 .220 (.849)3,4 -.703 (.867)1,2,4 -.272 (772)1,2,3 17.18** .143
Global b -.199 (.986)2 .328 (1.082)1,3 -.229 (.785)2 .139 (.995) 4.829** .045
Informative .083 (1.042)2 -.534 (1.048)1,3,4 .345 (.681)2 .273 (.847)2 5.43** .050
a
The values outside the brackets represent the countries’ means, whereas the values in the
brackets represent the corresponding standard deviations. The indices indicate which groups
are significantly different in statistical terms (Bonferroni post-hoc test).
b
The negative values represent a high emphasis of the meaning.
* p < .05; ** p < .01

The Post-hoc tests for the Intrusive dependent variable revealed that the
Austrians (M = .25; SD = 1.03) emphasised the Intrusive meaning
statistically significantly more than the Finns (M = -.16; SD = 1.08) and
Americans (M = -.39; SD = .85) did. The Indians’ (M = .14; SD = .84)
emphasis of the Intrusive Google meaning was the second highest and
statistically significantly different from the Americans’ lowest emphasis.

It was predicted that Intrusive would be influenced by in-group

collectivism, because individuals in cultures low in in-group collectivism
tend to look after themselves and view themselves as autonomous and
independent of groups (House et al 2004, p. 454). Such cultures were thus
also expected to consider Google’s private data gathering as a privacy
infringement. The results for Intrusive were in line with the Austrian (CA
= 5.27) lowest score, the Indian second lowest score; (CI = 5.32), and the
American highest score (CUS = 5.77) on the cultural dimension of in-group
collectivism.

169
Similarly, the post-hoc tests for the Inspired meaning revealed that the
Indians (M= .48; SD = .70) and the Americans (M = .56; SD = .90)
emphasised the Inspired meaning statistically significantly more than the
Austrians (M= -.35; SD = 1.08) and the Finns (M = -.32; SD = .88), while
the upper and lower subgroups did not differ statistically significantly
between each other.

The Inspired meaning was predicted to be affected by uncertainty

avoidance, assertiveness and performance orientation. The four studied
countries all have almost the same results on performance orientation, so
the influence of performance orientation had no influence on the results.
Cultures low in uncertainty avoidance facilitate new product development,
show less resistance to change and have a higher tolerance for breaking
rules (House et al. 2004, p. 618); while those high in assertiveness value
success, progress, taking initiative, and have a “can do” attitude (House et
al. 2004, p. 405). Thus, countries low in uncertainty avoidance and high in
assertiveness were both predicted to put a higher emphasis on Inspired
meaning. Since the studied countries have a reverse score on these values,
and since assertiveness has a stronger variance, it is also predicted to have
a stronger effect (as already mentioned for the Unhealthy meaning of
Coca-Cola). It can be claimed that the results for the Inspired meaning
were in line with the Indian high score (AI = 4.76), the Finnish (AF = 3.68)
and the United States (AUS = 4.32) moderate scores, and the Austrian (AA
= 2.81) low score on assertiveness.

According to the results of the post-hoc tests for the Fun meaning, the
Austrians (M = .46; SD = 1.00) emphasised the Fun meaning of Google
statistically significantly less than the secondly positioned United States
(M = -.27; SD = .77) and firstly positioned India (M = -.70; SD = .87).
The third lowest emphasis of Fun displayed by Finland (M = .22; SD =
.85) is also statistically significantly different from the score of the United

170
States and India. Finally, a difference in the emphasis of Fun was also
statistically significant between the United States and India.

Fun was predicted to be influenced by human orientation because cultures

that are less human oriented tend to value pleasure, comfort, and
enjoyment more (House et al. 2004, p. 570). The results for Fun were thus
in line with the prediction. In other words, India values human orientation
(HOI = 5.28) the least, followed by the United States (HOUS = 5.53),
which was also the order of the scores for the Fun meaning.

The post-hoc tests’ results for Global showed that the Finnish respondents
(M = .33; SD = 1.08) emphasised the Global meaning of Google
statistically significantly less than the Indian (M = -.23; SD = .79) and
Austrian respondents (M = -.20; SD = .99) did. The American (M = .14;
SD = 1.00) emphasis was slightly (not significantly) higher than the
Finnish one.

The Global meaning was predicted to be affected by the cultural

dimension of in-group collectivism, according to which cultures that have
a relatively higher appreciation of in-group collectivism emphasise global
relatedness and connectedness among members of the (Google) society
(House et al. 2004, p. 454). However, the results were opposite from what
was expected as Austria and India put the highest emphasis on the
meaning Global, and at the same time value in-group collectivism the
least (CI = 5.32; CA = 5.27).

The results of the post-hoc tests for the Informative dependent variable
revealed that the Informative meaning was statistically significantly less
emphasised in Finland (M = -.53; SD = 1.05) than in the other three
countries, i.e. Austria (M = .08; SD = 1.04), the United States (M = .27;
SD = .85), and India (M = .35; SD = .68). The latter three countries did
not differ in their emphasis of the meaning Informative.

171
Informative was predicted to be affected by the uncertainty avoidance
cultural dimension. In other words, cultures which greatly value
uncertainty avoidance, appreciate order, formalised interactions, and rules
also tend to keep experiences documented (House et al 2004, p. 618).
Hence, it was also expected that such cultures would highly appreciate the
Informative meaning. The results were partly confirmed. Since Austria
(UAA = 3.66) has the lowest and Finland (UAF = 3.85) the second lowest
score for uncertainty avoidance, the Austrian score for Informative should
have also been statistically significantly lower than the scores of the
United States and India.

Finally, there was no statistically significantly difference among the

countries in their emphasis of the Helpful meaning. This result was the
opposite from what was expected, since the Helpful meaning was
supposed to be affected by low assertiveness and high human orientation.

The results of the ANOVA tests for the independent variable, Type of
Google user, are displayed in table 6-21. The Type of Google user had a
statistically significant effect on the meanings: Almighty (F (1, 309) =
30.21; p < .01), Helpful (F (1, 309) = 15.3; p < .01), Inspired (F (1, 309) =
20.3; p < .01), Fun (F (1, 309) = 6.36; p < .05), Global (F (1, 309) =
25.41; p < .01), and Informative (F (1, 309) = 15.0, p < .01). The emphasis
of all the positive meanings (Almighty, Helpful, Inspired, Fun,
Informative, and Global) was, as expected, higher among the Engaged
users, as opposed to the Disengaged users.

Intrusive was the only identified and explored negative meaning of

Google. The difference in emphasis of Intrusive between the Engaged
users and the Disengaged users was not statistically significant. This was
not in line with the prediction that the Disengaged users would emphasise

172
the Intrusive meaning statistically significantly more than the Engaged
users.

The effect size was considered small for Helpful (partial Ș2 = .05), Fun
(partial Ș2 = .02), and Informative (partial Ș2 = .05), while it was moderate
for Almighty (partial Ș2 = .09), Inspired (partial Ș2 = .06), and Global
(partial Ș2 = .08). The exposed effect sizes indicated that between 2% and
9% of the variance in the meanings was explained by the variable Type of
Google user.

Table 6-21. The Effects of Type of Google User on the Google

Meanings a

Brand Engaged users Disengaged users F

Partial Ș2
Meaning (1) (2) (1, 309)
Almighty .504 (.781) -.369 (.983) 30.21** .089
Helpful b -.291 (.640) .213 (1.153) 15.3** .047
Intrusive -069 (.928) .051 (1.049) .126 .000
Inspired .461 (.856) -.337 (.964) 20.3** .062
Fun b -.349 (.878) .255 (1.009) 6.36* .020
Global b -303 (.719) .222 (1.114) 25.414** .076
Informative .313 (.788) -.229 (1.076) 15.0** .046
a
The values outside the brackets represent the countries’ means, whereas the values in the
brackets represent the corresponding standard deviations.
b
The negative values represent high emphasis of the meaning.
* p < .05; ** p < .01

In order to measure the interaction effects, the combination of the two

independent variables, Country and Type of Google user, resulted in a
total of eight subgroups. All of the subgroups had more than 20
observations, except for the group of the Disengaged American users. This
group has 14 respondents and hence the results had to be interpreted with
caution. The interaction effect was statistically significant for the
Informative meaning (p < .01). Graph 6-1 shows that the Engaged
Austrian users emphasised the Informative meaning less than the
Disengaged Austrian users, while the Engaged users in the other three

173
countries emphasised the Informative meaning more than the Disengaged
users. The interaction effect size is considered small (partial Ș2 = .05), i.e.
5% of variance in the meaning Informative was explained by this
interaction.

This interaction effect was not predicted. However, this result shed more
light on the results of the effect of Country on Informative reported
previously. The uncertainty avoiding dimension predicted that the
Austrians would emphasise Informative meaning the least. The results got
distorted because of the inexplicably high score of the Disengaged
Austrian users.

Graph 6-1. The interaction Effect on the Google Informative meaning

Country
Emphasis of the Informative Meaning

0,60
Austria
0,40 Finland
India
0,20 United
States
0,00
-0,20

-0,40

-0,60

-0,80

Engaged users Disengaged users

Type of Google user

For all the other dependent variables, the interaction effects between
Country and Type of Google user were not statistically significant as can
be seen from the following p values: Almighty (p = .61), Helpful (p = .78),
Intrusive (p = .69), Inspired (p = .51), Fun (p = .71), and Global (p = .48).

174
Differences in Nike’s Brand Meanings

A set of 4 (country) x 3 (type of Nike user) two-way ANOVAs was

conducted in order to determine whether differences in Nike’s meanings
existed among the different countries and the different types of users. The
five previously identified meanings were measured in continuous
numerical values and each represented a dependent variable in a separate
ANOVA. Variances of the meaning Sports were not homogeneous across
the groups (ap4-34), so non-parametric tests were applied for this
meaning. Country is a nominal independent variable which differentiates
between four groups of respondents. Type of Nike user represents an
ordinal numerical variable with three levels.

The results of the ANOVA tests for the independent variable Country are
displayed in table 6-22. Country had a statistically significant effect on the
following meanings: Achiever (F (3, 294) = 8.96; p < .01), Not socially
responsible (F (3, 294) = 6.96; p < .01) and Pricey (F (3, 294) = 17.46, p <
.01). The effect sizes were considered moderate for Achiever (partial Ș2 =
.08) and Not socially responsible (partial Ș2 = .07), and strong for Pricey
(partial Ș2 = .15), which indicated that between 7% and 15% of the
variance in the meanings was explained by the variable Country.

The post-hoc tests for the Achiever dependent variable revealed that the
Indians emphasised the Achiever meaning statistically significantly more
(M= .54; SD = .86) than the Austrians (M = -.21; SD = 1.03) and Finns
(M = -.22; SD = .96) did. The American score (M = .10; SD = .91) was
centrally situated and not statistically significantly different from any of
the other three scores.

The Achiever meaning was expected to be influenced by performance

orientation and assertiveness. Highly performance-oriented and highly
assertive cultures value competition, success and progress (House et al.

175
 2004, p. 245, 405). Hence, countries high in these cultural dimensions
were predicted to emphasise the Achiever meaning more than others.
Since all four countries display a similar appreciation of performance
orientation, the results were evaluated in terms of assertiveness and found
to be as predicted. India had a high score (AI = 4.76), the United States
(AUS = 4.32) a moderate score and Austria (AA = 2.81) and Finland (AF =
3.68) low scores on valuing assertiveness.

Table 6-22. The Effects of Culture on the Nike Meanings a

Brand Austria Finland India United States F Partial

Meaning (1) (2) (3) (4) (3, 294) Ș2
Achiever -.207 (1.032)3 -.218 (.964)3 .539 (.858)1.2 .102 (.908) 8.96 ** .084
Not weak -.012 (1.076) .016 (.867) -.148 (1.117) .195 (.888) 1.971 .020
Not socially
responsible .217 (1.008)2,3 -.140 (1.026)1,4 -.467 (.647)1,4 .439 (1.040)2,3 6.96** .066
Pricey b -.037 (.937)2,3 .581 (.864)1,3,4 -.634 (.753)1,2,4 -.136 (1.080)2,3 17.463** .151
Sports b .019 (1.085) .114 (.926) -.078 (.984) -.144 (.973) 2.235 .022
a
The values outside the brackets represent the countries’ means, whereas the values in the
brackets represent the corresponding standard deviations. The indices indicate which groups
are significantly different in statistical terms (Bonferroni post-hoc test).
b
The negative values represent high emphasis of the meaning
* p < .05; ** p < .01

The post-hoc tests for the Not socially responsible meaning revealed that
there was a statistically significant difference between the Austrians (M =
.22; SD = 1.01) and the Americans (M = .44; SD = 1.04) on one side, and
the Indians (M = -.47; SD = .65) and the Finns (M = -.14; SD = 1.03) on
the other side. The former two emphasise the Not socially responsible
meaning more than the later two.

Not socially responsible was predicted to be influenced by two cultural

dimensions; specifically, assertiveness and human orientation. Countries
low in assertiveness have sympathy for the weak (House et al. 2004, p.
405); while those high in human orientation value altruism, benevolence
and generosity (House et al. 2004, p. 570). Both sets of characteristics

176
belong to people who would emphasise the Not socially responsible
meaning. In comparison to the other countries, India values assertiveness
the most (AI = 4.76) and human orientation (HOI= 5.28) the least. This
was in line with its very low emphasis of the Not socially responsible
meaning for Nike. Austria, on the other hand, has the opposite results. It
values assertiveness (AA = 2.81) the least and human orientation the most
(HOA = 5.76). Consequently, Austria had the highest score on the Not
socially responsible meaning emphasis. Finland (HOF = 5.81) is high in
human orientation but moderate in assertiveness (AF = 3.68), which was
also in line with its position between Austria and India in emphasising the
Not socially responsible Nike meaning. However, the highest emphasis of
the Not socially responsible Nike meaning in the United States could not
be explained by its score on the afore-mentioned two cultural dimensions
(AUS = 4.32; HOUS = 5.53).

The post-hoc tests for the Pricey meaning revealed that the Indians (M = -
.63; SD = .75) emphasised this meaning statistically more than all the
other studied countries. Austria (M = -.04; SD = .94) and the United States
(M = -.14; SD = 1.08) followed. They did not statistically significantly
differ from each other in their Pricey meaning emphasis, but they
emphasised Pricey statistically significantly more than Finland did (M =
.58; SD = .86).

Pricey was not predicted to be influenced by culture, so the results were

not as predicted. However, the meaning Pricey was most probably not
influenced by cultural dimensions. It seemed that in the case of Nike,
India’s economic situation might have triggered the question of Nike’s
affordability and affected differences in its Pricey emphasis.

As predicted Sports was equally emphasised across countries. Finally, the

Not weak meaning was extracted in the qualitative phase as a result of

177
negative inquiry. It was predicted to vary along countries because it
should have been influenced by assertiveness, but the results most
probably pointed out that nobody associated Nike with being weak.
Hence, the differences in emphasis are very small and statistically
insignificant.

The results of the ANOVA tests for the independent variable, Type of Nike
user, are displayed in table 6-23. The Type of Nike user had a statistically
significant effect on meanings: Achiever (F (2, 294) = 19.66; p < .01), Not
weak ((F (2, 294) = 12.87; p < .05), Not socially responsible (F (2, 294) =
19.98; p < .01), and Sports (F (2, 294) = 19.64, p < .01). The effect size
was considered moderate for all statistically significant results, i.e.
Achiever (partial Ș2 = .12), Not weak (partial Ș2 = .08), Not socially
responsible (partial Ș2 = .12), and Sports (partial Ș2 = .12), which
indicated that between 8% and 12% of the variance in the meanings was
explained by the variable Type of Nike user.

Table 6-23. The Effects of Type of Nike User on the Nike Meanings a

Engaged users Disengaged Non-users F Partial

Brand Meaning
(1) users (2) (3) (2, 294) Ș2
Achiever .418 (.956)2,3 -.271 (.808)1 -.367 (.967)1 19.659** .117
Not weak .281 (.990)3 -.057 (.875) -.325 (.991)1 12.867** .080
Not socially responsible -.454 (.868)2,3 .314 (.970)1 .386 (.946)1 19.98** .120
Pricey b -.056 (.975) .123 (.960) -.005 (1.058) .869 .006
Sports b -.353 (.786)2,3 .156 (1.019)1 .356 (1.087)1 19.635** .118
a
The values outside the brackets represent the countries’ means, whereas the values in the
brackets standard deviations. The indices indicate which groups are statistically significantly
different (Bonferroni post-hoc test).
b
The negative values represent high emphasis of the meaning
* p < .05; ** p < .01

The post-hoc tests for the Achiever meaning revealed that the Engaged
users emphasised the Achiever meaning statistically significantly more (M
= .42; SD = .96) than the Disengaged users (M = -.27; SD = .81) and the

178
Non-users (M = -.37; SD = .97), whereas the latter two groups did not
statistically differ between each other. Likewise, the post-hoc tests for the
Sports dependent variable revealed that the Engaged users emphasised the
Sports meaning significantly more in statistic terms (M = -.35; SD = .79)
than the Disengaged users (M = .16; SD = 1.02) and the Non-users (M =
.36; SD = 1.09), whereas the latter two subgroups did not differ to each
other.

On the other hand, the post-hoc tests for the Not weak meaning revealed
that the Engaged users emphasised the Not weak meaning statistically
significantly more (M = .28; SD = .99) than the Non-users (M = -.33; SD
= .99). The Disengaged users (M = -.06; SD = .88) moderately
emphasised the Not weak meaning, but the difference was not statistically
significant from either of the former two. Similarly, the post-hoc tests for
the Not socially responsible meaning revealed that Nike’s Engaged users
emphasised the Not socially responsible meaning statistically significantly
less (M= -.45; SD = .87) than the Disengaged users (M = .31; SD = .97)
and the Non-users (M = .39; SD = .95), whereas the latter two subgroups
did not statistically differ to each other. Finally, the Pricey meaning was
not statistically different among the three types of users.

The general conclusion for the effects of the variable Type of Nike user on
the meanings of Nike was that the negative meaning (i.e. Not socially
responsible) was more emphasised by those respondents who favour the
brand less, whereas the positive meanings (Achiever, Not weak and
Sports) were more emphasised by those who favour it more. In other
words, for the Engaged users, Sports, Not weak and Achiever indicated
what Nike means to them more than for the Disengaged users and the
Non-users. In contrast, for the Non-users and the Disengaged users, the
Not socially responsible meaning indicated what Nike means to them

179
more than for the Engaged users. These results were in line with the
prediction.

To measure the interaction effects, the combination of the two

independent variables Country and Type of Nike user resulted in a total of
twelve subgroups. The interaction effects were not statistically significant
for any of the meanings as can be seen from the following p values:
Achiever (p = .58), Not weak (p = .56), Not socially responsible (p = .14),
Pricey (p = .56), and Sports (p = .22).

Overall, this section provided the necessary data for the evaluation of
hypotheses 1, 2, and 3. The results were individually commented in
regards to how they met the hypothesised predictions. Evaluation of
hypotheses on an overall basis is conducted in the discussion in
subchapter 6.2.4.

6.2.3.5. The Moderating Effect of Country

The data presented in the previous subchapters showed how the

respondents’ countries and usage habits influenced the respondents’
emphasis of the brand meanings. For the purpose of this study, in
particular hypothesis 4, it was also necessary to determine whether
differences in the emphasis of the brand meanings among different user
types were constant across cultures. It was also important to identify
meanings that were more emphasised among the Engaged users than the
other users in each of the countries for each of the brands. This was
important as such data provide useful guidelines on how to increase the
number of the Engaged users. For this analysis, the Type of user is
observed as an interval variable and renamed Affection towards a brand.
For each brand this variable ranges from the lowest affection for the Non-
users to the highest for the Engaged users.

180
The results in table 6-24 show that the all-countries overall sample
displayed a statistically significant correlation between the meanings
Refreshing (r = .29; p < .01), Energising (r = .18; p < .01) and Unhealthy
(r = -.42; p < .01) on one side and Affection towards Coca-Cola on the
other side. Therefore, this indicates that the emphasis of the meanings
Refreshing and Energising is positively related towards the respondents’
affection towards Coca-Cola.

Observing each country separately, Austria displayed the most similar

pattern to the overall sample with statistically significant correlations for
Refreshing (r = .25; p < .01), Energising (r = .19; p < .05) and Unhealthy
(r = -.18; p < .01). In addition, Globally recognised (r = .21; p < .05) was
also statistically significantly correlated with Affection towards Coca-Cola
in Austria, indicating that Austrians who had more affection for Coca-cola
placed a higher emphasis on the Globally recognised meanings than other
Austrians who did not.

Table 6-24. The Partial Correlations a for Affection towards Coca-

Cola b

Brand Overall Austria Finland India United States

Meaning (df = 358) (df =103) (df = 98) (df = 77) (df = 56)
Refreshing .289 (.000)** .252 (.010)** .280 (.005)** .219 (.052) .254 (.054)
Globally .056 (.289) .211 (.031)* -.140 (.169) .146 (.200) .046 (.730)
recognised
Unhealthy -.417 (.000)** -.179 (.000)** -.404 (.000)** -.330 (.003)** -.337 (.010)**
Christmas c -.052 (.328) -.056 (.569) -.197 (.049)* .090 (.430) .045 (.738)
Classic .007 (.896) -.006 (.954) .143 (.157) -.087 (.443) .180 (.175)
Energising .184 (.000)** .192 (.049)* .127 (.208) .195 (.085) .413 (.001)**
Bonding with .022 (.677) .011 (.912) -.055 (.958) .051 (.655) .215 (.105)
friends c
a
Each brand meanings’ correlations to Affection towards Coca-Cola were controlled for the
effects of the other six brand meanings listed in the first column of the table.
b
The values outside the brackets represent the correlation coefficients, whereas the values in
the brackets the significance levels.
c
The negative values represent high emphasis of the meaning.
* p < .05; ** p < .01

181
In Finland, Refreshing (r = .28; p < .01) and Unhealthy (r = -.40; p < .01)
also showed the same pattern as in the overall sample. In addition, the
correlation between Christmas and Affection towards Coca-Cola, (r = -
.20; p < .05) was also positive and statistically significant in Finland. This
indicated that the Christmas Coca-Cola meaning was more emphasised by
the Finns who have more affection for the brand, and less by those who do
not.

In India, only Unhealthy (r = -.33; p < .01) was statistically significantly

correlated to Affection towards Coca-Cola, while in the United States
Unhealthy (r = -.34; p < .01) and Energising (r = .42; p < .01) repeated the
same pattern as in the all-countries overall sample. Finally, when the
critical significance level was set just slightly less rigorously, Refreshing
also became significantly positively correlated with Affection towards
Coca-Cola in both India (r = .22; p = .05) and the United States (r = .25; p
= .05).

Table 6-25 shows the results for Google. The all-countries overall sample
displayed statistically significant correlations between the meanings
Almighty (r = .27; p < .05), Intrusive (r = -.12; p < .05), Inspired (r = .26; p
< .01), Fun (r = -.15; p < .01) and Informative (r = .13; p < .05) on one
side, and Affection towards Google on the other side. This indicated that
the meanings Almighty, Inspired, Fun, and Informative were more
emphasised by those respondents who were more affected towards
Google. On the other hand, the meaning Intrusive (r = -.12; p < .05) had a
negative direction of the correlation, so it was more emphasised by those
respondents who were less affected towards Google.

Observing each country separately, correlations occurred for far fewer

meanings in each of the countries. Finland displayed the most similar
pattern to the overall sample, with statistically significant correlations for

182
Almighty (r = .23; p < .05), Inspired (r = .23; p < .05) and Informative (r =
.24; p < .05). In Austria, only the meanings Almighty (r = .25; p < .05) and
Intrusive (r = -.22; p < .05) showed the same pattern as in the overall
sample, while all the other meanings did not correlate significantly to
Affection towards Google. In India, no meaning was statistically
correlated to Affection towards Google. Finally, in the United States only
the meaning Informative (r = .41; p < .01) had a statistically significant
positive correlation to Affection towards Google.

Table 6-25. The Partial Correlations a for Affection towards Google b

Brand Overall Austria Finland India United States

Meaning (df = 309) (df = 93) (df = 82) (df = 65) (df = 45)
Almighty .272 (.000)** .251 (.014)* .227 (.038)* .130 (.296) .136 (.362)
Helpful c -.029 (.616) -.095 (.360) .029 (.792) .010 (.934) -.114 (.446)
Intrusive -.124 (.028)* -.219 (.033)* .017 (.881) .023 (.856) -.198 (.183)
Inspired .263 (.000)** .179 (.083) .227 (.038)* .081 (.515) .007 (.965)
Fun c -.146 (.010)** -.074 (.477) -.062 (.578) -.014 (.908) -.259 (.179)
Global c -.105 (.065) -.093 (.369) -.213 (.052) -.159 (.200) -.069 (.644)
Informative .126 (.026)* -.139 (.178) .244 (.025)* .029 (.818) .410 (.004)**
a
Each brand meanings’ correlations to Affection towards Google were controlled for the
effects of the other six brand meanings listed in the first column of the table.
b
The values outside the brackets represent the correlation coefficients, whereas the values in
the brackets the significance levels.
c
The negative values represent high emphasis of the meaning.
* p < .05; ** p < .01

The results in table 6-26 show that almost all of Nike’s correlations were
statistically significant. In particular, Achiever (r = .24; p < .01), Not weak
(r = .24; p < .01), Not socially responsible (r = -.37; p < .01), and Sports (r
= -.16; p < .01) on one side, and Affection towards Nike on the other side
were all statistically significantly correlated. Such results indicated that
Achiever, Not weak and Sports were more emphasised by the respondents
the more they felt affection towards Nike Conversely, Not socially
responsible was more emphasised by the respondents who felt less
affection towards Nike.

183
Table 6-26. The Partial Correlations a for Affection towards Nike b

Brand Overall Austria Finland India United States

Meaning (df = 300) (df = 95) (df = 84) (df = 60) (df = 43)
Achiever .241 (.000)** .283 (.005)** .232 (.002)** .014 (.913) .066 (.665)
Not weak .236 (.000)** .293 (.004)** .190 (.080) .054 (.678) .388 (.008)**
Not socially -.371 (.000)** -.338 (.001)** -.434 (.000)** -.089 (.494) -.368 (.013)*
responsible
Pricey c .086 (.135) .049 (.636) -.072 (.510) .051 (.695) .149 (.329)
Sports c -.164 (.004)** -.075 (.467) -.218 (.043)* -.246 (.053) -.279 (.064)
a
Each brand meanings’ correlations to Affection towards Nike were controlled for the effects
of the other four brand meanings listed in the first column of the table.
b
The values outside the brackets represent the correlation coefficients, whereas the values in
the brackets the significance levels.
c
The negative values represent high emphasis of the meaning.
* p < .05; ** p < .01

The Meaning Achiever had the same relation to Affection towards Nike
only in Austria (r = .28; p < .01) and Finland (r = .23; p < .01), whereas in
the other countries it did not correlate to Affection towards Nike. Not weak
was also statistically significantly correlated to Affection towards Nike in
Austria (r = .29; p < .01) and in the United States (r = .39; p < .01). In
three of the studied countries: Austria (r = -.34; p < .01), Finland (r = -.43;
p < .01) and the United States (r = -.37; p < .01), Not socially responsible
was statistically significantly correlated to Affection towards Nike.

The emphasis of Pricey was not statistically significantly correlated to

Affection towards Nike in any of the countries. Finally, Sports was a
meaning that was statistically significantly correlated to Affection towards
Nike in Finland (r = -.22; p < .05) (and with a slightly more loosely
conditioned significance level in India (r = -.25; p = .06) and the United
States (r = -.28; p = .06)).

Overall, the results showed that correlations between a brand meaning and
affection towards a brand are not the same across cultures. A more

184
detailed discussion on the findings needed for the evaluation of hypothesis
4 is provided in subchapter 6.2.4.4.

6.2.4. Discussion

This chapter evaluates the hypotheses based on the summarised findings

provided in the previous subchapters.

6.2.4.1. Hypothesis 1

Most of the brand meanings differed across cultures not only in their
emphasis, but also in the importance of their contribution to the overall
brand meaning. Such results provided plenty of evidence to support
Hypothesis 1, which claimed that: The meaning of certain brands differs
for consumers in different cultures. These results are in accordance with
previous research (cf. Gram 2007; Phau and Lau 2000; Berthon, Pitt, and
Campbell 2008).

6.2.4.2. Hypothesis 2

Table 6-27 presents summarised evidence for Hypothesis 2, which

predicted that Brand meaning is affected by cultural dimensions. Out of
the thirteen H2 sub-hypotheses, which each predicted the effect of a
different cultural dimension, this research provided evidence to test eight
of these hypotheses. This is the case because of the nature of the research,
which allowed testing of only those particular meanings (and hypotheses)
that were identified by the interviewees as the meanings relevant for the
three studied brands.

As table 6-27 shows, three sub-hypotheses are fully supported (namely,

H2b - which predicted the effect of low in-group collectivism; H2e -
which predicted the effect of high power distance; and H2k - which
predicted the effect of low human orientation.) Five additional hypotheses

185
are partly supported (namely, H2a – which predicted the effect of high
collectivism; H2c - which predicted the effect of high uncertainty
avoidance; H2g – which predicted the effect of high assertiveness; H2h –
which predicted the effect of low assertiveness; and H2j - which predicted
the effect of high human orientation). Overall, this gives substantial
evidence to support hypothesis 2.

6.2.4.3. Hypothesis 3

Almost all of the positive meanings were found to be statistically more

emphasised by the Engaged users than by the others, and all the negative
meanings to be more emphasised by the Non-users than by the others. In
addition, those meanings that did not show such patterns were not
opposite to what was expected, but rather neutral (i.e. those meanings
showed no significant difference in the meaning emphasis between the
Engaged users and the Non-users). All of these arguments give a solid
basis for accepting hypothesis 3, which predicted that Brand users
(especially engaged ones) will emphasise positive brand meanings more
than non-users, whereas non-users will emphasise negative meanings
more than brand users (especially engaged ones).

186
Table 6-27. The Overall Effects of the Cultural Dimensions on the
Brand Meanings

Cultural
Coca-Cola H Google H Nike H Conclusion
dimension
Collectivism+ Bonding with S c Global NS H2a - partially
friends supported
Collectivism- Intrusive S H2b - supported

Uncertainty Globally NS Informative PS H2c – partially

avoidance + recognised supported
Uncertainty H2d – n.a.
avoiding -
Power Almighty S H2e – supported
distance +
Power H2f – n.a.
distance -
Assertiveness+ Inspired S Achiever S H2g – partly supported
a
Not weak NS
b
Assertiveness- Unhealthy S Helpful NS Not socially PS H2h – partly supported
responsible
Classic PS

Performance H2i – n.a.

orientation +
Human Christmas Sc Not socially PS H2j – partially
orientation+ responsible supported
Human Fun S H2k – supported
orientation -
Future H2l – n.a.
orientation +
Gender H2m – n.a.
egalitariani +
Not affected Refreshing Sd Sports S
by culture
Energising S Pricey Sd
a
Predicted to be affected by assertiveness, but no effect was captured.
b
Predicted to be affected by assertiveness and human orientation, but no effect was found.
c
The results are distorted for forces other than culture
d
No cultural effect predicted - statistically significantly higher results for India are ascribed
to environmental forces other than culture.
S = supported, PS= partly supported, NS = not supported, n.a. = not applicable

187
6.2.4.4. Hypothesis 4

Negative meanings showed consistency across all the countries in terms of

their different emphasis by Users vs. Non-users. In particular, such a
pattern is demonstrated by the Not socially responsible Nike meaning and
the Unhealthy Coca-Cola meaning. Apart from the negative meanings,
consistency across all the countries is also demonstrated for the functional
meanings Refreshing and Informative.

Except for the afore-mentioned meanings, the overall results differed

among the countries and hence provided substantial evidence for the
acceptance of hypothesis 4, which predicted that Differences in brand
meaning emphasis between different types of users are not equal in each
country.

188
 7. Conclusion

7.1. Summary of the Findings

The aim of this research was twofold. Firstly, to identify what brands
mean to consumers and specify the brand meaning concept, and secondly
to explore whether differences in brand meanings exist among consumers
of different cultures and whether they can be explained by differences in
cultural dimensions.

Investigation of past studies gave an insight into the existing concepts in

the world of brands. Although, or maybe even because, there are
numerous branding concepts (Keller 2003a), it remains unclear what
brands actually mean to consumers. Previous research analysis also
revealed several established paradigms that served as a firm basis for
constructing a relationship model between culture and brand meaning:
brand meaning is created in consumers’ minds (Brown, Kozinets, and
Sherry 2003; De Chernatony 2001, p. 19; Vargo and Lusch 2004); it is
based on the existing personal knowledge (Berthon, Pitt, and Campbell
2008) and value systems (Phau and Lau 2000); and knowledge and value
systems are culturally determined (Hofstede 2001).

To address the first aim of the paper, the research was conducted in the
spirit of scientific discovery (Snow and Thomas 1994, p. 465). The
empirical investigation was entered without a priori defined research
propositions. A series of in-depth interviews was conducted with
consumers in the four studied countries in order to understand what some
of the carefully chosen global brands mean to them. Transcript and visual
data were coded in three rounds until coherent groups of meanings were

189
obtained. Analysis of the evoked meanings showed that the concept of
brand meaning is both similar and different to the other established
branding concepts such as brand association, attitude, attribute,
personality, belief, image and benefit. Consequently, brand meaning was
defined as a set of cognitive and affective experience-based brand
associations.

To address the second aim of the research, i.e. to test the hypotheses that
brand meaning is culturally determined and that cultural values are
reflected in brand meanings, it was important to keep brand positioning
and the respondents’ profiles constant across the cultures. Constant brand
positioning was ensured by taking three uniquely positioned brands as the
objects of the research in all the studied countries. Similarly, the
respondents’ profiles were kept constant by conducting a survey on the
cross-cultural matched samples. The findings provided evidence which
supported all the hypotheses.

Since one of the goals of the research was also to explore whether
different types of users emphasise brand meanings differently and whether
these differences are constant across cultures, the appropriate tests were
conducted. The results are as predicted. In general, positive brand
meanings are more emphasised by users and negative by non-users, but
these differences are not present for all the meanings in all the studied
countries.

190
7.2. Managerial Implications

This research has several managerial implications as detailed below.

7.2.1. Brand Meaning is a set of Cognitive and Affective Experience-

Based Brand Associations

The results of this study show that a brand becomes more meaningful the
more cognitive and affective experience-based associations it has. In other
words, with experience more types of meanings become associated to a
brand, which results in higher awareness and a higher recall rate in various
consumption situations. Concretely, Coca-Cola takes the top position on
the Interbrand and Business Week (2008) list as the brand with the highest
brand value, Google the tenth position and Nike “only” the 29th. Coca-
Cola and Google are both richer in terms of the various positive meanings
that consumers attach to them. Moreover, when it comes to types of
meanings, symbolic meanings are becoming increasingly important to
consumers. Nowadays, many brands are primarily consumed for their
symbolic meanings (Veloutsou and Moutinho 2008). The importance of
symbolic meanings is also visible in the case of Coca-Cola. It has the
highest brand value and the highest number of symbolic meanings among
the studied brands.

Therefore, brand managers are advised to invest their creative energy in

creating as many experiences for their consumers and potential consumers
as possible so that consumers can expand their brand association
networks, become more connected to the brand and use it more often with
great passion. This will result in increased profits for the companies which
own the brands in question.

191
7.2.2. Consumer Understanding of Brands is Culturally Determined

The results of the research show that even the most global brands have
different meanings for consumers in different cultures. This is because of
the culturally embedded interpretation processes. Practitioners should thus
firstly seek information on the cultural specifics in the target countries.
Secondly, they should decide whether they consider unique positioning
beneficial, or not. In case they desire consistency of meanings across
cultures, they should, contrary to expectations, employ country-adapted
communication and even consider adapting other marketing elements.

In case practitioners do not consider cross-cultural brand meaning

consistency crucial and beneficial, they should adapt their brand messages
in a way that they evoke particular associations and behaviours on the side
of consumers.

7.2.3. Negative Brand Meanings Relate to Brand Usage

The results show that the emphasis of negative meanings between users
and non-users across brands and countries differs the most. The research
does not test for cause and effect, but the results give an important
indication that a negative meaning might be the reason for some
consumers’ avoidance of a particular brand, to a greater extent than
positive meanings might be the reason for opting for the brand. Therefore,
managers are advised to pay particular attention to the negative aspects of
their brand’s meaning and give priority to fighting such meanings.

Furthermore, both Coca-Cola and Nike received their negative meanings

when they failed to recognise consumer trends. Coca-Cola has been
building its favourable iconic position since 1886 almost solely on one
sweet and energising beverage. While this formula was successful for a
century, the trend changed in the 1990s when consumers became more

192
fitness and wellness-oriented. Isdel, Coca-Cola’s CEO, admits that Coca-
Cola did not understand the significance of the trend at the time (Theater
2004).

Similarly, since its beginnings in 1962, Nike has never owned its own
facility and still houses its production facilities with Asian subcontractors
(Riesenbeck and Perrey 2007). Observing this issue from the perspective
of brand management, it is evident that Nike has neglected a consumer
trend as well. That is, for decades nobody questioned Nike’s integrity, and
nobody even related fancy high-tech shoes to the (lack of) working and
living conditions in Nike’s subcontractors’ production facilities. Only, in
the 1990s, as the consumers grew increasingly aware and sensitive to
global societal problems, the Nike sweatshops scandal broke out and hurt
its image.

Therefore, from the point of view of brand management, managers are

advised to systematically and diligently observe consumer trends to avoid
negative meanings becoming attached to their brands. This advice is
aimed at market challengers, but even more so at market leaders who are
more prone to ignoring the trends while being comfortable ahead of their
competitors.

7.3. Scientific Implications

7.3.1. Major Scientific Insights of the Research

This research contributes scientifically to the fields of brand management,

cross-cultural studies and consumer research. Each of the contributions
will be further elaborated below.

193
7.3.1.1. Contribution to Brand Management Field

By taking an inductive approach to conceptualising brand meaning, this

research contributes to the field of brand management. After having
thoroughly analysed the existing branding literature, the need for a new
concept named brand meaning has been identified. This is despite the
many existing models and views on branding, and in line with the fact that
it is still not clear why consumers prefer certain brands and which
meanings those brands have for them. Previous literature has provided
evidence that brands become strong, successful and preferred by
consumers because of their appealing personalities (Phau and Lau 2000;
Riesenbeck and Perrey 2007, p. 151); good resonance with the target
consumers (Pringle and Thompson 1999, p. I); personal relationships with
consumers (Fournier 1998; Sweeney and Brandon 2006); or benefits for
consumers (Park, Jaworski, and Maclnnis 1986; Kapferer 2004, p. 23).

Because these are all related, but rather different concepts, this research
took a step away from the existing literature by turning directly to
consumers and asking them for the answers. The inductive approach was
considered superior to its deductive counterpart because research in brand
management field has reached a mature stage. Focusing on theory testing
alone at the current mature stage of brand management research might
hamper further knowledge creation (Montgomery, Wernerfelt and
Balakrishnan 1989).

Because the field research was entered theory-free, the direct contact with
consumers made it easier to understand what brands mean to consumers
and whether brand meaning is a concept most linked to benefits,
personalities, images, associations, relationships, and the like. The
interpretation of the results revealed that the concept of brand meaning has

194
some unique aspects and a value of its own even though it shares many
aspects of the just mentioned concepts.

7.3.1.2. Contribution to the Field of Cross-cultural Research

This research is embedded in the cross-cultural environment. Its

contribution to the field is related to observing cultural dimensions as a
priori predictors, which is an often neglected aspect of the research. A
more common approach in cross-cultural studies is either identifying the
studied phenomena’s differences among countries (e.g. in the studies by
Costa and Pavia 1992 and Gram 2007), or at best using cultural
dimensions as a posteriori constructs to explain the findings (e.g. in the
studies by Foscht et al. 2008 and Pankhania, Lee, and Hooley 2007).

This research showed that cultural dimensions can and should be used as a
priori predictors of certain consumer behaviours. Only when used a priori
can cultural dimensions really serve the purpose they were defined for in
the first place, and that is to help managers best meet the specifics of
particular local markets.

7.3.1.3. Contribution to Methodology in Consumer Research

Because this research was conducted on three continents, without modern

technologies its completion would have been much more complex, timely
and costly. Although it has by now already become an established practice
to conduct online surveys, online interviews conducted in written form
have not been widely used among researchers. This research proved that
for computer-literate respondents, online communication is not only just
as good as face-to-face communication, but in some respects even
superior to it.

The collected responses were compared to those of in-person conducted

pilot interviews and no difference was discovered in either the depth or

195
breadth of the responses. As the results of this research proved, the
feelings and moods of the respondents can also be expressed with the
mood icons integrated into most of the online messengers, and potentially
even better stored in the transcript. In addition, this mode of research
allows the respondents to stay totally anonymous and remain in the
comfort of their homes, with the possibility of even stopping the interview
and resuming it at a later stage. All of these things contributed greatly to
the validity of the responses. In addition, responses collected in written
form are much easier to process, which leaves less space for errors in
interpretation and coding.

Furthermore, although qualitative research usually serves to gain an initial

understanding of the studied phenomenon, and quantitative research
allows researchers to collect solid data to confirm hypotheses, this two-
stage research discovered that the in-depth interviews already yielded
results that greatly resembled those later obtained in the quantitative stage.
Hence this research provided further evidence to support Coulter,
Zaltman, and Coulter (2001) who claim that the ZMET method (that
served as a basis for this research) can provide up to 90% of the
information available from a larger set of classical interviews in just four
to five in-depth interviews.

7.3.2. Limitations of the Research

There were several limitations to this study. As for the qualitative stage,
the study was adjusted to an online setting. Although the Internet
environment provided a number of advantages over the face-to-face
interview setting, some of the association elicitation techniques such as
image grouping and mind mapping could not have been conducted. Such
techniques might have provided some additional relevant data on brand
meanings.

196
Furthermore, interview transcripts were analysed by applying the
hermeneutic approach. Such approach has an inherited problem of
interpretation. In other words, any interpretation is heavily dependent on
interpreter’s “personal history and cultural background” (Thompson,
Pollio, and Locander 1994). Hence, although the coding of the responses
was done with great care, the validity of the procedure could have been
increased by employing multiple, preferably cross-cultural coders. Since
the coded results were grouped and the groups of meanings were tested for
reliability in the quantitative stage, this problem was diminished. The
concern remains as to whether all the relevant meanings were recognised
in the qualitative stage and taken into the quantitative.

When it comes to quantitative phase of the research, the study was

conducted in English. Although, all the procedures for potential problem
reduction were applied, the question remains whether conducting a survey
in the respondents’ native languages would have yielded somewhat
different results. Furthermore, although the overall sample was
substantial, when divided into groups during the two-way ANOVA tests,
some of the groups were relatively small. Thus, it would have been useful
to collect more data. Finally, although cross-cultural research rarely
investigates more than three cases in four cultures, expanding the research
to include more cases and more cultures is a generally valid piece of
advice.

7.3.3. Future Research

Firstly, it would be interesting to address the limitations of the current

research. Expanding the study to more cultures and brands, as well as
addressing the language issue of the research, (i.e conducting the study in
the respondents’ native languages, and coding the responses with a team
of multicultural researchers) would be beneficial.

197
In addition, the results call for further investigation of the relationship
between brand meaning on one side and brand preference and loyalty on
the other. More specifically, although this research revealed some initial
understandings of which types of meanings are differentiated between
Engaged users, Disengaged users and Non-users, the conclusions of this
research would be strengthened with the results obtained in a study
designed in a cause and effect mode. Furthermore, it would also be
interesting to investigate how some associations in consumer minds
evolve to represent meanings; and what, apart from emphasised consumer
needs, makes certain meanings more relevant to consumer than others.

From a cross-cultural perspective, it would also be interesting to study

whether consumers in different cultures form brand meaning groups
differently. The results of the present study provide evidence of a rather
consistent pattern in that matter. However, focusing on this issue in
particular could provide even stronger conclusions. In addition, it would
be beneficial to study several brands of the same product category to see
whether the brand meanings of the preferred brand resemble the culturally
determined value systems of the respondents more than the non-preferred
brands.

Finally, during the research the question of whether the world is becoming
globalised, Westernised or Americanised arose. Although such a topic
would require a multidisciplinary perspective, from the point of view of
brand management, it would be interesting to firstly investigate whether
brands originating from the United States really convey American values
and spread them around the world, or is it that they are truly global, in the
sense that such brands drain different aspects of cultures from all parts of
the world and actually represent global values. Additionally, it would also
be interesting to analyse which role brands play in creating popular culture
locally and whether their role is dominant in comparison to some other

198
forms of multicultural exposure, like media, art, literature, personal
experiences and the like.

199
 References

Aaker, David A. (1991), Managing Brand Equity: Capitalizing on the

Value of a Brand Name. New York, NY: The Free Press.

------- (1994), “Building a Brand: The Saturn Story,” California

Management Review, 36 (2), 114–33.

-------, David A. (1996), Building Strong Brands. New York: The

FreePress.

-------, David A. and Erich Joachimsthaler (2002), Brand Leadership.

London: Simon & Schuster.

-------, David A., V. Kumar, and George S. Day (2001), Marketing

Research (7th ed.). New York: John Wiley & Sons, Inc.

Aaker, J.L. (1997), “Dimensions of Brand Personality,” Journal of

Marketing Research, 34 (3), 347-56.

-------, Jennifer L. (2000), “Accessibility or Diagnosticity? Disentangling

the Influence of Culture on Persuasion Processes and Attitude,”
Journal of Consumer Research, 28 (4), 340-57.

Abou Aish, Ehab M., Christine T. Ennew, and Sally A. McKechnie

(2003), “A Cross-Cultural Perspective on the Role of Branding in
Financial Services: The Small Business Market,” Journal of
Marketing Management, 19 (9/10), 1021-42.

Achenreiner, Gwen Bachmann and Deborah Roedder John (2003), “The

Meaning of Brand Names to Children: A Developmental
Investigation,” Journal of Consumer Psychology, 13 (3), 205-19.

200
AdAge (2008) “Top 10 Slogans of The Century,” (accessed 20th
December 2008) [available at:
http://adage.com/century/slogans.html].

Adigüzel, Feray and Michel Wedel (2008), “Split Questionnaire Design

for Massive Surveys,” Journal of Marketing Research, 45 (5), 608-
17.

Albert Noël, Dwight Merunka, and Pierre Valette-Florence, (2008),

“When Consumers Love Their Brands: Exploring the Concept and its
Dimensions,” Journal of Business Research, 61 (10) 1062–75.

Alden, Dana L., Jean-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.

Allen, I. Elaine and Christopher A. Seaman (2007), “Likert Scales and

Data Analyses,” Quality Progress, 40 (7), 64-65.

Ambler, Tim and Chris Styles (1997), “Brand Development versus New
Product Development: Toward a Process Model of Extension
Decisions,” Journal of Product & Brand Management, 6 (4), 222-34.

Apelbaum, Eidan, Eitan Gerstner, and Prasad Naik (2003), “The Effects of
Expert Quality Evaluations versus Brand Name on Price Premiums,”
Journal of Product & Brand Management, 12 (3), 154-65.

Aperia, Tony and Rolf Back (2004), Brand Relations Management.

Copenhagen: Liber, Copenhagen Business School Press, Abstrakt.

Azoulay Audrey and Jean-Noel Kapferer (2003), “Do Brand Personality

Scales Really Measure Brand Personality,” Journal of Brand
Management, 11 (2), 143-55.

201
Banerjee Saikat (2008), “Dimensions of Indian Culture, Core Cultural
Values and Marketing Implications: An Analysis,” Cross Cultural
Management: An International Journal, 15 (4), 367-78.

Barr, Pamela S. and Mary Ann Glynn (2004), “Cultural Variations in

Strategic Issue Interpretation: Relating Cultural Uncertainty
Avoidance to Controllability in Discriminating Threat and
Opportunity,” Strategic Management Journal, 25 (1), 59-67.

Baumgartner, Hans, Mita Sujan, and James R. Bettman (1992),

“Autobiographical Memories, Affect, and Consumer Information
Processing,” Journal of Consumer Psychology, 1 (1), 53-82.

Berthon, Pierre, Leyland F. Pitt, and Colin Campbell (2008), “Does Brand
Meaning Exist in Similarity or Singularity?,” Journal of Business
Research, In Press, Available online 3 August 2008
(doi:10.1016/j.jbusres.2008.05.015)
Bidney, David (1954), “Reviewed Work(s): Culture: A Critical Review of
Concepts and Definitions by A. L. Kroeber; Clyde Kluckhohn,” The
American Journal of Sociology, 59 (5), 488-89.

Bjerke, Rune and Rosemary Polegato (2006), “How Well Do Advertising

Images of Health and Beauty Travel across Cultures? A Self-concept
Perspective,” Psychology & Marketing, 23 (10), 865-84.

Black, Thomas R (1999), Doing Quantitative Research in The Social

Sciences: An Integrated Approach to Research Design, Measurement
and Statistics. London: Sage.

Blythe, Jim (2007), “Advertising Creatives and Brand Personality: A

Grounded Theory Perspective,” Journal of Brand Management, 14
(4), 284-94.

202
Bodkin, Charles D., Christie Amato, and Cara Peters (2008), “The Role of
Conflict, Culture, and Myth in Creating Attitudinal Commitment,”
Journal of Business Research, In Press, Available online 20 June
2008 (doi:10.1016/j.jbusres.2008.05.005)

Brannen Mary Yoko, Carolina Gómez, Mark F. Peterson, Laurence

Romani, Lilach Sagiv and Pei-Chuan Wu (2004), “People in Global
Organizations: Culture, Personality, and Social Dynamics,” in The
Blackwell Handbook of Global Management, Henry W. Lane,
Martha L. Maznevski, Mark E. Mendenhall and Jeanne Mcnett, eds.
Blackwell Publishing, 26 - 54.

Broderick, Amanda J. (2007), “A Cross-national Study of the Individual

and National–cultural Nomological Network of Consumer
Involvement,” Psychology & Marketing, 24 (4), 343-74.

Broniarczyk, Susan M. and Joseph W. Alba (1994), “The Importance of

the Brands in Brand Extension,” Journal of Marketing Research, 31
(2), 214-29.

Brown, Stephen P., Robert V. Kozinets, John F. Sherry Jr. (2003),

“Teaching Old Brands New Tricks: Retro Branding and the Revival
of Brand Meaning,” Journal of Marketing, 67 (3), 19-33.

-------, Steven P. and Douglas M. Stayman (1992), “Antecedents and

Consequences of Attitude toward the Ad: A Meta-analysis,” Journal
of Consumer Research, 19 (June), 34-51.

Brown, Tom J., Peter A. Dacin, Michael G. Pratt, and David A. Whetten
(2006), “Identity, Intended Image, Construed Image, and
Reputation: An Interdisciplinary Framework and Suggested
Terminology,” Journal of the Academy of Marketing Science, 34
(2), 99-106.

203
Bulmer, Ralph (1953), “Review: 211 Reviewed Work(s): Culture: A
Critical Review of Concepts and Definitions by A. L. Kroeber;
Clyde Kluckhohn,” Man, 53, 136-37.

Bulmer, Sandy and Margo Buchanan-Oliver (2006), “Advertising across

Cultures: Interpretations of Visually Complex Advertising,” Journal
of Current Issues & Research in Advertising, 28 (1), 57-71.

Burnkrant, Robert E. and H. Rao. Unnava (1995), “Effects of Self-

Referencing on Persuasion,” Journal of Consumer Research, 22 (1),
17-26.

Calder, Bobby J. and Steven J. Reagan (2001), “Brand Design,” in

Kellogg on Marketing, Dawn Iacobucci, ed. New York: John
Wiley, 58-73.

Carlson, Brad D., Tracy A. Suter, and Tom J. Brown (2008), “Social
versus Psychological Brand Community: The Role of Psychological
Sense of Brand Community,” Journal of Business Research, 61 (4),
284-91.

Cattell, Raymond. B. (1966), “The scree test for the number of factors,”
Multivariate Behavioral Research, 1 (2), 245-76.

Cavusgil, S. Tamer and Ajay Das (1997), “Methodological Issues in

Empirical Cross-cultural Research: A Survey of the Management
Literature and a Framework,” Management International Review, 37
(1), 71-96.

Chang, Pao-Long and Ming-Hua Chieng (2006), “Building Consumer–

brand Relationship: A Cross-Cultural Experiential View,” Psychology
& Marketing, 23 (11), 927-59.

204
Cho, Bongjin, Up Kwon, James W. Gentry, Sunkyu Jun; Fredric Kropp,
(1999), “Cultural Values Reflected in Theme and Execution: A
Comparative Study of U.S. and Korean Television Commercials,”
Journal of Advertising, 28 (4), 59-73.

Christodoulides, George (2008), “Breaking free from the industrial age

paradigm of branding,” Journal of Brand Management, 15 (4), 291-
93.

Corain, Livio and Luigi Salmaso (2007), “A Critical Review and a

Comparative Study on Conditional Permutation Tests for Two-Way
ANOVA,” Communications in Statistics: Simulation &
Computation, 36 (4), 791-805.

Costa, Janeen Arnold and Teresa M. Pavia (1992), “What it all adds up to:
Culture and alpha-numeric brand names,” Advances in Consumer
Research, 19 (1), 39-45.

Costello, Anna B. and Jason Osborne (2005). “Best Practices in

Exploratory Factor Analysis: Four Recommendations for Getting the
Most from Your Analysis,” Practical Assessment Research &
Evaluation, 10 (7). [Available online:
http://pareonline.net/getvn.asp?v=10&n=7].

Coulter, Robin Higie and Gerald Zaltman (1994), “Using The Zaltman
Metaphor Elicitation Technique to Understand Brand Images”
Advances in Consumer Research, 21 (1), 501-07.

-------, Robin Higie, Gerald Zaltman, and Keith S. Coulter (2001),

“Interpreting Consumer Perceptions of Advertising: An Application
of the Zaltman Metaphor Elicitation Technique,” Journal of
Advertising, 30 (4), 1-21.

205
Cova, Bernard and Stefano Pace (2006), “Brand Community of
Convenience Products: New Forms of Customer Empowerment - The
Case “My Nutella the Community”,” European Journal of Marketing,
40 (9/10), 1087-2005.

Craig, C. Samuel and Susan P. Douglas (2000), International Marketing

Research. New York: John Wiley and Sons.

Czerniawski, Richard D. and Michael W. Maloney (1999), Creating

Brand Loyalty: The Management of Power Positioning and Really
Great Advertising. New York: Amacom.

Davis, Teresa (2007), “Brand Meaning and Children: A Thematic

Categorisation Task,” Journal of Brand Management, 14 (3), 255-66.

De Chernatony, Lesley (2001), From Brand Vision to Brand Evaluation.

Oxford: Butterworth-Heinemann Ltd.

De Matos, Celso and Carlos Rossi (2008), “Word-of-mouth

Communications in Marketing: A Meta-analytic Review of the
Antecedents and Moderators,” Journal of the Academy of Marketing
Science, 36 (4), 578-96.

Denzin, Norman K. and Yvonna S. Lincoln (1994), Handbook of

Qualitative Research. Thousand Oaks & London: Sage.

Dillon, William R., Thomas J. Madden, Amna Kirmani, and Soumen

Mukherjee, (2001), “Understanding What's in a Brand Rating: A
Model for Assessing Brand and Attribute Effects and Their
Relationship to Brand Equity,” Journal of Marketing Research, 38
(4), 415-29.

Dilts, Rober and DeLozier Judith (2000), Encyclopedia of Systemic NLP

and NLP New Coding. Scotts Valley: NLP University Press.

206
Dittrich, Regina, Brian Francis, Reinhold Hatzinger, and Walter
Katzenbeisser (2007), “A Paired Comparison Approach for the
Analysis of Sets of Likert-scale Responses,” Statistical Modelling:
An International Journal, 7 (1), 3-28.

Dolnicar, Sara and John R. Rossiter (2008), “The Low Stability of Brand-
Attribute Associations is Partly Due to Market Research
Methodology,” International Journal of Research in Marketing 25 (2),
104–08.

Edwards, Jeffrey. R. and Robert P. Bagozzi (2000), “On the Nature and
Direction of Relationships Between Constructs and Measures,”
Psychological Methods, 5 (2), 155–74.

Eisenhardt, Kathleen M. (1989), “Building Theories from Case Study

Research,” Academy of Management Review, 14 (4), 532-32.

Erdem, Tuelin, Joffre Swait, and Ana Valenzuela (2006), “Brands as

Signals: A Cross-Country Validation Study,” Journal of Marketing,
70 (1), 34-49.

Escalas, Jennifer Edson and James R. Bettman (2005), “Self-Construal,

Reference Groups, and Brand Meaning,” Journal of Consumer
Research, 32 (3), 378-89.

Fabrigar, Leandre. R., Duane. T. Wegener, Robert C. MacCallum, and

Erin. J. Strahan (1999), “Evaluating the Use of Exploratory Factor
Analysis in Psychological Research,” Psychological Methods, 4 (3),
272-99.

Feldwick, Paul (2002), What is Brand Equity Anyway?. Oxfordshire, UK:

World Advertising Research Center.

207
Fine, Seymour H (1981), The Marketing of Ideas and Social Issues. New
York: Praeger.

Fishbein, Martin and Susan Middlestadt (1995), “Noncognitive Effects on

Attitude Formation and Change: Fact or Artifact?,” Journal of
Consumer Psychology, 4 (2), 181-202.

Foscht, Thomas, Cesar Maloles III, Bernhard Swoboda, Dirk Morschett,

and Indrajit Sinha (2008), “The Impact of Culture on Brand
Perceptions: A Six-nation Study,” Journal of Product & Brand
Management, 17 (3), 131-42.

Fournier, Susan (1998), “Consumers and Their Brands: Developing

Relationship Theory in Consumer Research,” Journal of Consumer
Research, 24 (4), 343-73.

Franzen, Giep and Margot Bouwman (2001), The Mental World of

Brands: Mind, Memory and Brand Success. Oxfordshire: World
Advertising Research Center.

Frith, Katherine T. And Barbara Mueller (2007), Advertising and

Societies: Global Issues (3rd ed.). New York: Peter Lang, Digital
Formations Vol. 14.

Furrer, Olivier, Ben Shaw-Ching Liu, and D. Sudharsan (2000), “The

Relationships between Culture and Service Quality Perceptions: Basis
for Cross-Cultural Market Segmentation and Resource Allocation,”
Journal of Service Research, 2 (4), 355-71.

Gaál, Zoltán, Lajos Szabó and Zoltán Kovács (2007), “Culture,

Competence, Competitiveness: Managing Diversity at Individual
and Community Level,” International Journal of Diversity in
Organisations, Communities & Nations, 7 (5), 131-41.

208
Gaines, Stanley O., William D. Marelich, Katrina L. Bledsoe, W. Neil
Steers, Michael C. Henderson, and Cherlyn S. Granrose, Lisa Barajas,
Diana Hicks, Michael Lyde, Yumi Takahashi, Nancy Yum, Diana I
Rios, Ben R Garcia, Karlyn R. Farris, and Mary S. Page (1997),
“Links between Race/Ethnicity and Cultural Values as Mediated by
Racial/Ethnic Identity and Moderated by Gender,” Journal of
Personality and Social Psychology, 72 (6), 1460-76.

Goldsmith, Ronald E. (2004), “Current and Future Trends in Marketing

and their Implications for The Discipline,” Journal of Marketing
Theory & Practice, 12 (4), 10-17.

Gram, Malene (2007), “Whiteness and Western Values in Global

Advertisements: An Exploratory Study,” Journal of Marketing
Communications, 13 (4), 291-309.

Gregory, Anne (2007), “Involving Stakeholders in Developing Corporate

Brands: The Communication Dimension,” Journal of Marketing
Management, 23 (1/2), 59-73.

Grier, Sonya and Anne M. Brumbaugh (1999), “Noticing Cultural

Differences: Ad Meanings by Target and Non-target Markets,”
Journal of Advertising, 28 (1), 79-99.

Griffin, Jill (1997), Customer Loyalty – How to Earn it, How to Keep it.
San Francisco: Jossey-Bass Publishers.

Haig, Matt (2003), Brand Failures. London: Kogan Page.

Hall, Edward T. (1989), Beyond culture [Reprint]. New York: Anchor

Books.

Hampden-Turner, Charles M. and Fons Trompenaars (1996), “A World

Turned Upside Down: Doing Business in Asia,” in Managing

209
 Across Cultures: Issues and Perspectives, Pat Joynt and Malcom
Warner eds. London: International Thomson Business Press, 275–
305.

Hankinson, Graham and Philippa Cowking (1996), The Reality of Global

Brands: Cases and Strategies for Successful Management of
International Brands. Columbus: McGraw-Hill.

Hatch, Mary Jo and James Rubin (2006), “The Hermeneutics of

Branding,” Journal of Brand Management, 14 (1/2), 40-59.

Hauser, John R. and Frank S. Koppelman (1979), “Alternative Perceptual

Mapping Techniques: Relative Accuracy and Usefulness,” Journal
of Marketing Research, 16 (4), 495-506.

Heath, Robert, David Brandt, and Agnes Nairn (2006), “Brand

Relationships: Strengthened by Emotion, Weakened by Attention”
Journal of Advertising Research, 46 (4), 410-19.

Henderson, Geraldine R., Dawn Iacobucci, and Bobby J. Calder (1998),

“Brand Diagnostics: Mapping Branding Effects Using Consumer
Associative Networks,” European Journal of Operational
Research, 111 (2), 306-27.

-------, Geraldine R., Dawn Iacobucci, and Bobby J. Calder (2002), “Using
Network Analysis to Understand Brands,” Advances in Consumer
Research, 29 (1), 397-405.

Henderson, Pamela W., Joseph A. Cote, Siew Meng Leong, Bernd

Schmitt (2003), “Building Strong Brands in Asia: Selecting the
Visual Components of Image to Maximize Brand Strength,”
International Journal of Research in Marketing, 20 (4), 297-313.

210
Henry, Jim (2007), “Company's Quality Crusade Was Launched by an
American,” Automotive News, 82, (2007, October 30) 58.

Herzberg, Frederick (1968), “One More Time: How Do You Motivate

Employees?,” Harvard Business Review, 46 (1), 53-62.

Herzog, Regula A. and Jerald G. Bachman (1981), “Effects of

Questionnaire Length on Response Quality,” Public Opinion
Quarterly, 45 (4), 489–504.

Hickman, Craig and Christopher Raia (2002), “Incubating Innovation,”

Journal of Business Strategy, 23 (3), 14-18.

Hofstede Geert (1997), “Cultural constraints in management theories” in

Leadership: Understanding the dynamics of power and influence in
organizations, Robert P. Vecchio, ed. Notre Dame, IN: University
of Notre Dame Press, 465 – 83.

-------, Geert (1991), Cultures and Organization: Software of the Mind,

London: McGraw-Hill.

-------, Geert (1980), “Motivation, Leadership, and Organization: Do

American Theories Apply Abroad?,” Organizational Dynamics, 9 (1),
42-63.

-------, Geert (2001), Culture's Consequences: Comparing Values,

Behaviors, Institutions, and Organizations across Nations (2nd ed.).
Thousand Oaks, CA: Sage.

-------, Geert and Michael Harris Bond (1988), “The Confucius

Connection: From Cultural Roots to Economic Growth,”
Organizational Dynamics, 16 (4), 5-21.

211
Hollenbeck, Candice R., Cara Peters, and George M. Zinkhan (2008),
“Retail Spectacles and Brand Meaning: Insights from a Brand
Museum Case Study,” Journal of Retailing, 84 (3), 334–53.

Holt, Douglas B. (1995), “How Consumers Consume: A Typology of

Consumption Practices,” Journal of Consumer Research, 22 (1), 1-
16.

-------, Douglas B. (2002), “Why Do Brands Cause Trouble? A Dialectical

Theory of Consumer Culture and Branding,” Journal of Consumer
Research, 29 (1), 70-90.

-------, Douglas B. (2004), How Brands Become Icons: The Principles of

Cultural Branding. Boston, MA: Harvard Business School Press.

Homack Susan R. (2001), “Understanding What ANOVA Post Hoc Tests

Are, Really,” 1st February 2001, (ERIC Document Reproduction
Service No. ED449222)

Homer, Pamela M. and Sun-Gil Yoon (1992), “Message framing and the
interrelationships among ad-based feelings, affect, and cognition,” Journal of
Advertising, 21 (1), 19-33.

House, Rober, Mansour Javidan, Paul Hanges, and Peter Dorfman (2002),
“Understanding cultures and implicit leadership theories across the
globe: an introduction to project GLOBE,” Journal of World
Business, 39 (1), 3-10.

-------, Robert, Paul Hanges, Mansour Javidan, Peter Dorfman and Vipin
Gupta (2004), Culture, leadership and organizations: The GLOBE
Study of 62 Societies. Thousand Oaks, London, New Delhi: SAGE
Publications.

212
Interbrand and Business Week (2008), “Best Global Brands 2008,”
(accessed December 5, 2008), [available at
(http://www.interbrand.com/best_global_brands.aspx?year=2008&typ
e=desc&col=7&langid=1000].

Jamieson, Susan (2004), “Likert scales: How to (ab)use them,” Medical

Education, 38 (12), 1217-18.

Jun, Jong Woo and Hyung-Seok Lee (2007), “Cultural Differences in

Brand Designs and Tagline Appeals,” International Marketing
Review, 24 (4), 474-91.

-------, Jong Woo, Chang-Hoan Cho, and Hyuck Joon Kwon (2008), “The
Role of Affect and Cognition in Consumer Evaluations of Corporate
Visual Identity: Perspectives from the United States and Korea,”
Journal of Brand Management, 15 (6), 382-98.

Kapferer, Jean-Noel (2004), The New Strategic Brand Management (3rd

ed.). London: Kogan Page.

Kates, Steven M. and Charlene Goh (2003), “Brand Morphing:

Implications for Advertising Theory and Practice,” Journal of
Advertising, 32 (1), 59-68.

Kay, Mark J. (2006), “Strong Brands and Corporate Brands,” European

Journal of Marketing, 40 (7/8), 742-60.

Kaynak, Erdener, Gulberk Gultekin Salman, and Ekrem Tatoglu (2008),

“An integrative framework linking brand associations and Brand
Loyalty in Professional Sports,” Journal of Brand Management, 15
(5), 336-57.

Keller, Kevin L. (2008), Best Practice Cases in Branding: Lessons from

the World’s Strongest Brands. Zagreb: Accent.

213
-------, Kevin L. (2003a), “Brand Synthesis: The Multidimensionality of
Brand Knowledge,” Journal of Consumer Research, 29 (4), 595-600.

-------, Kevin Lane and Donald R Lehmann (2006), “Brands and

Branding: Research Findings and Future Priorities,” Marketing
Science, 25 (6), 740-59.

-------, Kevin L. (1993), “The Effects of Corporate Branding Strategies on

Brand Equity,” Advances in Consumer Research, 20 (1), 27-27.

-------, Kevin L. (2003b), Strategic Brand Management: Building,

Measuring and Managing Brand Equity (2nd ed.). Upper Saddle
River, NJ: Pearson Education.

Kempf, Deanne S and Robert E. Smith (1998), “Consumer Processing of

Product Trial and the Influence of Prior Advertising: A Structural
Modeling Approach,” Journal of Marketing Research, 35 (3), 325-38.

Khalid, Halimahtun M. and Martin G. Helander (2004), “A Framework

for Affective Customer Needs in Product Design,” Theoretical Issues
in Ergonomics Science, 5 (1), 27-42.

Kim, Jooyoung, Jon D. Morris, and Joffre Swait (2008), “Antecedents of

True Brand Loyalty,” Journal of Advertising, 37 (2), 99–117.

Kim, W. Chan and Renee Mauborgne (2000), “Knowing a Winning

Business Idea When You See One,” Harvard Business Review, 78
(5), 129-138.

Kinra, Neelam (2006), “The Effect of Country-of-origin on Foreign Brand

Names in the Indian Market,” Marketing Intelligence & Planning,
24 (1), 15-30.

Klein, Naomi (2000), No Logo: No Space, No Choice, No Jobs. London:

Flamingo.

214
Kluckhohn, Clyde (1951), “The Study of Culture,” in The Policy
Sciences, Daniel Lerner and Harold D. Lasswell, eds. Stanford, CA:
Stanford University Press, 86-101.

Kluckhohn, R. Florence and Fred L. Strodtbeck (1961), Culture: A

Critical Review of Concepts and Definitions. Cambridge, MA:
Harvard University.

Krishnan, H.S. (1996), “Characteristics of Memory Associations: A

Consumer-Based Brand Equity Perspective,” International Journal
of Research in Marketing, 13 (October), 389-05.

Kumar, Archana, Hyun-Joo Lee, and Youn-Kyung Kim (2008), “Indian

Consumers' Purchase Intention toward a United States versus Local
Brand,” Journal of Business Research, In Press, Available online 24
September 2008 (doi:10.1016/j.jbusres.2008.06.018)

Lange, Fredrik and Micael Dahlen (2003), “Let's Be Strange: Brand

Familiarity and Ad-Brand Incongruency,” Journal of Product &
Brand Management, 12 (7), 449-461.

Langley, Ann (1999), “Strategies for Theorizing from Process Data,”

Academy of Management Review, 24 (4), 691-710.

LeBel, Jordan L. and Nathalie Cooke (2008), “Branded Food

Spokescharacters: Consumers' Contributions to the Narrative of
Commerce,” Journal of Product & Brand Management, 2008, 17 (3),
143-53.

Lee, Eun-Jung and Eun-Young Rhee (2008), “Conceptual Framework of

Within-Category Brand Personality Based on Consumers’ Perception
(WCBP-CP): The Case of Men’s Apparel Category in South Korea,”
Journal of Brand Management, 15 (6), 465-89.

215
Lee, Min-Young, Dee Knight, and Youn-Kyung Kim (2008), “Brand
Analysis of a US Global Brand in Comparison with Domestic Brands
in Mexico, Korea, and Japan,” Journal of Product & Brand
Management, 17 (3), 163-74.

Leo, Cheryl, Rebekah Bennett, and Charmine E. J. Härtel (2005), “Cross-

Cultural Differences in Consumer Decision-Making Styles,” An
International Journal of Cross Cultural Management, 12 (13), 32-62.

LePla, Joseph F. and Lynn M. Parker (2002), Integrated branding:

Becoming Brand-driven through Company-wide Action. London:
Kogan Page.

Levitt, Theodor (1983), “The Globalization of Markets,” Harvard

Business Review, 61 (3), 92-101.

Ligas, Mark and June Cotte (1999), “The Process of Negotiating Brand
Meaning: A Symbolic Interactionist Perspective,” in Advances in
Consumer Research, Vol. 26, no 1, Eric J. Arnould, and Linda M.
Scott, eds. Provo, UT: Association for Consumer Research, 609-14.

Lim, Kenny and O. Cass Aron (2001), “Consumer Brand Classifications:

An Assessment of Culture-of-Origin versus Country-of-Origin,”
Journal of Product & Brand Management, 10 (2), 120-36.

Lindstrom, Martin (2005), Build Powerful Brands through Touch, Taste

Smell, Sight and Sound. New York: Free Press.

Low, George S. and Charles W. Lamb Jr. (2000), “The Measurement and
Dimensionality of Brand Associations,” Journal of Product & Brand
Management, 9 (6), 350-68.

216
Luo, Xueming, K. Sivakumar and Sandra Liu (2005), “Globalization,
Marketing Resources, and Performance: Evidence from China,”
Journal of the Academy of Marketing Science, 33 (1), 50-65.

Martin, Ingrid, David Stewart, and Shashi Matta (2005), “Branding

Strategies, Marketing Communication, and Perceived Brand
Meaning: The Transfer of purposive, goal-oriented brand meaning
to Brand Extensions,” Journal of the Academy of Marketing
Science, 33 (3), 275-94.

Maslow, Abraham H. (1943) “A Theory of Human Motivation,”

Psychological Review, 50, 370-96.

Maznevski, Martha L, Joseph J. DiStefano, Gomez, Carolina B.,

Noorderhaven, Niels G, and Pei-Chuan Wu (2002), “Cultural
Dimensions at the Individual Level of Analysis,” International
Journal of Cross Cultural Management, 2 (3), 275-95.

McAlexander, James H., John W. Schouten, and Harold F. Koenig (2002),

“Building Brand Community,” Journal of Marketing, 66 (1), 38-54.

McClelland, David C. (1961), The Achieving Society. Princeton, NJ: Van

Nostrand.

McCracken, Grant (1986) “Culture and Consumption: A Theoretical

Account of the Structure and Movement of the Cultural Meaning of
Consumer Goods,” Journal f Consumer Research, 13 (June), 71-84.

Miliopoulou, Zozeta (2007), “Brand as A Symbolic Token: A Theoretical

Framework for the Consumer-Brand Relationship,” paper presented
at 2007 Thought Leaders International Conference on Brand
Management, Birmingham Business School, University of
Birmingham (April 24-25).

217
Monga, Alokparna Basu and Deborah Roedder John (2007), “Cultural
Differences in Brand Extension Evaluation: The Influence of Analytic
versus Holistic Thinking,” Journal of Consumer Research, 33 (4)
529-36.

Montgomery, Cynthia A., Birger Wernerfelt, and Srinivasan Balakrishnan

(1989), “Strategy Content and The Research Process: A Critique and
Commentary,” Strategic Management Journal, 10 (2), 189-97.

Moore, David J. and Pamela Miles Homer (2008), “Self-brand

Connections: The Role of Attitude Strength and Autobiographical
Memory Primes,” Journal of Business Research, 61 (7), 707-14.

Morgan, Susan E. and Tom Reichert (1999), “The Message Is in the

Metaphor: Assessing the Comprehension of Metaphors in
Advertisements,” Journal of Advertising, 28 (4), 1-12.

Moriarty, Sandra E. and Thomas R. Duncan (1990), “Global Advertising:

Issues and Practices,” Current Issues & Research in Advertising, 13
(2), 313-13.

Muniz Jr. Albert M. and Thomas C. O'Guinn (2001), “Brand

Community,” Journal of Consumer Research, 27 (4), 412-12.

-------, Jr. Albert M. and Hope Jensen Schau (2005), “Religiosity in the
Abandoned Apple Newton Brand Community,” Journal of Consumer
Research, 31 (4), 737-47.

Nairn Agnes, Christine Griffin, and Patricia Gaya Wicks (2008),

“Children’s Use of Brand Symbolism A Consumer Culture Theory
Approach,” European Journal of Marketing, 42 (5/6), 627-40.

218
Nisbett, Richard E., Kaiping Peng, Incheol Choi, and Ara Norenzayan
(2001), “Culture and Systems of Thought: Holistic versus Analytic
Cognition,” Psychological Review, 108 (2), 291-310.

Nunnally, Jum C. (1978), Psychometric Theory. New York: McGraw-Hill

Book Company.

Oakenfull, Gillian, Edward Blair, Betsy Gelb, and Peter Dacin (2000),
“Measuring Brand Meaning,” Journal of Advertising Research, 40
(5), 43-53.

O'Reilly, Daragh (2005), “Cultural Brands / Branding Cultures,” Journal

of Marketing Management, 21 (5/6), 573-88.

Orth, Ulrich R. and De Marchi, Renata (2007), “Understanding the

Relationships Between Functional, Symbolic, and Experiential Brand
Beliefs, Product Experiential Attributes, and Product Schema:
Advertising-Trial Interactions Revisited,” Journal of Marketing
Theory & Practice, 15 (3), 219-33.

-------, Ulrich R., Harold F. Koenig, and Zuzana Firbasova (2007), “Cross-
national Differences in Consumer Response to the Framing of
Advertising Messages An Exploratory Comparison from Central
Europe,” European Journal of Marketing,. 41 (3/4), 327-48.

Packer, Martin J. (1985), “Hermeneutic Inquiry in the Study of Human

Conduct,” American Psychologist, 40 (10), 1081-93.

Pankhania, Asha, Nick Lee, and Graham Hooley (2007), “Within-country

Ethnic Differences and Product Positioning: a Comparison of the
Perceptions of Two British sub-cultures,” Journal of Strategic
Marketing, 15 (2/3), 121-38.

219
Park, C. Whan, Bernard J. Jaworski, and Deborah J. Maclnnis (1986),
“Strategic Brand Concept-Image Management,” Journal of
Marketing, 50 (4), 135-45.

-------, C. Whan, Sandra Milberg, and Robert Lawson (1991), “Evaluation

of Brand Extensions: The Role of Product Feature Similarity and
Brand Concept Consistency,” Journal of Consumer Research, 18 (2),
185-85.

-------, C. Whan and Deborah J Macinnis (2006), “What's In and What's

Out: Questions on the Boundaries of the Attitude Construct,”
Journal of Consumer Research, 33 (1), 16-18.

Pawle, John and Peter Cooper (2006), “Measuring Emotion -- Lovemarks,

The Future Beyond Brands,” Journal of Advertising Research, 46 (1),
38-48.

Peter, Paul J. and Jerry C. Olson (2005), Consumer behaviour and

Marketing strategy, (7th ed.). New York, NJ: McGrow Hill.

Peters, Rolf-Herbert (2008), The Puma Story: The Remarkable

Turnaround of an Endangered Species into one of the World's
Hottest Sportlifestyle Brands. London: Marshall Cavendish.

Phau, Ian I.P. and Kong Cheen K.L. Lau (2000), “Conceptualising Brand
Personality: A Review and Research Propositions,” Journal of
Targeting, Measurement and Analysis for Marketing, 9 (1), 52-69.

Plassmann, Hilke, Tim Ambler, Sven Braeutigam, and Peter Kenning

(2007), “What can Advertisers Learn from Neuroscience?,”
International Journal of Advertising, 26 (2), 151-75.

Podsakoff, Philip M., Scott B. MacKenzie, Jeong-Yeon Lee and Nathan P.

Podsakoff (2003), “Common Method Biases in Behavioral

220
 Research: A Critical Review of the Literature and Recommended
Remedies,” Journal of Applied Psychology, 88 (5), 879-903.

Polegato, R. and R. Bjerke (2006), “The Link between Cross-Cultural

Value Associations and Liking: The case of Benetton and its
advertising,” Journal of Advertising Research, 46 (3), 263-73.

Pringle, Hamish and Marjorie Thompson (1999), Brand Spirit: How

Cause Related Marketing Builds Brands. New York: John Wiley &
Sons.

Quester, Pascale, Michael Beverland, and Francis Farrelly (2006), “Brand-

Personal Values Fit and Brand Meanings: Exploring the Role
Individual Values Play in Ongoing Brand Loyalty in Extreme Sports
Subcultures” Advances in Consumer Research, 33 (1), 21-27.

Raggio, Randle D. and Robert P. Leone (2007), “The Theoretical

Separation of Brand Equity and Brand Value: Managerial
Implications for strategic planning,” Journal of Brand Management,
14 (5), 380–95.

Raiklin, Ernest and Bulent Uyar (1996), “On The Relativity of The
Concepts of Needs, Wants, Scarcity and Opportunity Cost,”
International Journal of Social Economics, 23 (7), 49-56.

Ramberg, Bjørn and Kristin Gjesdal (2009), “Hermeneutics,” in The

Stanford Encyclopedia of Philosophy (Spring 2009 Edition), Edward
N. Zalta, ed. (accessed 4th January 2008) [available at
http://plato.stanford.edu/archives/spr2009/entries/hermeneutics/].

Raymond, Mary Anne, John D. Mittelstaedt, and Christopher D. Hopkins

(2003), “When Is a Hierarchy Not a Hierarchy? Factors Associated
with Different Perceptions of Needs, With Implications for

221
 Standardization - Adaptation Decisions in Korea,” Journal of
Marketing Theory & Practice, 11 (4), 12-25.

Reynolds, Thomas J. and Jonathan Gutman (1988), “Laddering Theory,

Method, Analysis, and Interpretation,” Journal of Advertising
Research, 28 (1), 11-31.

Richins, Marsha L. (1994), “Valuing Things: The Public and Private

Meanings of Possessions,” Journal of Consumer Research, 21 (3),
504-04.

Riesenbeck Hajo and Jesko Perrey (2007), Power Brands. Weinheim:

Wiley – VCH Verlag.

Rindfleisch, Aric, Alan J. Malter, Shankar Ganesan and Christine

Moorman (2008), “Cross-Sectional versus Longitudinal Survey
Research: Concepts, Findings, and Guidelines,” Journal of
Marketing Research, 45 (3) 261-79.

Roehm, Michelle L.; Brady, Michael K. (2007), “Consumer Responses to

Performance Failures by High-Equity Brands,” Journal of Consumer
Research, 34 (4), 537-45.

Romani, Laurence (2004), “Culture in Management: the Measurement of

Differences” in International Human Resource Management, Anne-
Wil Harzing and Ruysseveldt, Eds. 2 ed. London: Sage Publications.

Ross, Jill and Rod Harradine (2004), “I'm Not Wearing That!: Branding
and Young Children,” Journal of Fashion Marketing & Management,
8 (1), 11-26.

Roth, Martin S. (1995), “The Effects of Culture and Socioeconomics on

The Performance of Global Brand Image Strategies,” Journal of
Marketing Research, 32 (2), 163-63.

222
Ruigrok, Winfried, Michael Gibbert, and Barbara Kaes (2005), “In Search
of Rigorous Case Studies: Patterns of Validity and Reliability across
Ten Management Journals 1995-2000,” Working paper, University of
St. Gallen.

Ryan M. Richard and Edward L. Deci (2000), “Self-Determination Theory

and the Facilitation of Intrinsic Motivation, Social Development, and
Well-Being,” American Psychologist, 55 (1), 68-78.

Sawyer, Alan G. and Dwayne, A. Ball (1981), “Statistical Power and

Effect Size in Marketing Research,” Journal of Marketing Research,
18 (3), 275-90.

Schein, Edgar H. (1988), Organizational Culture and Leadership. San

Francisco: Jossey-Bass.

Schimmack, Ulrich, Shigehiro Oishi, and Ed Diener (2005),

“Individualism: A Valid and Important Dimension of Cultural
Differences Between Nations,” Personality & Social Psychology
Review, 9 (1), 17-31.

Schmitt, Bernd H. (1999), Experiential marketing: How To Get Customers

To Sense, Feel, Think, Act and Relate to Your Company and Brands.
New York: The Free Press.

Schönemann, Peter. H. (1990), “Facts, Fictions, and Common-Sense

About Factors and Components,” Multivariate Behavioral Research,
25 (1), 47-51.

Schroeder, Jonathan E. (2005), “The Artist and The Brand,” European

Journal of Marketing, 39 (11/12), 1291-305.

223
Schuh, Arnold (2007), “Brand Strategies of Western MNCs as Drivers of
Globalization in Central and Eastern Europe,” European Journal of
Marketing, 41 (3/4), 274-91.

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

Implications for Work,” Applied Psychology: An International
Review, 48 (1), 23-47.

Scitovsky, Tibor (1986), Human Desire and Economic Satisfaction.

Brighton: Wheatsheat.

Sheth, Jagdish N., Bruce I. Newman, and Barbara L. Gross (1991), “Why
We Buy What We Buy: A Theory of Consumption Values,” Journal
of Business Research, 22 (2), 159-70.

Shook, Christopher L., David J. Ketchen Jr, Cynthia S. Cycyota, and

Dilene Crockett (2003), “Data Analytic Trends and Training in
Strategic Management,” Strategic Management Journal, 24 (12),
1231-37.

Smith, Robert E. and William R. Swinyard (1983), “Attitude-Behavior

Consistency: The Impact of Product Trial versus Advertising,”
Journal of Marketing Research, 20 (3), 257-67.

Snow, Charles C. and James B. Thomas (1994), “Field Research Methods

in Strategic Management: Contributions to Theory Building and
Testing,” Journal of Management Studies, 31 (4), 457-80.

Srinivasan, V. (1979), “Network Models for Estimating Brand-Specific

Effects in Multi-Attribute Marketing Models,” Management Science,
25 (1), 11-21.

224
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.

Stern Barbara B. (2006), “What Does Brand Mean? Historical-Analysis

Method and Construct Definition,” Journal of Academy of
Marketing Science, 34 (2), 216-23.

Sung Youngjun and Spencer F. Tinkham (2005), “Brand Personality

Structures in the United States and Korea: Common and Culture-
Specific Factors,” Journal of Consumer Psychology, 15 (4), 334-
50.

Sweeney C. Jillian and Brandon Carol (2006), “Brand Personality:

Exploring the Potential to Move from Factor Analytical to
Circumplex Models,” Psychology & Marketing, 23 (8), 639-63.

------- C. Jillian and Geoffrey N. Soutar (2001), “Consumer perceived

value: The development of a multiple item scale,” Journal of
Retailing, 77 (2), 203–220.

Synodinos, Nicolaos E. Charles F. Keown, Laurence W. Jacobs (1989),

“Transnational Advertising Practices: A Survey of Leading Brand
Advertisers in Fifteen Countries,” Journal of Advertising Research,
29 (2), 43-50.

Szymanski, David M., Sundar G. Bharadwaj, and Rajan P. Varadarajan

(1993), “Standardization versus Adaptation of International
Marketing Strategy: An Empirical Investigation,” Journal of
Marketing, 57 (4), 1-17.

Tabachnick, Barbara G. and Linda S. Fidell (2001), Using Multivariate

Statistics. Boston: Allyn and Bacon.

225
Taylor, Edward (1871), Primitive cultures. London: John Murray.

Teather, David (2004), “Coca-Cola Reduces Profit Targets,”

The Guardian (November 12, 2004) (accessed 15th December 2008)
[available at: http://www.indiaresource.org/news/2004/1053.html].

Thompson, Craig J. (2004), “Beyond Brand Image: Analyzing the Culture

of Brands,” Advances in Consumer Research, 31 (1), 98-100.

-------, Craig J., Aric Rindfleisch, and Zeynep Arsel (2006), “Emotional
Branding and the Strategic Value of the Doppelgaenger Brand
Image,” Journal of Marketing, 70 (1), 50-64.

-------, Craig J., Howard R Pollio, and William B. Locander (1994), “The
Spoken and the Unspoken: A Hermeneutic Approach to
Understanding the Cultural Viewpoints That Underlie Consumers'
Expressed Meanings,” Journal of Consumer Research, 21 (3), 432-
52.

Thompson, Jim (2008), “Brandjunkies on the Influence of Brands: The

Brandjunkie Survey Result,” Brandchannel (accessed 12th
November, 2008) [available at
http://www.brandchannel.com/features_effect.asp?pf_id=415]

Triandis, Harry C. and Michael J. Gelfand (1998), “Converging

Measurement of Horizontal and Vertical Individualism and
Collectivism,” Journal of Personality and Social Psychology, 74 (1),
118-28.

Trochim William and James P. Donnelly (2007), The Research Methods

Knowledge Base (3rd ed.). Mason, OH: Atomic Dog Publishing.

Trompenaars, Fons (1994), Riding the Waves of Culture. London:

Nicholas Brealey.

226
Vargo, Stephen L and Robert F. Lusch (2004), “Evolving to a New
Dominant Logic for Marketing,” Journal of Marketing, 68 (1), 1-17.

Velicer, Wayne F. and Douglas N. Jackson (1990), “Component Analysis

versus Common Factor-Analysis – Some Further Observations,”
Multivariate Behavioral Research, 25 (1), 97-114.

Veloutsou, Cleopatra (2007), “Identifying the Dimensions of the Product-

Brand and Consumer Relationship,” Journal of Marketing
Management, 23 (1/2), 7-26.

-------, Cleopatra (2008), “Branding: A Constantly Developing concept,”

Journal of Brand Management, 15 (5), 299-300.

-------, Cleopatra and Luiz Moutinho (2008), “Brand Relationships

Through Brand Reputation and Brand Tribalism,” Journal of Business
Research, In Press, Available online 6 August 2008
(doi:10.1016/j.jbusres.2008.05.010)

Verhage, Bronislaw J., Ugur Yavas, and Robert T. Green (1990),

“Perceived Risk: A Cross-cultural Phenomenon,” International
Journal of Research in Marketing, 7 (4), 297-303.

Ward, James C. and Amy L. Ostrom (2003), “The Internet as Information

Minefield: An Analysis of the Source and Content of Brand
Information Yielded by Net Searches,” Journal of Business Research,
56 (11) 907-14.

Watson, John, Steve Lysonski, Tamara Gillan, and Leslie Raymore

(2002), “Cultural Values and Important Procession: A Cross-cultural
Analysis,” Journal of Business Research, 55 (11), 923–31.

Weick, E. Karl (1995), “What Theory Is Not, Theorizing Is,”

Administrative Science Quarterly, 40 (3), 385-90.

227
Wilcox, Rand (2002), “Methodological article: Understanding the
Practical Advantages of Modern ANOVA Methods,” Journal of
Clinical Child and Adolescent Psychology, 31 (3), 399-412.

Wipperfurth, Alex (2006), Brand Hijack: Marketing Without Marketing:

New York: Portfolio Trade.

Yin, K. Robert Ed. (1998), The Abridged Version of Case Study Research.
Thousand Oaks, CA: Sage.

Zaltman, Gerald (2003), How Customers Think: Essential Insights into the
Mind of the Market. Boston: Harvard Business School Press.

Zhang, Yong and James P. Neelankavil (1997), “The Influence of Culture

on Advertising Effectiveness in China and the USA,” European
Journal of Marketing, 31 (1/2), 134-49.

Zhou, Lianxi, Lefa Teng, and Patrick S. Poon (2008), “Susceptibility to

Global Consumer Culture: A Three-Dimensional Scale,” Psychology
& Marketing, 25 (4), 336-51.

Zumpano, Anthony (2007), “Similar Search Results: Google Wins”,

Brandchannel, (accessed 12th November, 2008) [available at
http://www.brandchannel.com/features_effect.asp?pf_id=352]

228
Appendices

I
 Appendix 1. The Images Collected in the Interviews
Table ap1-1. Visual Indications of What Coca-cola Means and Does
Not Mean to the Interviewees
Interviewee Images that indicate what Images that indicate what
Coca-cola means to Coca-cola does not mean to
interviewees interviewees
Hanna
(Finland)

Minna
(Finland)

II
Tiia
(Finland)

Kari
(Finland)

Hannu
(Finland)

Anisha
(India)

Praveen
(India)

III
 Samyuktha
(India)

Hemant
(India)

Himanshu
(India)

Jane
(US)

Samantha
(US)

IV
Lucy
(US)

Robert
(US)

Peter
(Austria)

V
 Michaela
(Austria)

Christina
(Austria)

Ronald
(Austria)

Petra
(Austria)

VI
Table ap1-2. Visual Indications of What Google Means and Does Not
Mean to the Interviewees
Interviewee Images that indicate what Images that indicate what
Google means to interviewees Google does not mean to
interviewees
Hanna
(Finland)

Minna
(Finland)

Tiia
(Finland)

Kari
(Finland)

VII
 Hannu
(Finland)

Anisha
(India)

Praveen
(India)

Samyuktha
(India)

Hemant
(India)

Himanshu
(India)

VIII
Jane
(US)

Samantha
(US)

Lucy
(US)

Robert
(US)

IX
Peter
(Austria)

Michaela
(Austria)

Christina
(Austria)

Ronald
(Austria)

Petra
(Austria)

X
Table ap1-3. Visual Indication of what Nike Means and Does Not
Mean to the Interviewees
Interviewee Images that indicate what Images that indicate what
Nike means to interviewees Nike does not mean to
interviewees
Hanna
(Finland)

Minna
(Finland)

Tiia
(Finland)

XI
 Kari (Finland)

Hannu
(Finland)

Anisha
(India)

Praveen
(India)

XII
Samyuktha
(India)

Hemant
(India)

Himanshu
(India)

Jane
(US)

XIII
Samantha
(US)

Lucy
(US)

Robert
(US)

XIV
Peter
(Austria)

Michaela
(Austria)

Christina
(Austria) Cheap prices

Ronald
(Austria)

XV
Petra
(Austria)

XVI
Table ap1-4. Visual Indication of What Nokia Means and Does Not
Mean to the Interviewees
Interviewee Images that indicate what Images that indicate what
Nokia means to interviewees Nokia does not mean to
interviewees
Hanna
(Finland)

Minna
(Finland)

Tiia
(Finland)

Kari (Finland)

Hannu
(Finland)

XVII
Anisha
(India)

Praveen
(India)

Samyuktha
(India)

XVIII
Hemant
(India)

Himanshu
(India)

Jane
(US)

Samantha
(US)

Lucy
(US)

XIX
Robert n.a.
(US)
Peter
(Austria)

Michaela
(Austria)

Christina
(Austria)

Ronald
(Nokia)

Petra
(Austria)

XX
Table ap1-5. Visual Indication of What Toyota Means and Does Not
Mean to the Interviewees
Interviewee Images that indicate what Images that indicate what
Toyota means to Toyota does not mean to
interviewees interviewees
Hanna
(Finland)

Minna
(Finland)

Tiia
(Finland)

Kari (Finland)

Hannu
(Finland)

XXI
Anisha
(India)

Praveen
(India)

Samyuktha
(India)

Hemant
(India)

Himanshu
(India)

XXII
Jane
(US)

Samantha
(US)

Lucy
(US)

Robert
(US)

XXIII
Peter
(Austria)

Michaela
(Austria)

Christina
(Austria)

Ronald
(Austria)

Petra
(Austria)

XXIV
Appendix 2. The Product Categories and Images Used in
the Pilot Phase Survey
Table ap2-1. The Product Categories for Coca-cola, Google and Nike
Coca-cola Google Nike
Mineral water City guides Sports TV
Cakes Space shuttles Training on how to win presidential
Cigarettes Fashion items campaign
Alcohol In class lectures on Sleepwear
Toys global issues Spa centres
Ice cream Life vests Music
Energy bars Books Luxurious handbags
Jeans GPS navigation Diamond mining in Africa
Playing cards gadgets High quality furniture
Artificial fertilizers Info stalls in the cities Adventurous travels agency
World sports games Sandals
Virtual big brother Cushions for chairs
show Innovative electrical appliances
Music Jogging fields
Art Sport utility vehicles (SUVs)
Training people how to achieve what
they though was impossible

Table ap2-2. The Images for Coca-cola

Table ap2-3. The Images for Google

Table ap2-4. The Images for Nike

XXV
Appendix 3. The Questionnaire
Cross-cultural Branding
I. Default Section

Thank you for participating in my research on differences in brand meaning

among cultures. This questionnaire will take about 10 minutes of your time. The
answers are anonymous.

1. Which University do you attend? ____________________________

2. What is your major? ____________________________

3. What type of degree are you studying for?

a) BSc b) BA c) MSc d) MBA e) other

4. Which semester are you currently enrolled in?

a) 1 b) 3 c) 5 d) 7 e) 9 f) other

5. How old are you? ____________________________

6. Have you lived abroad for longer than 6 months at a stretch?

a) yes b) no

7. If the answer is yes, please indicate when, where, and for how long.
________________________________________________________

8. Do you use the following brands?

Coca-Cola a) yes b) no
Google a) yes b) no
Nike a) yes b) no

II. Coca-Cola

Coca-Cola is a drink we often encounter. Some of us like to drink it and some of

us don’t. As you look at the following questions, please answer them by thinking
about how it feels to experience Coca-Cola.

9. What does Coca-Cola mean to you?

____________________________

10. To what extent do the associations listed below indicate what Coca-Cola
means to you?
It means very little it means a lot
American 1 2 3 4 5 6 7
traditional 1 2 3 4 5 6 7
worldwide 1 2 3 4 5 6 7
classic 1 2 3 4 5 6 7
warm summer 1 2 3 4 5 6 7

XXVI
 original 1 2 3 4 5 6 7
American lifestyle 1 2 3 4 5 6 7
Christmas 1 2 3 4 5 6 7
calories 1 2 3 4 5 6 7
drink it when tired 1 2 3 4 5 6 7
hot day 1 2 3 4 5 6 7
bonding with friends 1 2 3 4 5 6 7
global 1 2 3 4 5 6 7
healthy 1 2 3 4 5 6 7
unknown 1 2 3 4 5 6 7
recognisable 1 2 3 4 5 6 7
cold 1 2 3 4 5 6 7
fun 1 2 3 4 5 6 7
identifiable 1 2 3 4 5 6 7
obesity 1 2 3 4 5 6 7
sugar 1 2 3 4 5 6 7
Christmas song 1 2 3 4 5 6 7
unhealthy 1 2 3 4 5 6 7
energetic 1 2 3 4 5 6 7
freshness 1 2 3 4 5 6 7
Santa Claus 1 2 3 4 5 6 7
exciting 1 2 3 4 5 6 7
no sleeping 1 2 3 4 5 6 7
cooling 1 2 3 4 5 6 7
familiar 1 2 3 4 5 6 7
natural 1 2 3 4 5 6 7
parties 1 2 3 4 5 6 7
multinational 1 2 3 4 5 6 7
keeps me going 1 2 3 4 5 6 7
boring 1 2 3 4 5 6 7
cool 1 2 3 4 5 6 7
sweet 1 2 3 4 5 6 7
great brand 1 2 3 4 5 6 7
refreshing 1 2 3 4 5 6 7

10. What do you feel for Coca-Cola?

a) I love it b) I like it c) I am indifferent d) I don’t like it e) I hate it

III. Google

When something smells ‘fishy’ and you want to check whether it holds true, you
will most likely hear a voice urging you to Google. With Google in mind, please
answer the following questions.

11. What does Google mean to you?

____________________________

XXVII
12. To what extent do the associations listed below indicate what Google means
to you?
It means very little it means a lot
fun 1 2 3 4 5 6 7
fast 1 2 3 4 5 6 7
old 1 2 3 4 5 6 7
world 1 2 3 4 5 6 7
powerful 1 2 3 4 5 6 7
life saver 1 2 3 4 5 6 7
no privacy 1 2 3 4 5 6 7
useful 1 2 3 4 5 6 7
entertainment 1 2 3 4 5 6 7
happiness 1 2 3 4 5 6 7
huge 1 2 3 4 5 6 7
connecting with the 1 2 3 4 5 6 7
easy ldto use 1 2 3 4 5 6 7
reliable 1 2 3 4 5 6 7
provides data 1 2 3 4 5 6 7
new 1 2 3 4 5 6 7
worldwide 1 2 3 4 5 6 7
basic necessity 1 2 3 4 5 6 7
freedom 1 2 3 4 5 6 7
helpful 1 2 3 4 5 6 7
smart 1 2 3 4 5 6 7
convenient 1 2 3 4 5 6 7
trustful 1 2 3 4 5 6 7
all knowing 1 2 3 4 5 6 7
high speed 1 2 3 4 5 6 7
global 1 2 3 4 5 6 7
almighty 1 2 3 4 5 6 7
very important 1 2 3 4 5 6 7
no independency 1 2 3 4 5 6 7
no problem 1 2 3 4 5 6 7
informative 1 2 3 4 5 6 7
knowledge 1 2 3 4 5 6 7
observes people's habits 1 2 3 4 5 6 7
encyclopaedia 1 2 3 4 5 6 7
At any time 1 2 3 4 5 6 7
efficient 1 2 3 4 5 6 7
young 1 2 3 4 5 6 7
dominant 1 2 3 4 5 6 7
innovative 1 2 3 4 5 6 7
creative 1 2 3 4 5 6 7

XXVIII
 satisfaction 1 2 3 4 5 6 7
multibillion industry 1 2 3 4 5 6 7

13. What do you feel for Google?

a) I love it b) I like it c) I am indifferent d) I don’t like it e) I hate
it

IV. Nike

Most of us have seen a commercial speaking about features and benefits a new
Nike product offers. Whether you have worn Nike shoes or not, please answer
the following questions.

14. What does Nike mean to you?

____________________________

15. To what extent do the associations listed below indicate what Nike means to
you?
It means very little it means a lot
sports 1 2 3 4 5 6 7
achiever 1 2 3 4 5 6 7
freedom 1 2 3 4 5 6 7
athletes 1 2 3 4 5 6 7
cool 1 2 3 4 5 6 7
expensive 1 2 3 4 5 6 7
sweat shop 1 2 3 4 5 6 7
good quality 1 2 3 4 5 6 7
lazy 1 2 3 4 5 6 7
life 1 2 3 4 5 6 7
crowd cheering 1 2 3 4 5 6 7
comfortable 1 2 3 4 5 6 7
jogging 1 2 3 4 5 6 7
Michael Jordan 1 2 3 4 5 6 7
powerful 1 2 3 4 5 6 7
self-assured 1 2 3 4 5 6 7
athletic 1 2 3 4 5 6 7
trendy 1 2 3 4 5 6 7
pricey 1 2 3 4 5 6 7
child labour 1 2 3 4 5 6 7
low quality 1 2 3 4 5 6 7
just do it 1 2 3 4 5 6 7
fitness 1 2 3 4 5 6 7
innovative 1 2 3 4 5 6 7
running 1 2 3 4 5 6 7

XXIX
 sporty 1 2 3 4 5 6 7
victory 1 2 3 4 5 6 7
failure 1 2 3 4 5 6 7
young 1 2 3 4 5 6 7
durable 1 2 3 4 5 6 7
air technology 1 2 3 4 5 6 7
socially responsible 1 2 3 4 5 6 7
innovation 1 2 3 4 5 6 7
winning 1 2 3 4 5 6 7
weak 1 2 3 4 5 6 7
Tiger Woods 1 2 3 4 5 6 7
positive thinking 1 2 3 4 5 6 7

16. What do you feel for Nike?

a) I love it b) I like it c) I am indifferent d) I don’t like it
e) I hate it

17.If you have any general comments, please write them in the space below:
_____________________________________________________________
_____________________________________________________________
_____________________________________________________________

XXX
 Appendix 4. The Statistics
Scree Plot

10

8
Eigenvalue

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39

Component Number

Figure ap4-1. The Initial Scree plot for Coca-Cola

Table ap4-1. The Communalities for Coca-Cola

Extraction Extraction
American .475 sugar .628
traditional .720 Christmas song .795
worldwide .631 unhealthy .656
classic .667 energetic .551
warm summer .639 freshness .659
original .448 Santa Claus .767
Christmas .764 exciting .609
calories .626 no sleeping .448
drink it when tired .659 cooling .712
hot day .636 familiar .371
bonding with friends .717 not natural .421
global .710 parties .565
not healthy .466 multinational .649
not unknown .292 keeps me going .650
recognisable .476 cool .649
cold .537 sweet .457
fun .640 great brand .419
identifiable .437 refreshing .768
obesity .642
Extraction Method: Principal Component Analysis.

XXXI
Table ap4-2. The Total Variance Explained for Coca-Cola
Rotation Sums
Extraction Sums of Squared of Squared
Initial Eigenvalues Loadings Loadings a
% of Cumulative % of Cumulative
Component Total Variance % Total Variance % Total
1 8.339 22.539 22.539 8.339 22.539 22.539 5.611
2 3.959 10.701 33.240 3.959 10.701 33.240 3.447
3 2.827 7.639 40.879 2.827 7.639 40.879 3.973
4 2.332 6.302 47.181 2.332 6.302 47.181 3.616
5 1.620 4.378 51.559 1.620 4.378 51.559 2.840
6 1.554 4.200 55.759 1.554 4.200 55.759 3.945
7 1.326 3.584 59.343 1.326 3.584 59.343 4.398
8 1.118 3.022 62.365
9 0.968 2.616 64.981
Extraction Method: Principal Component Analysis.
a
When components are correlated sums of squared loadings cannot be added to obtain a
total variance.

Table ap4-3. The Structure Matrix for Coca-Cola

Component
1 2 3 4 5 6 7
refreshing .818 .160 -.277 -.236 .228 .414 -.281
cooling .817 .148 -.201 -.123 .134 .273 -.385
cool .739 .207 -.204 -.228 .093 .339 -.454
cold .717 .176 -.048 -.101 .068 .194 -.287
freshness .685 .126 -.304 -.331 .214 .483 -.262
great brand .530 .353 -.312 -.214 .127 .131 -.216
familiar .451 .257 -.166 .065 .347 -.032 -.199
global .068 .796 .179 -.088 .218 .107 -.074
multinational .260 .781 .005 -.082 .057 .054 -.129
worldwide .057 .737 .082 -.044 .369 .004 .096
recognisable .294 .618 .074 -.136 .236 -.108 .193
identifiable .427 .507 -.019 -.228 .240 -.112 .070
not unknown .184 .418 .051 .050 -.047 -.222 .247
unhealthy -.165 .060 .801 -.003 .017 -.102 .240
sugar -.109 .097 .775 -.120 .042 -.100 .200
obesity -.239 .014 .774 .080 -.041 -.023 .019
calories -.102 .047 .759 -.048 -.033 -.079 -.071
sweet .286 .270 .491 -.196 -.083 .111 .082
not natural -.235 .080 .448 .174 -.348 -.195 .391
not healthy -.077 .022 .441 .332 -.110 -.395 .431
Christmas song .086 .092 .091 -.881 .081 .137 -.065

XXXII
 Santa Claus .137 .038 -.024 -.869 .153 .245 -.052
Christmas .014 .025 .072 -.852 .192 .153 -.171
traditional .078 .115 -.011 -.225 .838 .128 -.005
classic .232 .240 -.109 -.135 .799 .023 -.037
original .286 .348 -.033 -.376 .511 .126 -.227
American -.102 .318 .363 -.069 .487 .044 .186
drink it when tired .135 .008 -.060 -.147 .085 .806 -.296
keeps me going .311 .109 -.278 -.186 .046 .754 -.328
energetic .377 .046 -.127 -.250 .219 .657 -.102
no sleeping .082 -.050 .054 -.176 -.010 .654 -.214
bonding with friends .243 -.024 -.251 -.218 -.017 .416 -.814
hot day .522 -.024 -.065 -.182 .124 .308 -.692
warm summer .467 -.048 -.072 -.175 .292 .314 -.681
fun .555 .085 -.241 -.310 .213 .360 -.653
parties .367 .301 -.081 -.124 -.106 .252 -.622
exciting .445 .037 -.314 -.433 .082 .446 -.591
Extraction Method: Principal Component Analysis.
Rotation Method: Oblimin with Kaiser Normalization.

XXXIII
 Scree Plot

14

12

10
Eigenvalue

1 2 3 4 5 6 7 8 9 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 4 4 4
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2

Component Number

Figure ap4-2. The Initial Scree Plot for Google

Table ap4-4. The Communalities for Google

Extraction Extraction
fun .585 trustful .571
fast .568 all knowing .585
not old .407 high speed .518
world .627 global .660
life saver .544 almighty .564
no privacy .657 very important .667
useful .720 no independency .515
entertainment .639 no problem .497
happiness .734 informative .672
huge .520 knowledge .631
connecting with the world .572 observes people's habits .593
easy to use .642 encyclopaedia .608
provides data .389 efficient .494
worldwide .735 young .677
basic necessity .562 dominant .604
freedom .587 innovative .708
helpful .729 creative .679
smart .636 satisfaction .603
convenient .639 multibillion industry .521
Extraction Method: Principal Component Analysis.

XXXIV
Table ap4-5. The Total Variance Explained for Google
Rotation Sums
Extraction Sums of Squared of Squared
Initial Eigenvalues Loadings Loadings a
Compo % of Cumulative % of Cumulative
nent Total Variance % Total Variance % Total
1.000 11.898 31.310 31.310 11.898 31.310 31.310 6.575
2.000 3.178 8.364 39.674 3.178 8.364 39.674 6.740
3.000 2.390 6.290 45.963 2.390 6.290 45.963 2.253
4.000 1.565 4.118 50.082 1.565 4.118 50.082 6.128
5.000 1.423 3.744 53.826 1.423 3.744 53.826 4.557
6.000 1.283 3.375 57.201 1.283 3.375 57.201 5.909
7.000 1.123 2.954 60.156 1.123 2.954 60.156 3.689
8.000 0.999 2.629 62.785
Extraction Method: Principal Component Analysis.
a
When components are correlated sums of squared loadings cannot be added to obtain a
total variance.

Table ap4-6. The Structure Matrix for Google

Component
1 2 3 4 5 6 7
very important .766 -.425 .059 .295 -.191 -.305 .323
almighty .685 -.106 .296 .329 -.296 -.321 .172
all knowing .648 -.221 .031 .343 -.327 -.381 .460
life saver .640 -.270 -.006 .328 -.530 -.194 .252
trustful .630 -.322 -.270 .367 -.337 -.144 .307
basic necessity .629 -.492 .133 .222 -.375 -.249 .176
satisfaction .618 -.410 -.023 .551 -.421 -.312 .345
no problem .594 -.280 -.029 .358 -.426 -.200 .396
dominant .509 -.145 .393 .434 -.203 -.347 -.232
helpful .353 -.836 -.092 .296 -.158 -.395 .312
useful .205 -.815 -.028 .033 -.156 -.343 .224
easy to use .149 -.773 -.046 .273 -.169 -.456 .177
convenient .285 -.760 -.102 .300 -.088 -.402 .410
fast .235 -.725 -.019 .323 -.070 -.424 .141
provides data .203 -.526 -.014 .321 -.044 -.426 .340
no privacy -.195 .104 .762 -.090 .014 -.103 -.143
observes people's habits .164 .021 .743 .201 -.020 -.165 .003
no independency .158 .024 .703 .080 -.118 -.052 -.021
innovative .360 -.206 .115 .828 -.300 -.229 .228
creative .319 -.252 .099 .812 -.300 -.213 .235
young .281 -.034 .102 .735 -.450 -.282 .127
smart .400 -.536 .057 .664 -.348 -.228 .310

XXXV
efficient .446 -.462 -.029 .545 -.230 -.272 .400
multibillion industry .284 -.315 .270 .536 -.169 -.468 -.110
high speed .483 -.476 -.015 .492 -.026 -.412 .283
happiness .471 -.189 .018 .369 -.809 -.292 .215
entertainment .238 -.205 .105 .278 -.778 -.214 .179
fun .278 -.230 -.014 .407 -.707 -.166 .225
freedom .565 -.285 .040 .482 -.605 -.264 .153
not old -.173 -.288 -.229 .243 .348 -.017 .265
worldwide .176 -.489 .120 .211 -.113 -.837 .139
global .264 -.321 .066 .295 -.048 -.792 .199
world .127 -.334 .119 .166 -.162 -.773 .243
connecting with the world .182 -.472 .137 .276 -.355 -.646 .314
huge .378 -.358 .117 .155 -.304 -.630 -.026
knowledge .405 -.386 -.030 .332 -.224 -.433 .690
informative .385 -.473 .005 .348 -.142 -.490 .686
encyclopaedia .393 -.233 .151 .272 -.348 -.234 .682
Extraction Method: Principal Component Analysis.
Rotation Method: Oblimin with Kaiser Normalization.

XXXVI
 Scree Plot

12,5

10,0
Eigenvalue

7,5

5,0

2,5

0,0

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

Component Number

Figure ap4-3. The Initial Scree Plot for Nike

Table. Ap4-7. The Communalities for Nike

Extraction Extraction
sports .397 pricey .812
achiever .532 child labour .636
freedom .576 just do it .382
athletes .515 fitness .714
cool .618 innovative .654
expensive .825 running .737
sweat shop .599 sporty .732
good quality .557 victory .713
not lazy .554 not failure .657
life .475 young .407
crowd cheering .532 durable .512
comfortable .446 air technology .422
jogging .558 not socially responsible .512
Michael Jordan .503 innovation .656
powerful .515 winning .699
self-assured .526 not weak .598
athletic .628 Tiger Woods .411
trendy .558 positive thinking .637
Extraction Method: Principal Component Analysis.

XXXVII
Table ap4-8. The Total Variance Explained for Nike
Rotation Sums
Extraction Sums of Squared of Squared
Compo Initial Eigenvalues Loadings Loadings a
nent % of Cumulative % of Cumulative
Total Variance % Total Variance % Total
1 12.130 33.695 33.695 12.130 33.695 33.695 10.723
2 2.853 7.926 41.621 2.853 7.926 41.621 2.699
3 2.617 7.270 48.891 2.617 7.270 48.891 2.721
4 1.797 4.991 53.882 1.797 4.991 53.882 3.332
5 1.410 3.915 57.797 1.410 3.915 57.797 7.786
6 1.306 3.627 61.424
7 1.056 2.934 64.358
8 .987 2.742 67.100
Extraction Method: Principal Component Analysis.
a
When components are correlated sums of squared loadings cannot be added to obtain a total
variance.

Table ap4-9. The Structure Matrix for Nike

Component
1 2 3 4 5
victory .825 .151 -.073 -.265 -.520
winning .823 .135 -.034 -.237 -.477
innovation .773 .048 -.319 -.264 -.406
positive thinking .772 .133 -.230 -.044 -.371
innovative .761 .017 -.340 -.252 -.438
cool .725 .072 -.320 -.248 -.510
self-assured .724 .040 -.120 -.193 -.363
freedom .714 .009 -.320 -.115 -.409
crowd cheering .709 -.060 .012 -.256 -.345
achiever .706 .123 -.120 -.232 -.482
powerful .699 .041 -.073 -.280 -.439
life .670 -.085 -.182 -.112 -.288
air technology .615 -.024 .036 -.027 -.361
trendy .610 -.017 -.326 -.457 -.425
durable .599 .081 -.339 -.383 -.439
young .596 .042 -.203 -.314 -.300
Tiger Woods .557 .119 .196 -.039 -.180
Michael Jordan .504 .086 .418 -.176 -.242
just do it .504 .226 .176 -.102 -.416
not failure .127 .787 -.199 .053 -.171
not weak .095 .761 -.071 -.087 -.265
not lazy -.015 .738 -.061 .073 -.161

XXXVIII
 child labour -.128 -.250 .755 -.045 -.008
sweat shop .003 -.287 .675 -.241 -.076
not socially -.419 .287 .530 .085 .103
ibl
expensive .120 -.032 .101 -.897 -.180
pricey .149 -.071 .093 -.890 -.169
running .365 .119 -.014 -.187 -.853
sporty .439 .253 -.073 -.222 -.851
fitness .507 .199 .009 -.176 -.829
athletic .493 .324 -.034 -.205 -.757
jogging .315 .012 .008 -.182 -.730
athletes .429 .290 -.002 -.270 -.684
sports .193 .349 .051 -.239 -.564
comfortable .512 .079 -.320 -.262 -.531
good quality .387 .332 -.421 -.425 -.483
Extraction Method: Principal Component Analysis.
Rotation Method: Oblimin with Kaiser Normalization.

XXXIX
Table ap4-10. The Levene's Tests of Equality of Error Variances for
the Coca-Cola’s Meanings
Dependent variable F df1 df2 Sig.
Refreshing 2.429 11 354 .006
Globally recognised 1.847 11 354 .045
Unhealthy 1.729 11 354 .066
Christmas .545 11 354 .872
Classic 3.993 11 354 .000
Energising 1.298 11 354 .224
Bonding with friends 1.066 11 354 .388

Table ap4-11. The Non-parametric Kruskal-Wallis Test for the

Dependent Variable Refreshing
Independent v. Country Type of Coca-Cola user
Chi-Square 23.339 56.977
df 3 2
Asymp. Sig. .000 .000

Table ap4-12. The Non-parametric Mann-Whitney Tests for the

Dependent Variable Refreshing and the Independent Variable Country
Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4
Mann-Whitney U 4737.0 2988.0 2966.0 3288.0 3312.0 2061.0
Wilcoxon W 10953. 9204.000 9182.0 8959.0 5392.0 4141.0
Z -2.479 -4.394 -1.815 -3.205 -.257 -2.527
Asymp. Sig. (2-tailed) .013 .000 .069 .001 .797 .011
1 = Austria; 2 = Finland; 3 = India; 4 = United States

Table ap4-13. The Non-parametric Mann-Whitney Tests for the

Dependent Variable Refreshing and the Independent Variable Type of
Coca-Cola User
Independent v. 1 vs. 2 1 vs. 3 2 vs. 3
Mann-Whitney U 5703.0 3463.0 3412.0
Wilcoxon W 10956.0 6238.0 6187.0
Z -5.796 -6.401 -1.085
Asymp. Sig. (2-tailed) .000 .000 .278
1 = Engaged users; 2 = Disengaged users; 3 = Non-users

Table ap4-14. The Non-parametric Kruskal-Wallis Tests for the

Dependent Variable Globally Recognised
Independent v. Country Type of Coca-Cola user
Chi-Square 10.730 2.240
df 3 2
Asymp. Sig. .013 .326

XL
Table ap4-15. The Non-parametric Mann-Whitney Tests for the
Dependent Variable Globally recognised and the Independent
Variable Country
Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4
Mann-Whitney U 5798.0 3920.0 2694.0 3790.0 2591.0 2549.0
Wilcoxon W 11469.0 7575.0 4774.0 7445.0 4671.0 4629.0
Z -.184 -2.026 -2.658 -1.883 -2.576 -.656
Asymp. Sig. (2-tailed) .854 .043 .008 .060 .010 .512
1 = Austria; 2 = Finland; 3 = India; 4 = United States

Table ap4-16. The Non-parametric Kruskal-Wallis Tests for the

Dependent Variable Classic
Independent v. Country Type of Coca-Cola
Chi-Square 100.296 11.887
df 3 2
Asymp. Sig. .000 .003

Table ap4-17. The Non-parametric Mann-Whitney Tests for the

Dependent Variable Classic and the Independent Variable Country
Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4
Mann-Whitney U 2575.0 3983.0 1642.0 1510.0 3213.0 959.0
Wilcoxon W 8791.0 7638.0 7858.0 5165.0 5293.0 4614.0
Z -7.155 -1.866 -5.917 -7.888 -.576 -6.753
Asymp. Sig. (2-tailed) .000 .062 .000 .000 .565 .000
1 = Austria; 2 = Finland; 3 = India; 4 = United States

Table ap4-18. The Non-parametric Mann-Whitney Tests for the

Dependent Variable Classic and the Independent Variable Type of
Coca-Cola User
Independent v. 1 vs. 2 1 vs. 3 2 vs. 3
Mann-Whitney U 7304.0 6582.0 3126.0
Wilcoxon W 12557.0 9357.0 8379.0
Z -3.469 -.804 -1.942
Asymp. Sig. (2-tailed) .001 .421 .052
1 = Engaged users; 2 = Disengaged users; 3 = Non-users

XLI
Table ap4-19. The Levene's Tests of Equality of Error Variances for
Google’s meanings
Dependent variable F df1 df2 Sig.
Almighty 1.776 7 309 .092
Helpful 3.503 7 309 .001
Intrusive 2.330 7 309 .025
Inspired 3.347 7 309 .002
Fun 1.193 7 309 .306
Global 4.238 7 309 .000
Informative 3.331 7 309 .002

Table ap4-20. The Non-parametric Kruskal-Wallis Test for the

Dependent Variable Helpful and the Independent Variable Country
Independent v. Country
Chi-Square 4.597
df 3
Asymp. Sig. .204

Table ap4-21. The Non-parametric Mann-Whitney Test for the

Dependent Variable Helpful and the Independent Variable Type of
Google User
Independent v. Type of Google user
Mann-Whitney U 8254.000
Wilcoxon W 17299.000
Z -4.971
Asymp. Sig. (2-tailed) .000

Table ap4-22. The Non-parametric Kruskal-Wallis Test for the

Dependent Variable Intrusive and the Independent Variable Country
Independent v. Country
Chi-Square 16.204
df 3
Asymp. Sig. .001

Table ap4-23. The Non-parametric Mann-Whitney Tests for the

Dependent Variable Intrusive and the Independent Variable Country
Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4
Mann-Whitney U 3635.0 3470.0 1758.0 2716.0 2079.0 1291.0
Wilcoxon W 7730.0 6171.0 3189.0 6811.0 3510.0 2722.0
Z -2.386 -.660 -3.493 -1.899 -1.279 -3.180
Asymp. Sig. (2-tailed) .017 .509 .000 .058 .201 .001
1 = Austria; 2 = Finland; 3 = India; 4 = United States

XLII
Table ap4-24. The Non-parametric Mann-Whitney Test for the
Dependent Variable Intrusive and the Independent Variable Type of
Google User
Independent v. Type of Google user
Mann-Whitney U 11787.000
Wilcoxon W 20832.000
Z -.588
Asymp. Sig. (2-tailed) .557

Table ap4-25. The Non-parametric Kruskal-Wallis Test for the

Dependent Variable Inspired and the Independent Variable Country
Independent v. Country
Chi-Square 56.020
df 3
Asymp. Sig. .000

Table ap4-26. The Non-parametric Mann-Whitney Tests for the

Dependent Variable Inspired and the Independent Variable Country
Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4
Mann-Whitney U 4536.0 1996.0 1392.0 1557.0 1133.0 1667.0
Wilcoxon W 9687.0 7147.0 6543.0 5652.0 5228.0 4368.0
Z -.024 -5.155 -4.885 -5.767 -5.233 -1.322
Asymp. Sig. (2-tailed) .981 .000 .000 .000 .000 .186
1 = Austria; 2 = Finland; 3 = India; 4 = United States

Table ap4-27. The Non-parametric Mann-Whitney Test for the

Dependent Variable Inspired and the Independent Variable Type of
Google User
Independent v. Type of Google user
Mann-Whitney U 6547.000
Wilcoxon W 23383.000
Z -7.088
Asymp. Sig. (2-tailed) .000

Table ap4-28. The Non-parametric Kruskal-Wallis Test for the

Dependent Variable Global and the Independent Variable Country
Independent v. Country
Chi-Square 20.723
df 3
Asymp. Sig. .000

XLIII
Table ap4-29. The Non-parametric Mann-Whitney Tests for the
Dependent Variable Global and the Independent Variable Country
Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4
Mann-Whitney U 3030. 3635.0 2141.0 2175.0 2155.0 1516.0
Wilcoxon W 8181. 8786.0 7292.0 4876.0 3586.0 4217.0
Z -3.973 -.157 -2.037 -3.704 -.961 -2.068
Asymp. Sig. (2-tailed) .000 .875 .042 .000 .336 .039
1 = Austria; 2 = Finland; 3 = India; 4 = United States

Table ap4-30. The Non-parametric Mann-Whitney Test for the

Dependent Variable Global and the Independent Variable Type of
Google User
Independent v. Type of Google user
Mann-Whitney U 8772.000
Wilcoxon W 17817.000
Z -4.328
Asymp. Sig. (2-tailed) .000

Table ap4-31. The Non-parametric Kruskal-Wallis Test for the

Dependent Variable Informative and the Independent Variable
Country
Independent v. Country
Chi-Square 39.565
df 3
Asymp. Sig. .000

Table ap4-32. The Non-parametric Mann-Whitney Test for the

Dependent Variable Informative and the Independent Variable Country
Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4
Mann-Whitney U 2996.0 3158.0 2377.0 1524.0 1270.0 1924.0
Wilcoxon W 7091.0 8309.0 7528.0 5619.0 5365.0 3355.0
Z -4.062 -1.612 -1.139 -5.877 -4.660 -.052
Asymp. Sig. (2-tailed) .000 .107 .255 .000 .000 .959
1 = Austria; 2 = Finland; 3 = India; 4 = United States

Table ap4-33. The Non-parametric Mann-Whitney Test for the

Dependent Variable Informative and the Independent Variable Type
of Google User
Independent v. Type of Google user
Mann-Whitney U 8319.000
Wilcoxon W 25155.000
Z -4.890
Asymp. Sig. (2-tailed) .000

XLIV
Table ap4-34. The Levene's Test of Equality of Error Variances for
Nike’s meanings
Dependent variable F df1 df2 Sig.
Achiever 1.477 11 294 .139
Weak 1.471 11 294 .142
Not socially 1.750 11 294 .062
Pricey 1.601 11 294 .098
Sports 4.056 11 294 .000

Table ap4-35. The Non-parametric Kruskal-Wallis Test for the

Dependent Variable Sports
Independent v. Country Type of Nike user
Chi-Square 4.317 32.152
df 3 2
Asymp. Sig. .229 .000

Table ap4-36. The Non-parametric Mann-Whitney Tests for the

Dependent Variable Sports and the Independent Variable Type of
Nike user
Independent v. 1 vs. 2 1 vs. 3 2 vs. 3
Mann-Whitney U 3016.0 4253.0 3051.0
Wilcoxon W 12196.0 13433.0 5262.0
Z -3.716 -5.312 -1.314
Asymp. Sig. (2-tailed) .000 .000 .189
1 = Engaged users; 2 = Disengaged users; 3 = Non-users

XLV
Curriculum Vitae
Last Name: First
First Name: Ivana
Born: March 17, 1978 in Rijeka, Croatia

Education
04/05-09/09 University of St. Gallen, Doctoral Programme in
Multicultural Management, St. Gallen, Switzerland
10/02-09/04 University of St. Gallen, M.Sc. Programme in International
Management, St. Gallen, Switzerland
08/03-05/05 University of St. Gallen, CEMS M.A. in International
Management, St. Gallen, Switzerland
08/03-12/03 Stockholm School of Economics, Exchange Semester within
M.Sc. and CEMS M.A. Programmes, Stockholm, Sweden
10/96-06/01 Faculty of Economics in Rijeka, B.Sc. in Economics, Rijeka,
Croatia

Working Experience
12/01-... University of Rijeka, Faculty of Economics, Rijeka, Croatia
Research and Teaching Assistant

06/04-09/04 Siemens VDO Automotive AG, Frankfurt Germany

Intern in HVAC business unit

08/01-11/01 Beling d.o.o., Rijeka, Croatia

Kaba-Elzett and other security products manager

01/00-05/01 Silca-Unican-Elzett Rt., Budapest, Hungary

AIESEC intern in marketing department

Contact Information
E-mail: ifirst@efri.hr

XLVI