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
and
The President:
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
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?
1
1.2. Problem Definition and Purpose of the Study
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).
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).
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.
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.
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).
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:
7
1.3. Structure of the Thesis
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.
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.
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
Meaning
Cultural
Brand
Consumer Brand
Needs Benefits
Type of
User
4. RESEARCH METHODOLOGY
7. CONCLUSION
10
2. Definitions and Theoretical Background
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
13
perceived image, the inner self of the brand, i.e. its identity, is what should
really be managed.
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
RELATIONSHIP
14
Company actions
Strategy Programmes
(Satisfaction)
What customers do
about a brand
Financial market
impact
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.
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.
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.
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).
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.
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).
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).
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).
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.
22
2.1.1.8. Brand Attitudes
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.
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.
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.
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.
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.
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.
28
Table 2-1. An Overview of the Existing Brand Meaning Definitions
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).
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.
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).
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
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.
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).
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).
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).
35
be careful to maintain an equal level of understanding of their brands’
meanings as consumers (Berthon, Pitt, and Campbell 2008).
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.
37
2.2. The Concept of Culture
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).
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).
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).
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
Reinforcement
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.
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,
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).
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
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:
45
activities at the same time. Trompenaars refers to this difference
as “times as a race” vs. “time as a dance”,
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).
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).
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
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.
51
3.1.1. The Influence of Cultural on Brand Communication
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.
53
families (De Matos and Rossi 2008). In other words, the personality
perception is influenced by the personality preference of the consumers.
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
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.
56
3.1.3. Cultural Dimensions as an Antecedent to Brand Meaning
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.
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:
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.
59
shapes) that stem from a customer’s culturally emphasised needs
(represented by the black triangle and the rectangle).
Legend
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.
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.
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.
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.
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.
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).
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:
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:
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:
69
by a sportsman is very often an appropriate source of identity construal for
consumers from cultures high in assertiveness and performance
orientation.
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.
71
And finally, along similar lines, for the last two cultural dimensions of
future orientation and gender egalitarianism, the following hypotheses can
be 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
73
3.3. The Effect of Brand Usage and Brand Affection
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:
75
Culture
Cultural H1
Dimensions
H2 Brand Meaning
H4
H3
Type of User
76
4. Research Methodology
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.
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.
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”.
Experts’
questionnaire analysis
Quantitative
stage
Cross
cultural
Questionnaire on a Questionnaire on a differences
pilot sample cross-cultural sample in brand
meaning
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.
81
4.2.2. Units of Analysis
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.
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.
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).
85
branding literature despite its position as the sixth most valuable global
brand in the Interbrand and Business week list (2008).
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
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).
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.
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.
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.
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.
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.
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.
94
in response quality between the Skype communication and in-person
communication was discovered.
5.1. 2. Results
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
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.
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.
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).
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).
5.2.3. Results
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.
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
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.
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.
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.
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.
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
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.
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:
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
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.
…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.
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
…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.
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.
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.
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.
113
Table 5-6. The Types of Responses for Coca-Cola
Brand
Attribute/
Meaning Stimulus Type of benefit meaning
Personality
category
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.
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.
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.
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.
118
Table 5-8. The Types of Responses for Nike
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
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.
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
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
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:
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”.
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 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.
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:
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.
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
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.
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.
133
Table 6-2. Google’s Groups of Meanings and Their Hypothesised
Relationships to the Cultural Dimensions
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.
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.
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).
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).
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.
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).
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).
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.
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
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.
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.
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.
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.
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)).
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.
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).
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
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.
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
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.
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.
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.
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 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
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).
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**).
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
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
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.
168
higher scores on power distance than the Austrian (UAA = 2.44) and
Finish (UAF = 2.19) lower scores on this dimension.
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.
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.
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.
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 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.
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.
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.
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.
0,60
Austria
0,40 Finland
India
0,20 United
States
0,00
-0,20
-0,40
-0,60
-0,80
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
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.
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.
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.
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).
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
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.
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.
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.
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.
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.
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.
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
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.
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
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
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
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
187
6.2.4.4. Hypothesis 4
188
7. Conclusion
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.
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
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.
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.
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.
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.
193
7.3.1.1. Contribution to Brand Management Field
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.
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.
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.
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.
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.
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
200
AdAge (2008) “Top 10 Slogans of The Century,” (accessed 20th
December 2008) [available at:
http://adage.com/century/slogans.html].
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.
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.
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.
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)
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.
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.
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.
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.
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.
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.
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.
207
Fine, Seymour H (1981), The Marketing of Ideas and Social Issues. New
York: Praeger.
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.
Griffin, Jill (1997), Customer Loyalty – How to Earn it, How to Keep it.
San Francisco: Jossey-Bass Publishers.
209
Across Cultures: Issues and Perspectives, Pat Joynt and Malcom
Warner eds. London: International Thomson Business Press, 275–
305.
-------, Geraldine R., Dawn Iacobucci, and Bobby J. Calder (2002), “Using
Network Analysis to Understand Brands,” Advances in Consumer
Research, 29 (1), 397-405.
210
Henry, Jim (2007), “Company's Quality Crusade Was Launched by an
American,” Automotive News, 82, (2007, October 30) 58.
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.
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].
-------, 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.
213
-------, Kevin L. (2003a), “Brand Synthesis: The Multidimensionality of
Brand Knowledge,” Journal of Consumer Research, 29 (4), 595-600.
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.
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.
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.
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.
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.
-------, Jr. Albert M. and Hope Jensen Schau (2005), “Religiosity in the
Abandoned Apple Newton Brand Community,” Journal of Consumer
Research, 31 (4), 737-47.
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.
Oakenfull, Gillian, Edward Blair, Betsy Gelb, and Peter Dacin (2000),
“Measuring Brand Meaning,” Journal of Advertising Research, 40
(5), 43-53.
-------, 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.
219
Park, C. Whan, Bernard J. Jaworski, and Deborah J. Maclnnis (1986),
“Strategic Brand Concept-Image Management,” Journal of
Marketing, 50 (4), 135-45.
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.
220
Research: A Critical Review of the Literature and Recommended
Remedies,” Journal of Applied Psychology, 88 (5), 879-903.
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.
221
Standardization - Adaptation Decisions in Korea,” Journal of
Marketing Theory & Practice, 11 (4), 12-25.
Ross, Jill and Rod Harradine (2004), “I'm Not Wearing That!: Branding
and Young Children,” Journal of Fashion Marketing & Management,
8 (1), 11-26.
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.
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.
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.
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.
225
Taylor, Edward (1871), Primitive cultures. London: John Murray.
-------, 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.
226
Vargo, Stephen L and Robert F. Lusch (2004), “Evolving to a New
Dominant Logic for Marketing,” Journal of Marketing, 68 (1), 1-17.
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.
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.
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
XXV
Appendix 3. The Questionnaire
Cross-cultural Branding
I. Default Section
7. If the answer is yes, please indicate when, where, and for how long.
________________________________________________________
II. Coca-Cola
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
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.
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
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.
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
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
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.
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
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.
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
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.
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
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
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
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
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
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
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
Contact Information
E-mail: ifirst@efri.hr
XLVI