Brand Meaning and its Creation in a Cross-Cultural Context

D I S S E R T A T I O N of the University of St. Gallen,

Graduate School of Business Administration, Economics, Law and Social Sciences (HSG)

to obtain the title of Doctor Oeconomiae

submitted by

Ivana First

from

Croatia

Approved on the application of

Prof. Dr. Torsten Tomczak

and

Prof. Dr. Marcus Schögel

Dissertation no. 3616

Gutenberg AG, Schaan 2009

The University of St. Gallen, Graduate School of Business Administration, Economics, Law and social Sciences (HSG) hereby consents to the printing of the present dissertation, without hereby expressing any opinion on the views herein expressed. St. Gallen, May 19, 2009.

The President:

Prof. Ernst Mohr, PhD

Preface Looking back at the process of writing my doctoral thesis, I feel it was a challenging but rewarding task that required devotion, curiosity, and diligence. It was a process that would have been impossible without the professional and personal support of several people. I would like to take this opportunity to express my gratitude to them. Firstly, I would like to thank my supervisor, Prof. Dr. Torsten Tomczak, and co-supervisor, Prof. Dr. Marcus Schögel, for giving me freedom as well as guidance in this research. I would also like to thank Prof. Dr. Bruno Grbac from the University of Rijeka (Croatia). Being a part of his team provided me with continuous constructive discussion and helped me to develop my research ideas. Furthermore, I am very grateful to my colleague and friend Deepali Sinha, a doctoral student herself, who was always available for ongoing concerns ranging from prosaic to complex issues. I would also like to thank Dr. Daniel Wentzel for his insight into various issues throughout the process of writing this thesis, Dr. Darko Lon ari for his advice on statistical analyses, and Joanna Niederer for correcting the English side of things. In addition, I would like to express my heartfelt appreciation to all the participants in this research and especially to the interviewees. Without their time and effort this research would not have been accomplished. Finally, I am deeply grateful to my family, who gave me the opportunity to work on this thesis and, together with my boyfriend Andrej, provided me with backing and encouragement whenever I needed it. I dedicate this work to them. Rijeka, July 31, 2009.

Ivana First

I

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

II

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

III

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

IV

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

V

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

VI

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

VII

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.

VIII

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.

1

1. Introduction

1.1. Setting the Scene

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

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

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

2

1.2. Problem Definition and Purpose of the Study

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

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

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

The meaning of a brand is not predetermined in some brand identity booklet. Brand meaning is constructed in consumers’ minds. People experience brands in different settings, situations and psychological states,

3

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

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

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

4

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

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

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

5

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

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

Hence, Bulmer and Buchanan-Oliver (2006) advise advertisers to craft scenes and sequences to provide the viewer with a brand meaning, keeping in mind that their interpretation is dependent on knowledge of culturally constituted signs or symbols. Kates and Goh (2003) recognise this phenomenon as “brand morphing” and define it as a way that brand meaning, facilitated by practitioners’ efforts to accommodate, reinforce and create diverse cultural meaning, changes among different groupings of

6

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

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

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

Understand what brands mean to consumers and consequently conceptualise brand meaning,

Reveal how different cultural backgrounds affect global brands’ meanings among otherwise homogeneous consumers in terms of demographic, professional and occupational characteristics.

7

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:

What do brands mean to consumers?

How do brands acquire meaning?

How are other brand concepts such as brand benefits, associations, beliefs, and attributes related to brand meanings?

What is culture and how do cultures differ?

How do differences in cultural dimensions affect brand meanings?

How are brand consumption frequency and brand affection related to brand meanings across and within countries?

What are the scientific and managerial implications of the findings?

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

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

8

1.3. Structure of the Thesis

This thesis is structured into seven chapters as presented in Figure 1-1. The first introductory chapter sets the stage for the research. It briefly provides an overview of the current state of scientific thought on the concept of brands and, in particular, brand meaning. It also explicitly states the purpose of the thesis articulated through the research questions, and finally presents the structure of the paper.

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

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

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

9

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

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

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

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

10

Figure 1-1. Structure of the Thesis

5. QUALITATIVE EMPIRICAL RESEARCH

6. QUANTITATIVE EMPIRICAL RESEARCH

1. INTRODUCTION

3. PREVIOUS RESEARCH AND HYPOTHETICAL MODEL

2. DEFINITIONS AND THEORETICAL BACKGROUND C

ultu

ral

Dim

ensi

ons

4. RESEARCH METHODOLOGY

7. CONCLUSION

Bra

nd

Mea

ning

DEFINING BRAND MEANING Research Procedure

Results Discussion

EXPLORING DIFFERENCES IN BRAND MEANINGS ACROSS CULTURES Research Procedure

Results Discussion

Summary of the Findings

Managerial Implications

Scientific Implications

Type of User

Consumer Needs

Brand Benefits

11

2. Definitions and Theoretical Background

This chapter provides an introduction to two key concepts of this study, namely brand and culture. Furthermore, it gives a short overview of how both concepts evolved and indicates the current stage both of them have reached in the scientific research.

2.1. The Concept of Brand

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

Achenreiner and John (2003) studied the evolution of a child’s understanding of brands and realised that younger children are more attuned to brands on a perceptual level (serving as a cue for products with familiar names and perceptual features), and that only later at some point of their personal development do conceptual meanings become salient (serving to evaluate a brand as such and not as the name of a product, or inferring the personality characteristics of the consumers owning it). The learning process also takes place in adults, however here it is not an

12

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)

13

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.

Figure 2-1. The Brand Identity Prism

Source: Kapferer (1998, p. 107)

For example, Kapferer (2004) developed a hexagonal identity prism (Figure 2-1) as a model that defines what a brand really is and what it stands for. Identity is what a brand wants to be perceived as, i.e. what it “transmits”. On the other end of the communication channel, i.e. what customers receive, is the brand’s image. Due to the market noise, the brand image is not identical to the brand identity. Obviously, the perceived image is what matters in the end; nevertheless, to manage a

Reflection Self-image

Relationship Culture

Physique Personality

PICTURE OF SENDER

PICTURE OF RECIPIENT

EXT

ERN

AL

ISA

TIO

N

INE

RN

AL

ISA

TIO

N

14

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

Similarly, Aaker (1996), proposed a brand identity system (Figure 2-2) that encompasses the whole complexity surrounding how a brand evolves. His model emphasises the need to consider product-related associations, organization-related associations, personality, and symbolic associations in order to ensure that brand identity has texture and depth. He also differentiates between core and extended brand identity, and later with Joachimsthaler (Aaker and Joachimsthaler 2002) adds an even more focussed dimension named “brand essence”, which might be useful for brands that possess several compactly summarised core identities.

Figure 2-2. The Brand Identity System

Source: Aaker (1996, p. 79)

Extended

BRAND IDENTITY

Core

Brand as product:- product scope - product attributes - quality/value -uses -users - country of origin

Brand as organisation: - organisation atributes - local vs. global

Brand as person: - personality - customer / brand relationship

Brand as symbol: - visual image and metaphors - brand heritage

VALUE PROPOSITION -functional -emotional -self-esteem benefits benefits benefits

CREDIBILITY Support other brands

RELATIONSHIP

15

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

Source: Keller and Lehmann (2006)

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

Partners' actions: chanels,

employees

Strategy

Quality Direction

Programmes

Specifics: type, budget

Quality

Company actions

What customers think and feel about a brand

What customers do about a brand

Financial market impact

Competitor's actions

Industry / environmental

conditions

Attitude Attachment Activity

(Satisfaction)

Awareness Associations

16

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

2.1.1. Branding Concepts

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

2.1.1.1. Brand Awareness

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

17

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

Keller (2003a) further differentiates between the depth and breadth of brand awareness. Depth is understood as the ease of brand recognition and recall, and breadth refers to the number of purchase and consumption situations for which the brand comes to mind. Kapferer (2004, p. 17) differentiates between aided and unaided awareness, and claims that both contribute to brand equity. According to Aaker and Joachimsthaler (2002, p. 17), although awareness is often taken for granted in branding and is considered an insufficient condition for measuring a brand’s strength, in their opinion it strongly contributes to brand equity.

In summary, awareness can be defined as a multilevel understanding of the brand, which is a prerequisite for brand existence, and at the same time contributes to brand equity.

2.1.1.2. Brand Associations

Krishnan (1996) defines associations as a link between nodes in the consumer's mind. In other words, memory is composed of knowledge that is organised as a network of connections. The building blocks of this network are mental nodes that represent any piece of information like a brand, an attribute, a situation or similar, and connections between the nodes that represent associations.

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

18

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

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

Overall, brand associations are understood as links in the consumer’s memory between brand on one side and attributes, benefits, and attitudes on the other.

2.1.1.3. Brand Attributes

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

Brand managers who focus on attributes to develop a brand’s identity are paving a path to weak customer loyalty (Phau and Lau 2000). Such a

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

19

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

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

2.1.1.4. Brand Personality

Aaker (1997) defines brand personality as the set of human characteristics associated with a brand. This set of human characteristics is often the most differentiating dimension between brands (Aperia and Back 2004). According to Riesenbeck and Perrey (2007, p. 151), brand personality accounts for 70% of a brand’s strength, and according to LePla and Parker (2002, p. 86) it is the foundation for each customer’s emotional relationship with a brand and its company as a whole.

In Kapferer’s (2004, p. 107) identity prism, among the six edges, personality is the innermost edge of a brand. It is internal and it is seen from the brand’s perspective. Personality explains the character of a brand by answering the following question: ‘Were it a person, what kind of person would it be’? Personality, as an expression of a brand’s identity, has been experimented with in brand advertising as far back as the 1970s (Czerniewski and Maloney 1999, p. 99). It was then considered that the best way to communicate a brand’s traits was to have a celebrity endorse it. By this, the celebrity transfers his or her own personality to the brand,

20

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

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

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

In summary, brand personality is an expression of brand identity, an important contributor to brand strength, and a basis for the consumer brand relationship.

2.1.1.5. Brand Beliefs

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

21

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

Overall, the most important characteristic of brand beliefs is its personal and cognitive aspects. Researchers also claim brand beliefs are the reason for brand strength.

2.1.1.6. Brand Knowledge

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

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

In summary, brand knowledge is a concept that is often related or even equated to brand beliefs.

22

4.1.1.7. Brand Emotions

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

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

Based on a study by Homer and Yoon (1992), Orth, Koenig, and Firbasova (2007) propose that emotions influence cognition, which further influences attitude and finally the purchase intention. A similar model was also proposed by Kim, Morris, and Swait (2008), in which they propose and empirically confirm that brand credibility influences brand affection and brand conviction, which then influences brand attitude strength.

More specifically, emotions act together with cognition to strengthen the representations of brands in consumers’ minds.

23

2.1.1.8. Brand Attitudes

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

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

In summary, brand attitudes are summary judgments, and although they are clearly related to both emotions and cognition, the research has not reached common ground in understanding their relationship.

24

2.1.1.9. Brand Benefits

Brand benefits are personal values and meanings that consumers attach to a brand's product attributes. Frequently, a major distinction is made between three basic categories of benefits according to the underlying

motivations to which they relate – functional, experiential, and symbolic

benefits (Keller 2003a; Park, Jaworski, and Maclnnis 1986). Functional benefits are the more intrinsic advantages of product consumption, and they usually correspond to product attributes. These benefits are linked to basic motivations, such as well-being and health. Experiential benefits relate to what it feels like to use the product and they also usually correspond to product attributes. These benefits satisfy experiential needs such as sensory pleasure and cognitive stimulation. Symbolic benefits are the more extrinsic advantages of product consumption. They usually correspond to non-product-related attributes and relate to underlying needs for social approval, personal expression, and outer-directed self-esteem (Orth and De Marchi 2007).

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

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

25

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

2.1.2. Brand Meaning

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

In a study by Henderson et al. (2003), brand meaning is considered equal to a brand association; that is, the participants in their study are asked to list the first meaning or association that comes to mind when looking at the brand’s logo. Moore and Homer (2008) also use the terms ‘brand associations’ and ‘brand meanings’ as synonyms, and Feldwick (2002)

26

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

Another widely cited paper by McCracken (1986) focuses on how meaning is manufactured and transferred from celebrities to brands and further to consumers. In that process of cascaded meaning transitions, consumers opt to acquire an image of their beloved celebrity by using the same brands that the celebrities use. McCracken, however, also fails to define what he understands by brand meaning. Since he speaks of meaning being transferred from a celebrity to a brand, it can be assumed that brand personality is what is referring to.

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

Keller (2003b) suggest that providing a brand with meaning actually implies explaining to consumers what a particular product can do for them

27

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

28

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

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

As shown in subchapter 2.1.1, different concepts within the world of brands are used interchangeably and for multiple manifests. Subchapter 2.1.2 has provided evidence that the concept brand meaning is not devoid of complexity and ambiguity either. The above-mentioned definitions of brand meaning are analysed in terms of their constituting elements and an overview is presented in table 2-1. From the summarised definitions, it can be seen that the terms most often equated or related to brand meaning are brand associations and brand attributes.

29

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

Author(s) Brand meaning is Brand meaning originates from

Brown, Kozinets, and Sherry 2003

Aura, Allegory, Arcadia, Antinomy

Chang and Chieng 2006 Knowledge Related to: Awareness, Association, Perceived quality, Attitude, Personality, and Image

Davis 2007 Attributes

Escalas and Betmann 2005 Personality Associations

Feldwick 2002 Associations, Beliefs

Franzen and Bouwman 2001 Impression, Associations

Mental links between brand names, images and cognitions

Hendersson et al. 2003 Associations

Hollenback, Peters, and Zinkhan 2008

Identity

Keller 2003b Benefits

Martin, Stewart, and Matta 2005

Knowledge and attitudes

Belongs to a network of associations that also include brand name, concrete and abstract attributes, and occasions

McCracken 1986 Personality , Image

Moore and Homer 2008 Associations, Image, Personality

Oakenfull et al. 2000 Attributes

Park, Milberg, and Lawson 1991

Attributes

Veloutsou and Moutinho 2008 Image Before moving on to exploring how brands acquire meanings, a small detour into the field of psychology and neuro-linguistic programming (NLP) offers the following definition of meaning in general. Meaning is related to the intention or significance of a message or experience. It is the natural consequence of interpreting experience using inner representations or experiences that are associated with external cues and events.

30

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

This definition serves as a good introduction to take a step forward in the quest to understand what brands mean to consumers, and to trace the path along which a brand acquires meaning. At this stage, it is particularly important to analyse how brands come to acquire meaning, and who influences brand meaning the most.

2.1.3. How Brands Acquire Meanings

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

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

2.1.3.1. Brand Strategies’ Role in Brand Meaning Creation

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

31

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

Figure 2-4. Traditional and Contemporary Brand Image Creation

Source: Blythe 2007.

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

Brand manager instigates dialogue

New product is created

Consumers enter dialogue about brand image

Central pool of

meanings Creatives and agency enter

dialogue about communications

strategy and tactics

Brand consultant s enter dialogue about

brand identity

“Pool of Reality” model of Brand Image Creation

New product is created

Marketers decide on

desired brand identity

Brand identitiy is communicated to consumers via vision, culture,

positioning, personality,

relationshp and presentation

Brand image is absorbed and accepted by consumers

The Traditional View of Brand Meaning Creation

32

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

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

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

33

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

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

2.1.3.2. Consumers’ Role in Brand Meaning Creation

When consumers are exposed to a company’s stories or brand information, they do not passively absorb those messages. Instead, they create their own meanings by mixing received information with their own

34

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

What matters in the construction of the brand relationship is not simply the idea that managers intend for them, but what consumers do with brands to add meaning to their lives (Fournier 1998; Cova and Pace 2006). As Vargo and Lusch (2004) conclude in their paper: The consumer is always a co-producer and he determines the perceived value of a brand on the basis of “value in use”. The firm can only make value propositions, but the final meaning is the result of an interpretation process (Richins 1994). Thus, although marketers initially design brands, final brand meanings reside in consumers’ minds (Keller 2003; De Chernatony 2001, p. 19).

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

35

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

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

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

36

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

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

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

Gregory (2007) proposed a process of how stakeholders can contribute to brand creation based on the concept of a ‘negotiated brand’. According to him, the negotiated approach starts by a corporation identifying its core values, which are a blend of management vision and internal stakeholder views. Then the brand goes out to the world to be modelled by external stakeholders and takes its final form. Gregory depicts this process as a spiral development.

37

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

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

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

38

2.2. The Concept of Culture

Culture (and cultural dimensions) is the second field of research relevant for the present study. This chapter provides an overview of what culture is and how it differs among various groups of people.

2.2.1. Culture, Cultural Differences and Boundaries of Culture

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

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

39

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

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

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

Furthermore, culture is a multilayered phenomenon that consists of artefacts, values and underlying assumptions. Artefacts are the most easily observed manifestation of culture, whereas underlying assumptions are the

40

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

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

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

41

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.

Figure 2-5. The Stabilising of Culture Patterns

Source: Hofstede 2001.

2.2.2. Cultural Dimensions

Despite the debate on the appropriate definition and understanding of culture, most social scientists (e.g. Kluckhohn and Strodtbeck 1961; Schwartz 1999; Hofstede 2001; Trompenaars 1994) have long recognised

Origins Ecological factors: Geography History Demography Hygiene Nutrition Economy Technology Urbanisation

Outside influence Forces of nature Forces of man: Trade Domination

Scientific discovery

Societal norms Value systems of major groups of a population

Consequences Structure and functioning of institutions: Family patterns Role differentiation Social stratification Socialisation emphases Education systems Religion Political systems Legislation Architecture Theory development

Reinforcement

42

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

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

As Hampden-Turner and Trompenaars (2002) note, virtually all of the members of the human race engaged in enterprise face the same problems or dilemmas, but their responses to these vary widely. Table 2-2 presents the key societal dilemmas and cultural dimensions that various researchers have identified. According to them, these serve as a key to understand how various cultures react when faced with such societal problems.

Hofstede (2001) greatly contributed to the field of international management by being the first to operationalise variables into measurable components – cultural dimensions. Before his study, which was conducted during the 60s and 70s, culture was considered vague, intangible and soft (Romani 2004). He conducted two survey rounds between 1967 and 1973 on 116,000 employees of IBM (at that time, but also today one of the largest multinational corporations) in 72 countries. He then identified four dimensions along which cultures differ. These are:

43

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

b) Power distance: the extent to which less powerful members of organisations and institutions accept and expect power to be distributed equally,

c) Uncertainty avoidance: the extent to which a culture programmes its members to feel either uncomfortable or comfortable in unstructured situations (situations that are novel, unknown, surprising, or different from usual),

d) Masculinity vs. Femininity: the distribution of emotional roles between the genders in which tough societies are considered masculine, and tender societies are considered feminine.

In the 1980s, a continuation of cross-cultural study uncovered the existence of a fifth dimension, at the time named Confucian dynamism (Hofstede and Bond 1988) and later changed to Long-term vs. short-term orientation. This dimension is defined as the extent to which a culture programmes its members to accept delayed gratification of their material, social, and emotional needs. This dimension is particularly important for capturing some of the behaviours and values in South-East Asian culture.

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

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

44

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

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

Author(s) Nature of the individual

Relationship with other beings

Relationship to nature

Time perception

Hofstede 2001

Individualism vs. collectivism

Masculinity vs. Femininity

Power distance

Uncertainty avoidance

Long-term vs. short-term orientation

Trompenaars 1994

Universalism vs. particularism

Individualism vs. communitarism

Analysed specifics vs. integrated wholes

Achievement vs. ascription

Equality vs. hierarchy

Inner vs. outer direction

Sequential time vs. synchronised time

Schwartz 1999

Conservatism vs. autonomy

Hierarchy vs. egalitarianism

Mastery vs. harmony

House et al. 2004

Assertiveness

Performance orientation

In group collectivism

Institutional collectivism

Power distance

Gender egalitarianism

Uncertainty avoidance

Humane orientation

Future orientation

Source: adapted from Brannen et al. 2004.

Like Hofstede, Trompenaars (1994), conducted a study in a business setting, i.e. he collected responses from about 30,000 managers from 55 countries. His research was motivated by the idea that everyone in the

45

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

a) Individualism vs. communitarism; according to which people belonging to individualistic cultures focus upon the enhancement of each individual’s rights, motivations, rewards and capacities, as opposed to those belonging to communitarian cultures in which attention is paid to the advancement of the community,

b) Achievement vs. ascription; societies more directed to achievement are those that appreciate individuals for their achievement and performance, while societies more directed to ascription are those that appreciate individuals for who they are

c) Universalism vs. particularism; which differentiates between societies in which rules should always be applied, from those in which one should allow for exceptions,

d) Inner vs. outer direction; which defines whether a society is predominantly populated by individuals directed to their own inner feelings and judgments, or by individuals directed to signals and trends in the outside world to which they must adjust,

e) Analysed specifics vs. integrated (diffuse) wholes; this defines whether decision makers in society are more effective when analysing parts, specifics, facts, numbers and units, or when all of these are taken together to form integrated wholes in the wider contexts,

f) Sequential time vs. synchronised time is the difference between the ability to perform one activity as fast as possible before moving on to the next, or to synchronise and coordinate more

46

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

g) Equality vs. hierarchy determines whether it is more important to treat people equally so they are intrinsically motivated to give their best, or to establish a hierarchical authority system and judge and evaluate people to extrinsically motivate them to achieve their best.

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

Although some of the dimensions from Hofstede’s (2001) study are repeated in the studies by Trompenaars (1994) and Schwartz (1999), both Trompenars and Schwartz contribute with new dimensions. Trompenars’ research focuses in particular on the nature of the individual, and it discovers four dimensions that deal with this question. Apart from Hofstede’s dimensions of individualism, power distance and uncertainty avoidance, those most often cited and used as antecedents in various cross-cultural studies are Trompenaar’s specific vs. diffuse (cf. Jun and Lee 2007) and achievement vs. ascription dimensions (cf. Gaál, Szabó, and Kovács 2007).

47

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

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

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

48

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).

49

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

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

Figure 2-6. Cultural Differences Distribution

Source: Romani 2004.

50

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.

51

3. Previous Research and Hypothetical Model

This chapter takes a deeper look at previous research exploring a direct interaction between brand and culture. Then two auxiliary concepts (brand benefits and consumer needs) are introduced into the aforementioned relationship. Finally, a hypothetical model of this research is developed.

3.1. The Influence of Culture on Brand Meaning

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

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

52

3.1.1. The Influence of Cultural on Brand Communication

A survey of leading brand advertisers in 15 countries discovered major differences in creative approaches among countries in terms of the appeals and messages portrayed (Synodinos, Keown, and Jacobs 1989). For example, while no Swedish advertisements refer to brand benefits, 77% of the Singaporean ones do; at the same time, Swedish advertisements most often use spokespersons and refer to product itself. The reason for creative communication differences among cultures might be ascribed to different marketing and advertising traditions; the practice of different advertising reflecting different national cultures; the processes of persuasion in general, and the patterns of emotions in particular (Orth, Koenig, and Firbasova 2007).

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

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

53

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

3.1.2. The Influence of Cultural on Brand Communication Interpretation

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

54

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

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

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

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

55

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).

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

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

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

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

S R2

R3 KR1

KR2

KR3

R1 M1

M2

M3

C

S = sender of communication «C» R = recipient of communication K = knowledge base of recipient M = meaning ascribed to

communication

56

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

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

Thus, the studies discussed above present a solid basis to acknowledge culturally embedded meaning creation and to introduce the first hypothesis of this research:

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

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

57

3.1.3. Cultural Dimensions as an Antecedent to Brand Meaning

Phau and Lau (2000) suggested that understanding cultural meaning requires linking it to the cultural dimensions identified by Hofstede (2001) and Trompenaars (1994). They propose that the meanings of brand personality dimensions differ between members of individualistic and collectivistic cultures. Brand personality traits that are consistent with interdependent self-construals (e.g. dependent, peaceful and harmonious) are emphasised by members of the collectivistic cultures, while in contrast, those consistent with independent self-construals (e.g. achievement, competition and independence) are emphasised by members of individualistic cultures.

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

Similarly, Pankhania, Lee, and Hooley (2007) incorporated the effect of the cultural dimensions of power distance and collectivism in their study. Their findings show that collectivist cultures assign more importance to the social value of brands and at the same time perceive a certain brand to be more socially-oriented than individualistic cultures do.

The afore-mentioned studies by Foscht et al. (2008) and Pankhania, Lee, and Hooley (2007) provide clear evidence that cultural dimensions influence brand perceptions by altering brand personality, brand attributes

58

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

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

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

H2: Brand meaning is affected by cultural dimensions.

In order to be able to predict how cultural dimensions affect brand meanings, a model that connects cultural dimensions to brand meanings

59

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

3.2. From Culturally Determined Needs to Brand Meanings

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

The process of brand meaning creation in the consumer’s mind is depicted in figure 3-2. As the figure shows, apart from key brand benefits (represented by a white ellipse), a brand holds those meanings (grey

60

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

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

The following paragraphs will introduce the concept of consumer needs and brand benefits, and then explain in which way they mediate the relationship between cultural dimensions and brand meanings.

3.2.1. Consumer Needs

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

Culturally determined consumer needs

Brand benefits Brand meanings

Emphasised consumer needs

Emphasised brand meanings

Legend

Key brand benefit

61

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

The evolution of these needs resembles Maslow’s (1943) hierarchy of needs, which classifies needs from the most essential physiological needs to the most advanced self-actualisation needs. He suggests that at the most basic level, the most essential and primitive needs, which are usually taken as the starting point for motivation theory, are the so-called physiological needs. They include hunger, thirst, sleep and the like. The next level of needs represents safety needs, which are expressed as the need for shelter from any kind of real or perceived threat. The third level of needs is represented by love needs as they were originally named by Maslow, but later rephrased into love and belonging needs or social needs. The fourth level of Maslow’s needs refers to esteem needs, which are defined as needs or desires for a stable, firmly based evaluation of oneself based upon real capacity, achievement and respect from others. The highest level of needs, self-actualisation, is characterised by people’s wish to fulfil their potential and possibly contribute to their society’s overall welfare for the sake of doing good and not for the sake of being recognised and merited by that act.

It is important for this particular study to stress that later cross-cultural research questioned and negated the hierarchical order of Maslow’s needs.

62

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

Apart from Maslow (1943), many other psychologists and anthropologists have studied needs as drivers of human decision-making and behaviour. One of the most important and widely cited works is that of Herzberg (1968). He conducted the study in a working environment and as a result discovered satisfactory needs and dissatisfactory needs. The former motivate people and can enforce the level of satisfaction, but their non-fulfilment will not result in dissatisfaction. The later, which are also called hygiene needs, result in dissatisfaction when non-fulfilled, whereas their fulfilment does not significantly affect the level of satisfaction.

In another study, Scitovsky (1968) classifies needs into a more parsimonious version of Maslow’s needs. He differentiates between physiological, social and intellectual needs. Raiklin and Uyar (1996) identified biological and socio-cultural needs, while Ryan and Deci (2000) suggest competence needs, autonomy needs and relatedness needs, especially in a job environment.

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

63

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

3.2.2. Brand Benefits

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

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

64

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

Another categorisation is provided by Sheth, Newman, and Gross (1991), who differentiate between functional value (functional, utilitarian, or physical performance); emotional value (the capacity to arouse feelings or affective states); social value (the association with one or more specific social groups); conditional value (the utility acquired as the result of a specific situation or a set of circumstances facing the choice maker); and epistemic value (the capacity to arouse curiosity, provide novelty, and/or satisfy a desire for knowledge). Later, Sweeney and Soutar (2001) proposed a reduced categorisation of benefits that consisted of only four values: functional value in terms of performance and quality, functional value in terms of value for money, emotional value, and social value.

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

65

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

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

Because of their brand specific context, Kapferer’s types of benefits will be used in the model of this study.

3.2.3. The Relationship between Culture, Consumer Needs, Brand Benefits and Brand Meanings

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

66

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

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

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

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

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

67

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

H2a: For consumers from cultures displaying higher collectivism, brand meanings will be more related to group activities, belonging, network, sacrifices for others, and common goals than for consumers from cultures expressing lower collectivism.

H2b: For consumers from cultures displaying lower collectivism, brand meanings will be more related to themes like individualism, autonomy, freedom, responsibility, self-sustainability than for consumers from cultures expressing higher collectivism.

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

68

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

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

H2c: For consumers from cultures displaying higher uncertainty avoidance, brand meanings will be more related to high quality, safety, value for money, standards, reliability, order, definitions, and consistency than for consumers from cultures displaying lower uncertainty avoidance.

H2d: For consumers from cultures displaying lower uncertainty avoidance, brand meanings will be more related to risk, adventure, innovation, excitement, unpredictability than for consumers from cultures displaying higher uncertainty avoidance.

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

69

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

H2e: For consumers from cultures displaying higher power distance, brand meanings will be more related to power, admiration, class difference, honour, and being better than others than for consumers from cultures displaying lower power distance.

H2f: For consumers from cultures displaying lower power distance, brand meanings will be more related to being approachable, equal, humble, and having a low profile than for consumers from cultures displaying higher power distance.

House et al. (2004) elaborate that countries low in assertiveness focus on the quality of life rather than performance; and on relationships between people rather than money and things. On the other hand, consumers in cultures displaying higher assertiveness will have a greater need for esteem and self-actualisation. At the same time, consumers in countries high in performance orientation will also value development, good performance, individual achievement and competition.

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

70

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

Park, Milberg, and Lawson (1991) discovered that Rolex is associated with status, wealth, luxury, fashion, etc. These are clearly the expressions of the self image of a consumer from a culture high in assertiveness and performance orientation. Therefore, such consumers are more likely than others to use brands mainly because brands help them to express their personal or business successes. However, this is not to say that the symbolism only lies in some particular brands, but rather it is in a consumer’s interpretation of brands in general. Proof of this can be found in a study by Coulter and Zaltman (1994). In a brand mapping process, they discovered that consumers connect the strong and dependable Tide detergent to self-confidence. They further claim that such associations to Tide help the brand’s consumers to be perceived as self-confident people. Thus, if Tide as a non-personal brand can provide such a benefit for some consumers, then it is plausible that any brand might have this ability. Hence, it can be concluded that a brand’s success is not entirely an objective category, but rather depends on subjective perception in a way that:

H2g: For consumers from cultures displaying higher assertiveness, brand meanings will be more related to assertiveness, advancement, toughness, aggression, control, and competition than for consumers from cultures displaying lower assertiveness.

H2h: For consumers from cultures displaying lower assertiveness, brand meanings will be more related to care, weakness, helpl, solidarity, tradition, life quality, and ambiguity than for consumers from cultures displaying higher assertiveness.

71

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

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

H2j: For consumers from cultures displaying higher human orientation, brand meanings will be more centred on love, care, social responsibility, and the environment than for consumers from cultures displaying lower human orientation.

H2k: For consumers from cultures displaying lower human orientation, brand meanings will be more centred on self-interest, material pleasures, parties, and hedonism than for consumers from cultures displaying higher human orientation.

72

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

H2l: For consumers from cultures displaying higher future orientation, brand meanings will be more related to investing, planning, saving, and sacrifices for the future than for consumers from cultures displaying lower future orientation.

H2m: For consumers from cultures displaying higher gender egalitarianism, brand meanings will be more related to gender equality and fairness than for consumers from cultures displaying lower gender egalitarianism.

In conclusion, table 3-1 summarises the afore-hypothesised connections between the four studied concepts: cultural dimensions (by House et al 2004), consumer needs (by Maslow 1943), brand benefits (by Kapferer 2004) and brand meanings. Most of the cultural dimensions are bipolar, i.e. cultures that are high in a certain dimension display a highly specific set of characteristics; whereas cultures that are low in that same dimension do not display a low level of the same set of characteristics. Instead, they display a specific set of their own characteristics. Therefore, for most cultural dimensions two hypotheses are constructed.

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

73

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

H Cultural dimension a(House et al. 2004)

Consumer need b(Maslow 1943)

Brand Benefit (Kapferer 2004)

Brand Meaning

H2a Collectivism + Social Badge Belonging, network, sacrifices for others, common, group

H2b Collectivism - Esteem

Hedonistic Continuity

Individual, autonomy, freedom, responsible, self-sustainable

H2c Uncertainty avoidance +

Safety Recognition cue Quality guaranteeOptimisation Continuity

High quality, safety, value for money, standards, reliable, order, definitions, consistency

H2d Uncertainty avoidance -

Esteem Badge Hedonistic

Risk, adventure, innovation, excitement, unpredictable

H2e Power distance + Esteem

Badge Hedonistic

Power, admirable, class difference, honourable, better than others

H2f Power distance - Social Badge Approachable, equal, humble, low profile

H2g Assertiveness + Esteem Badge Assertiveness, advanced, tough, aggression, in control, competition

H2h Assertiveness - Social Badge Ethical

Helpful, care, weakness, solidarity, tradition, life quality, ambiguity

H2i Performance orientation +

Esteem

Practicality Identification Badge

Efficiency, success effectiveness, progress competitiveness, target

H2j Human orientation + Social

Ethical Loving, care, social responsibility, environment

H2k Human orientation - Hedonistic Self-interest, material pleasure parties, tasty

H2l Future orientation + Safety Continuity Investing, planning, saving, sacrifice for the future

H2m Gender egalitarian.+ Social Ethical Gender equality, fair a the high pole of cultural dimension is indicated by +, whereas the low by – b The self-actualisation need is a very individually expressed type of need that might result in very different types of benefit extractions and very different brand meanings. For this reason, it is not included in table.

74

3.3. The Effect of Brand Usage and Brand Affection

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

Grier and Brumbaugh (1999) stress that the meanings constructed by target and non-target consumers may differ. Although they do not explicitly speak about users and non-users, their target and non-target consumer classification closely resembles the user and non-user classification in terms of how these two groups might understand the brand.

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

75

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

H3: Brand users (especially engaged ones) will emphasise positive brand meanings more than non-users, whereas non-users will emphasise negative meanings more than brand users (especially engaged ones).

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

H4: Differences in brand meaning emphasis between different types of users are not equal in each country.

Figure 3-3 presents a hypothetical model of this study. Because hypothesis 2 contains 13 sub-hypotheses, only the main one is depicted in the figure.

76

Figure 3-3. A Hypothetical Model

The following chapter will provide an overview of the methodology used in this thesis.

Culture

Type of User

Brand Meaning

H1

H2

H3H4

Cultural Dimensions

77

4. Research Methodology

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

4.1. Research Approach

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

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

78

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

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

4.2. Research Design

4.2.1. Research Methods

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

79

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

Table 4-1. Methodological Approaches

Stage Purpose Question Direction Methodological approach

Method

1 Define brand meaning What? Inductive Descriptive theory building

Interview

2 Test influence of culture on brand meaning

Is it as predicted?

Deductive Explanatory theory testing

Questionnaire

Source: Adapted from Snow and Thomas 1994.

Eisenhardt (1989) suggests that two-phase studies should be used to neutralise some problems inherited by method choice and get more accurate results. Recent studies embrace such an approach, even adding more phases (Abou Aish, Ennew, and McKechnie 2003; Ross and Harradine 2004). Both stages of this research were thus conducted in two phases (figure 4-1). The first phase in each stage served to test the measurement tool on a pilot sample, whereas the second served to apply the fine-tuned measurement tool on the multicultural sample to reach the targeted findings. In addition, the analysis of the output of the qualitative stage was used in the subsequent quantitative stage of the research.

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

80

“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”.

Figure 4-1. A Research Sequence

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

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

Cross cultural

differences in brand meaning

In-depth interviews with experts

In-depth interviews on a cross-cultural panel

Questionnaire on a cross-cultural sample

Questionnaire on a pilot sample

Experts’ questionnaire analysis

Pilot phase Main phase

Qua

litat

ive

stag

e Q

uant

itativ

e st

age

Brand meaning definition

Output

81

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

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

82

4.2.2. Units of Analysis

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

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

83

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

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

In addition to these three major conditions, it was considered necessary to think of the sample population when choosing appropriate brands for the study. This means the respondents had to be very familiar with the brands

84

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

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

Table 4-2. The Studied Brands

Brand Country of origin Product category Rank a Targeted consumer need

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

Nokia Finland Consumer electronics 5 Social

Toyota Japan Automobile 6 Safety, (esteem)

Google United States Internet Services 10 Social, Self-actualisation, (physiological)

Nike United Stated Sporting Goods 29 Self-actualisation (esteem)a based on brand value Source: according to Best Global Brands 2008 survey (Interbrand and Business Week 2008).

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

85

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

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

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

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

86

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

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

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

87

4.2.3. Population and Sampling

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

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

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

88

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

4.2.4. The Studied Countries

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

89

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

4.3. Variable Operationalisation

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

4.3.1. The Construct of Culture

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

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

90

Table 4-3. The Studied Countries’ Cultural Dimensions

Austria Finland India United StatesCultural dimension

practice value practice value practice value practice value

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

4.3.2. The Construct of Brand Meaning

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

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

91

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

Alber, Merunka, and Valette-Florence (2007) used a similar methodology in which subjects select one image among 19 offered for each brand. The chosen image symbolises their relationship with the brand. However, Coulter, Zaltman and Coulter (2001) emphasise that ZMET method’s strength lies in the fact that respondents, not researchers, supply the stimuli (i.e., visual images such as pictures or photographs) for the subsequent in-depth interviews. Being proactive in this task, the respondents are able to better articulate their thoughts and feelings, and identify issues that are important to them and potentially unknown to the researcher.

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

92

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

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

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

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

93

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.

94

5. Empirical Research on Brand Meaning

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

5.1. Pilot Research

In order to enable testing of the reliability of the findings, firstly the research procedure is transparently described (Yin 1994). In that way, subsequent researchers can repeat the research and arrive at the same conclusions (Denzin and Lincoln 1994). Subsequently, the results will be presented.

5.1.1. Research Procedure

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

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

95

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

5.1. 2. Results

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

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

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

96

5.2. Main Phase Research

5.2.1. Research Procedure

This subchapter offers a detail description of data gathering and data analysis procedures employed in the main phase of the qualitative research.

5.2.1.1. Data gathering

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

These conditions resembled the conditions set for the quantitative research, with one difference. The interviewees had to be business graduates, whereas the questionnaire respondents were still students. It was considered important to choose business graduates for the interviews as the interviewees then acted not only as consumers, but also as experts in the field during the interviews. In that way, they helped the researcher to grasp culture-specific issues that might have occurred in the later stage of the research. 5 In the validation studies of the ZMET procedure, Coulter, Zaltman, and Coulter (2001) discovered that four to five focused in-depth interviews can provide up to 90% of the information available from a larger set of interviews.

97

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

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

5.2.1.2. Data Analysis Procedure

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

98

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

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

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

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

99

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

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

5.2.3. Results

Altogether about 200 images (Appendix 1) and 150 pages of transcript were collected. For Coca-Cola, most of the interviewees submitted items depicting the Coca-Cola logo, Christmas associations, sugar and ice. In the case of Google, most of the images frequently depicted the Google logo, followed by the symbols representing ideas and searching activity. The vast majority of Nike’s images represented sport or activity in general. Nokia’s and Toyota’s images were dominated by phones and cars respectively.

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

100

Graph 5-1. The Percentages of Symbolic Images

0

20

40

60

80

100

Coca-Cola Google Nike Nokia ToyotaBrand

The

Perc

enta

ges o

f Sym

bolic

Im

ages

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

101

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

5.2.3.1. The Meanings of Coca-Cola

From 93 groups of meanings for Coca-Cola, as graph 5-2 shows, Coca-Cola was most often associated to refreshing (represented by the manifested meanings of refreshing, cooling, fresh and freshness). Then followed not healthy (with the manifested variables not healthy, no nutrition value, not safe, harmful and unhealthy), cold (with the manifested variables cold, cool, chilled, chilling, not hot), Christmas (with the manifested variables Christmas, Christmas song, Santa Claus and Christmas trucks) and sweet (with the manifested variables sweet, sugar and not soar). Tiia7 from Finland summarised it all to:

Traditional…original…refreshing…dr. pepper…Santa Claus…warm summer …refreshing drink8

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

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

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

102

Graph 5-2. The Meanings of Coca-Cola

0 2 4 6 8 10 12 14

Refreshing/cooling/fresh/freshnessNot healthy*

Cold/cool/chilled/chilling/not hotChristmas/Santa Claus/Christmas

song/Christmas truckSweet/sugar/not soar

Global/not localWarm summer/hot day

American/American lifestyleFizzy**

Energetic***Big/large/huge

Exciting****Recognisable/familiar/not

unknown/identifiableOpening bottle

Traditional/classic/not new/originalBonding with friends/parties/fun

Not natural

Finland India US Austria

*Not healthy/no nutrition value/not safe /harmful/ unhealthy

**Fizzy/carbonated/aerated/carbonation on tongue/ sound of bubbles/ sparkling

***Energetic/drink it when tired/no sleeping/ keeps me going/keeps me working

*** Exciting/not boring /not serious/not uncool /not sad

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

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

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

103

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

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

Country Meaning 1 Meaning 2 Meaning 3 Meaning 4

Austria Not healthy Refreshing Christmas Finland Refreshing Not healthy Christmas Global India Refreshing Cold Bonding with friends United States Cold Sweet

a Only the first meaning is mentioned – it represents the whole group of meanings

5.2.3.2. The Meanings of Google

In the analysis of the responses for Google, among 59 groups of meanings, the following were the most often mentioned: gives answers (with the manifested meanings gives answers, provides data, all knowing, informative, source of knowledge, key to everything, encyclopaedia, Wikipedia, information) and fast (with the manifested variables fast, quick, speed, not slow, efficient, timely, at any time) as graph 5-3 also shows. As for the individual answers, Anisha from India sent an image of a smiley face, and in the accompanying comment she explained that the image indicated she was satisfied and further added:

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

Google, like Coca-cola, also evoked some country-specific associations. For example, while the Finnish respondents considered it a life saver and generally had very positive opinions towards it, almost all of the Austrian

104

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

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

Graph 5-3. The Meanings of Google

0 2 4 6 8 10 12 14 16

Gives answers*Fast/quick/high speed/not slow/efficient /timely/at any

timeNot old**

World/worldwide/global

Powerful***

Life saver/basic necessity/very importantNo privacy/no freedom/no independence/

omnipresent/taking over our livesUseful/helpful

Google logo

Happiness/satisfaction

Huge/extensive/not small

Not reliable

Easy to use/convenient/no problem

Search

Not stupid/not dumb/not ignorant/smart/savvy

Reliable/trustful

Creative

Finland India US Austria

*Gives answers/provides data/ all knowing/ informative/source of information/key to everything /encyclopedia/wikipedia/ information/knowledge

**Not old/new/not antique/not historic/not old fashioned/ not out of date/up to date/young /innovative

***Powerful/potent/almighty/took earth control/dominant/ multi billion industry/not poor

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"

105

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

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

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

Country Meaning 1 Meaning 2 Meaning 3 Meaning 4

Austria Gives answers Not old Fast No privacy Finland Gives answers Fast World Life saver India Gives answers Fast United States Fast Powerful

a Only the first meaning is mentioned – it represents the whole group of meanings

5.2.3.3. The Meanings of Nike

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

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

106

Graph 5-4. The Meanings of Nike

0 2 4 6 8 10 12 14 16

Sports/ sportiness/ sportyAchiever/winner/victory/winning/not

failure/not losingNot formal/casual/be myself

SweatAthletes/athletic

Cool/trendyExpensive/pricey/not cheap

Sweat shops*Good quality/not low quality/durable

Activity**Sport shoes

Crowd cheeringCushion for feet***

Hi-tech/innovative/innovationJogging/runningMichael Jordan

PowerfulRough/not weepy/not weak

Self-assured/just do it/positivethinking

Swoosh

Finland India US Austria

*Sweat shops/child labor/not fairness/ exploitativeness/not socially responsible/not sustainable

**Activity/not lazy/active/not being immobile/not being still/movement

*** Cushion for feet/walking on clouds/ soft-footed/air technology

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).

107

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

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

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

Country Meaning 1 Meaning 2 Meaning 3

Austria Sports Sweat Sweat shopsFinland Sports India Sports United States Sports Expensive a Only the first meaning is mentioned – it represents the whole group of meanings

5.2.3.4. The Meanings of Nokia

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

Wooooooowwwwwww, very positive feelings…Nokia always indicates in India battery back up. Nokia means good battery that’s all.

108

Graph 5-5. The Meanings of Nokia

0 1 2 3 4 5 6 7 8 9

Reliable*Hi-tech/technology/engineered

Good/good investment/quality/safe choiceConnecting people

FinnishEasy to use /user friendly/functional/not complicated

Sturdy/strong/tough/hardCommunication

Smooth keypads/sleek/clean designFriends/family

Not a camera/just phone/phone/not diverseModern/trendy/not old fashioned/not old

First cell phoneInnovative

Not cutting edge**Global player/not local

Not exciting/not cool/not exoticNokia original ring tone

Durable/break-proof/long lasting

Finland India US Austria

*Reliable/will not ditch you/feeling of security/ handy/ there for you

** Not cutting edge/ not up to date/not innovator/not state of the art

Another interesting observation is that although most of the respondents strongly associated Nokia with their slogan connecting people, an Indian respondent viewed Nokia’s role in human relations quite differently. He said that:

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

Similarly a Finnish respondent admitted that Nokia represented:

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

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

109

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

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

Country Meaning 1 Meaning 2 Meaning 3

Austria Innovative Reliable Communication Finland Finnish Not exciting India Sturdy Reliable Good United States Hi-tech Smooth keypads

a Only the first meaning is mentioned – it represents the whole group of meanings

5.2.3.4. The Meanings of Toyota

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

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

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

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

110

Graph 5-6. The Meanings of Toyota

0 1 2 3 4 5 6 7 8 9 10

Reliable*Average/ok/good/practical/grey/safe

choice/rationalSafe

Not luxury**Innovative/hybrid car

Japanese****

Comfort/comfortable/relaxingExpensive/not cheap/overpriced

Not exciting***Family brand/children/family car

High qualityNot American/imported (for US)

Value for money/worth the moneyClass/luxury/high-end/regal/grace

Not for young/not for the nextgeneration

Not expensive/quite cheap

Finland India US Austria

*Reliable/no surprises /dependable/not unreliable/ not undependable/long lasting

**Not luxury/not premium /not exclusive/not high class/not status symbol/not the highest brand value

***Not exciting/not adventurous/not passion/ boring/not extravagant/not sexy/no emotions

****Japanese hard working engineering team/JIT/ robots/Japanese car

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

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

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

111

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

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

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

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

Country Meaning 1 Meaning 2 Meaning 3

Austria Reliable Not luxury Finland Reliable Average Not exciting India Comfort Expensive Class United States Reliable High quality

a Only the first meaning is mentioned – it represents the whole group of meanings

5.2.4. Discussion

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

112

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

5.2.4.1. Defining Brand Meaning

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

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

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

113

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

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

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

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

114

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

Meaning Stimulus Type of benefit Attribute/ Personality

Brand meaning category

Refreshing/cooling/fresh/freshness Meaning Functional/Experiential -

Attribute Product related

Not healthy/no nutrition value a/not safe a/harmful/unhealthy

Negative meaning

- Attribute Product related

Cold/cool/chilled a/chilling a/not hot a

Meaning Experiential/Functional Attribute Product related

Christmas/Santa Claus/Christmas song/Christmas truck a

Meaning Symbolic - Situational

Sweet/sugar/not soar b Meaning Experiential/Functional Attribute Product related

Global/not local Meaning Symbolic Attribute Provenance

Warm summer/hot day Meaning Experiential - Situational

American/American lifestyle Meaning Symbolic - Provenance

Fizzy b/carbonated/aerated b /carbonation on tongue b/sound of bubbles b/sparkling

Sensory image

Experiential Attribute Product related

Energetic/drink it when tired/no sleeping/keeps me going/keeps me working a

Meaning Functional/Experiential - Situational

Big b /large b /huge Adjective Symbolic Attribute Provenance

Recognisable/familiar/not unknown/identifiable

Meaning Symbolic Attribute Product related Provenance

Sound of opening a bottle b Sensory image

Experiential Attribute Product related

Traditional/classic/not new a /original

Meaning Symbolic Attribute Provenance

Bonding with friends/parties/fun Meaning Experiential /Symbolic - Situational

Exciting/not boring/not serious/cool

Meaning Symbolic/ Experiential Personality Symbolic

Not natural Negative meaning

- Attribute Product related

a Item excluded from the quantitative stage to retain parsimony. b Item excluded from the quantitative stage as it is a response to a non-direct-meaning question.

115

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

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

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

116

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

Table 5-7. The Types of Responses for Google

Meaning Stimulus Benefit Attribute/ personality

Brand meaning category

Gives answers/provides data /all knowing/ informative/ source of information a /key to everything a

/encyclopaedia/ Wikipedia a/ information a /knowledge

Meaning

Functional Attribute Product related

Fast/quick a /high speed/not slow a

/efficient /timely a /at any time/ Meaning Experiential

/Functional Attribute Product related

Not old/new/not antique a /not historic

a /not old fashioned/ not out of date/up to date a /young/ innovative

Meaning

Symbolic Attribute Product related

World/worldwide/global Meaning Symbolic Attribute Product related / provenance

Powerful/potent a /almighty/ took earth control /dominant /multi billion industry/not poor a

Meaning

Symbolic Personality Provenance

Life saver/basic necessity/very important

Meaning Experiential Attribute Symbolic/ product related/

No privacy/no freedom/no independence /omnipresent a /taking over our lives a

Meaning Experiential - Product related

Useful/helpful Adjective Functional Attribute, personality

Product related

Google logo c Meaning - - Sign Happiness/satisfaction Meaning Experiential - Product related Huge/extensive b /not small b Adjective Functional Attribute Product related/

provenance Not reliable b Adjective - Personality Product related Easy to use/convenient/no problem Meaning Experiential

/Functional Attribute Product related

Search a Meaning Functional - Product related Not stupid/not dumb/not ignorant/ smart/savvy b

Adjective Symbolic Personality Symbolic

Reliable/trustful Adjective Experiential Personality Product related Creative Adjective Experiential Personality Symbolic

a Item excluded from the quantitative stage to retain parsimony. b Item excluded from the quantitative stage as it is a response to a non-direct-meaning question. c Item excluded from the quantitative stage as no difference across cultures was predicted.

117

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

In relation to brand beliefs, as explained in chapter 2.1.1.5, the terminology and the semantics are rather complex. According to Orth and De Marchi (2007), beliefs are defined as features, attributes and benefits, and sometimes as brand cognitions. As afore elaborated, brand meanings include features, attributes and benefits, but are not limited to them.

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

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

118

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

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

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

However, although Franzen and Bouwman (2001) did not provide a unique consolidated definition of brand meaning, they clearly characterise it as a cognitive concept. This exploratory study provided evidence that many emotions are included in the brand meanings as well. Thus, the evidence from this study is more in line with the view of Richins (1994), who claims that an object’s meanings represent the totality of an individual’s thoughts and feelings about the importance of an object. He further differentiates public meanings from private ones, so that public meanings are likely to have an influence on shaping desire, while private meanings on consumers' feelings about the things they already possess.

119

Table 5-8. The Types of Responses for Nike

Meaning Stimulus Benefit Attribute / Personality

Brand meaning category

Sports/sportiness/sporty Meaning Functional/ Symbolic

Attribute Product related

Achiever/winner a /victory/winning/not failure/not losing a

Meaning Symbolic Personality Symbolic

Not formal/casual/be myself b Adjective Symbolic Personality Symbolic

Sweat Sensory image

Experiential - Product related

Athletes/athletic Meaning Functional/ Symbolic

Attribute Symbolic

Cool/trendy Adjective Symbolic Personality Symbolic

Expensive/pricey/not cheap Adjective - Attribute Perceived price

Sweat shops/child labour/not fairnessa /exploitativeness a /not socially responsible/not sustainable

Meaning - - Provenance

Good quality/not low quality/durable Adjective Functional Attribute Perceived quality

Not lazy/active/not being immobile a /not being still/activity/movement a

Meaning Symbolic Attribute Symbolic

Sport shoes Meaning Functional - Product related

Crowd cheering Sensory image

Experiential - Product related/ Symbolic

Cushion for feet/walking on clouds/soft-footed a /air technology

Sensory image

Experiential/Functional

Attribute Product related

Hi-tech/innovative/innovation Meaning Functional Attribute Product relatedJogging/running Meaning Functional - Product related /

situational Michael Jordan Meaning Symbolic - Advertising

Powerful Meaning Symbolic Personality Symbolic

Rough/not weepy b/not weak Adjective Symbolic Personality Symbolic

Self-assured/just do it/positive thinking

Meaning Symbolic Personality Symbolic

Swoosh c /Nike logo c Meaning - - Sign a Item excluded from the quantitative stage to retain parsimony. b Item excluded from the quantitative stage as it is a response to a non-direct-meaning question. c Item excluded from the quantitative stage because no difference across cultures was predicted.

120

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

Meaning Stimulus Benefit Attribute / Personality

Category of brand meaning

Reliable/will not ditch you/ feeling of security/ handy/ there for you

Meaning Experiential Personality Product related

Hi-tech/technology/ engineered Meaning Functional Attribute Product related

Good/good investment/quality/ safe choice

Meaning Experiential Attribute Perceived quality

Connecting people Meaning Functional Attribute Advertising

Finnish Meaning Symbolic Attribute Provenance

Easy to use /user friendly/ functional/not complicated

Meaning Experiential Attribute Product related

Sturdy/strong/tough/hard Meaning Experiential/Functional

Attribute Product related

Communication Meaning Functional - Product related

Smooth keypads/sleek/clean design

AdjectiveSensory i.

Experiential Attribute Product related

Friends/family Meaning Symbolic - Situational

Not a camera/just phone/phone/ not diverse

Meaning - - Product related

Modern/trendy/not old fashioned/ not old

Adjective Symbolic Personality Product related

First cell phone Meaning Symbolic - Situational/ provenance

Innovative Meaning Functional Attribute Product related

Not cutting edge/ not up to date/ not innovator/not state of the art

Negative adjective

- Attribute Product related

Global player/not local Meaning Symbolic Attribute Provenance

Not exciting/not cool/not exotic Negative adjective

- Personality Symbolic

Nokia original ring tone Meaning Experiential Attribute Brand sign

Durable/break-proof/long lasting Meaning Experiential/Functional

Attribute Product related

a The brand was not studied in the quantitative stage, so there was no need to identify the items for exclusion.

121

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

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

Robert actually implied that he has certain associations to Nokia, but because he did not have any experience with it, he has no feelings towards it and thus Nokia does not have any meaning to him.

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

122

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

Meaning Stimulus Benefit Attribute / Personality

Brand meaning category

Reliable/no surprises /dependable/not unreliable/ not undependable/long lasting

Meaning Functional Attribute Product related

Average/ok/good/practical/ grey/safe choice/rational

Meaning Functional Attribute Product related

Safe Meaning Functional Attribute Product related

Not luxury/not premium /not exclusive/not high class/not status symbol/not the highest brand value

Negative meaning

- Attribute Product related

Innovative/hybrid car Meaning Functional Attribute Product related

Japanese hard working engineering team/JIT/robots/ Japanese car

Meaning Symbolic - Provenance

Comfort/comfortable/relaxing Meaning Functional/ Experiential

Attribute Product related

Expensive/not cheap/overpriced Meaning - Attribute Perceived price

Not exciting/not adventurous/ not passion/ boring/not extravagant not sexy/no emotions

Negative meaning

- Personality Symbolic

Family brand/children/family car Meaning Symbolic - Symbolic

High quality Meaning Functional Attribute Perceived quality

Not American/imported (for Americans)

Adjective Symbolic Attribute Provenance

Value for money/worth the money Adjective Functional/ Experiential

Attribute Perceived price

Class/luxury/high-end/regal/grace Meaning Symbolic Personality Symbolic

Not for young/not for the next generation

Negative meaning

- Attribute Symbolic

Not expensive/quite cheap Negative adjective

Functional Attribute Perceived price

a The brand was not studied in the quantitative stage, so there was no need to identify the items for exclusion.

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

123

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 perspectiveBrand association Does not demand experience with the brand

Defining a construct in terms of what it is not is generally an important, but only a first step in defining what it is. The findings of this study provide a solid ground for understanding the brand meaning concept as a set of cognitive and affective experience-based brand associations.

5.2.4.2. Differences in the Brand Meanings across Cultures

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

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

124

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

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

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

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

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

Nokia is such a common phone brand in Finland - it is hard for me to separate general mobile from Nokia. It is the only product that we can take pride in on a global scale.

125

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

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

Question Brand 1 Brand 2 Brand 3 Brand 4 Brand 5

What brand would you most like to sit next to at a dinner party?

x x

What brand, if sent back 100 years, would have the biggest impact on history?

x x x

What brand that no longer exists would you resurrect?

x x x x x

Which brand do you want to argue with? x x x x

Which brand inspires you the most? x x If you were to describe yourself as being a brand, what brand would you be?

x x x

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

x x x x

What brand can you not live without? x x x

What brand is most likely to revolutionize the branding industry in the next five years?

x x x

What brand do you think is truly (going) "green"?

x x x x

Source: Thompson 2008.

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

126

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.

127

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

The second stage of the empirical research was conducted quantitatively with the aim of testing the hypotheses and discovering the differences in brand meanings among cultures.

6.1. Pilot Research

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

6.1.1. Research Procedure

6.1.1.1. Questionnaire Design

Dolnicar and Rossiter (2008) demonstrated that the low stability of consumers' brand-attribute associations (temporary associations) can be partly explained by the prevailing methods used in market research; hence, they propose seven methodological factors that influence the increased stability of brand-attribute associations. They are as follows: using the respondent’s favourite brand, making the evaluation task easier, directing the brand association relationship from the brand to the association, and using brands and languages the respondent is familiar

128

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

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

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

Coca-Cola is a drink we often encounter. Some of us like to drink it and some don't. As you look at the following questions, please answer them while keeping in mind how it feels to be in contact with Coca-Cola.

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

129

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

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

In addition to these four questions aimed at discovering brand meanings, the respondents also rated the extent to which they liked or disliked the brand in question, and whether they consumed it.

6.1.1.2. Questionnaire Distribution

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

10 www.surveymonkey.com

130

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

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

6.1.2. Results

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

131

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 = .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

132

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

Table 6-1. Coca-Cola’s Groups of Meanings and Their Hypothesised Relationships to the Cultural Dimensions

Group of meanings Meanings Cron.

Excluded items

Cron.

Cultural dimension

Refreshing Refreshing/cooling/fresh/freshness/cold /warm summer/hot day/ image of ice

.871 Image of ice, fresh

.878 No effect

Not healthy/harmful /unhealthy/not natural

.531 Harmful Unhealthy

Sweet/sugar/image of sugar .703 Image of sugar a

.676 Assertiveness –

Performance orientation -

Christmas Christmas/Santa Claus/Christmas song/image of Santa Claus

.900 Image of Santa Claus

.894 Human orientation +

Global Global/not local/huge/image of globe

.122 Not local, huge, image of globe b

- Power distance +

American American/American lifestyle/image of American flag

.740 Image of American flag

.684 Collectivism +

Carbonated Carbonated/sparkling .519 Carbonated, sparkling

- -

Energizing Energetic/drink it when tired/no sleeping/keeps me going

.728 - .728 No effect

Recognisable Recognisable/familiar/not unknown/identifiable

.428 - .428 Uncertainty avoidance+

Classic Traditional/classic/original .569 - .569 Uncertainty avoidance+

Assertiveness-

Bonding with friends

Bonding with friends/parties/fun/ image of friends

.624 Parties, image of friends

.707 Collectivism -

Cool Exciting/not boring/not serious/cool

.550 Not serious .650 Assertive +

a obesity, lot of calories added from Indian responses. b great brand, worldwide and multinational added from the pilot survey.

133

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

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

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

Meanings often mentioned in the pilot questionnaire were observes people’s habits, fun, entertainment, connecting to the world. The groups of meanings they belonged to are marked in table 6-3.

134

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

Group of meanings Meanings Cron.

Excluded items Cron.

Cultural dimension

Gives answers/provides data/all knowing/ informative/ encyclopaedia/knowledge

.797 Provides data, gives answers

Informative

Not dumb/ not stupid/smart/not ignorant

.650 Not dumb, not stupid, not ignorant

.713 Uncertainty avoidance +

Efficient Fast/high speed/efficient/timely/at any time/easy to use/convenient/image of speed

.633 Image of speed .677 Performance orientation +

Inspired Not old/new/not old fashioned/ not out of date/up to date/young/innovative/creative

.666 Not out of date/ not old fashion

.662 Uncertainty avoidance – Assertiveness + Performance orientation +

World/worldwide/global/huge/image of globe

.476 Image of globe a Almighty

Powerful/almighty/took earth control/dominant /multi billion industry/image of dominant

.596 Took earth control image of dominant

.705 Power distance +

Fun Happiness/satisfaction/no problem/image of smiley

.676 Image of smiley b .730 Human orientation -

Life saver/basic necessity/very important

.574

Reliable/trustful .680

Helpful

Useful/helpful .581

.742 Human orientation +

Assertiveness -

Intrusive No privacy/no freedom/no independence/image of freedom

-.821 Image of freedom c .466 Collectivism -

a connecting to the world added from the pilot survey b fun, entertainment added from the pilot survey c observes people’s habits added from the pilot survey In the case of Nike, several separate groups of meanings were strongly correlated, so for example sport, jogging and sport shoes formed one common group; achiever, self-assured and crowd cheering the second group; and pricey, good quality, cool and innovative the third group. All

135

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

Table 6-3. Nike’s Groups of Meanings and their Hypothesised Relationships to the Cultural Dimensions

Group of meanings Meanings Cron.

Excluded items Cron.

Cultural dimension

Sports Sports/sportiness/sporty/athletes/ athletic/jogging/running/sport shoes

.855 Sportiness, sport shoes a

.842 No effect

Activity (Renamed to Not weak)

Not lazy/active/not being still/activity

.374 Active, activity, not being still (Added not weak, not failure)

.598 Assertiveness

Sweat Sweat n.a. sweat - -

Achiever/victory/winning/not failure/image of victory

.678 Image of victory; not failure

Self-assured/just do it/positive thinking/ image of self-assured/crowd cheering

.558 Image of self-assured

Achiever

Michael Jordan n.a. b

.815 Performance orientation +

Assertiveness +

Powerful Powerful/image of powerful .281 Image of powerful - Power distance +

Rough Rough, not weak .113 Rough, not weak - -

Casual Not formal/casual -.212 Not formal, casual - -

Good quality/not low quality/durable .609 -

Expensive/pricey/not cheap .669 Not cheap

Cool/trendy .557 -

Innovative

Hi-tech/innovative/innovation/ image of high tech

.545 High tech, image of high tech

.746 Performance orientation +

Not CSR Sweat shops/child labour/not socially responsible/not sustainable

.612 Not Sustainable .662 Assertiveness –

Human orientation +

Cushioning technology

Cushion for feet/walking on clouds/air technology

.345 Cushion for feet, walking on clouds c

n.a. No effect

a added fitness from the Indian respondents’ interviews b added Tiger Woods from the American respondents’ interviews, whereas life and freedom from the pilot survey and young from the Austrian respondents’ interviews c added comfortable from the pilot survey

136

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

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

6.2. Main Phase Research

6.2.1. Research Procedure

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

137

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

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

The questionnaire was available online for 18 days in December 2008. It consisted of three sets of questions for the three studied brands and an introductory demographic set of questions. Three demographic questions were crucial to filter out the respondents who had lived abroad for longer than a year and longer than six months at a stretch during the last five years; those older than 31; and those not studying business or in a business related field. Such a filtering procedure ensured that the samples were as homogenous as possible.

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

138

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

6.2.2. Data Analysis Procedures

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

6.2.2.1. Data reduction method

There were 39 items representing 39 Coca-Cola’s meanings, 42 items representing Google’s meanings and 37 items representing Nike’s meanings. Such high numbers of variables are usually reduced to manageable number of factors or components. Barlett’s tests of sphericity were statistically significant (=.000 for all three brands) and Kaiser-Meyer-Olkin measures of sample adequacy were high (>.80 for all three brands), which indicated that items of each of the three brands correlated and were suitable for data reduction analyses. Because data reduction analysis is a highly influenced statistical method (i.e. it involves many analytical decisions by a researcher), it requires rigorous due diligence (Fabrigar et al. 1999).

139

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

According to Velicer and Jackson (1990), the correct number of extractions will reduce differences among different sub-methods. Furthermore, factor and principal component analyses are generally large sample methods that become increasingly stable with an increasing number of observations. However, the proposed ratios between the number of observed variables and the number of observations is just a rule of thumb and should not be used as a rule (Costello and Osborne 2005). In other words, when at least three or four measured variables represent each common factor or component and the communalities are an average of .70 or higher, even 100 observations suffice; whereas samples of even 800 are necessary with weaker results (Fabrigar et al. 1999).

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

140

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

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

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

6.2.2.2. Analysis of Variance

In order to investigate whether different meanings have different importance in different countries and for different types of users, a two-way analysis of variance (ANOVA) for independent samples was

141

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 2. Partial

2 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 2 < .06, medium if .06 < 2 < .14 and large if 2 > .14 (Sawyer and Ball 1981).

6.2.3. Results

6.2.3.1. Sample Characteristics

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

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

142

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

Altogether, 676 students participated in the research (205 Austrian, 148 Finnish, 178 Indian and 145 the United States respondents), out of which 59% evaluated all three brands. This percentage varied from 52% for India, 59% for Austria and the United States, to 68% for Finland. The total numbers of responses per each brand per country are presented in table 6-4.

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

Table 6-4. The Sample Characteristics

Initial sample Excluded a Final sample

Country TotalCoca-Cola Google Nike All

brands Inter.exp. 31+ Non-

bus. Total Coca-Cola Google Nike All

brands

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

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

143

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

6.2.3.2. Brand Meaning

The Brand Meanings for Coca-Cola

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

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

In order to prepare data for the ANOVA tests (providing equality of variance is met), a principal components analysis was conducted on 39 initially measured items. A total of 366 observations presented a ratio of items to observations of 1:9. The Scree plot (figure ap4-1.) suggested

144

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

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

Item Median b Countries c Item Median b Countries d

worldwide 7 US, F, A refreshing 5 recognisable 7 US, I, F, A obesity 5 not unknown 7 I, F, A cooling 5 global 7 A cool 5 identifiable 7 US, F traditional 5 I not healthy 7 US freshness 5 multinational 6 I parties 5 American 6 warm summer 5 sugar 6 fun 5 great brand 6 I hot day 5 not natural 6 energetic 5 sweet 6 drink it when tired 4 American lifestyle 6 Christmas 4 I unhealthy 6 Santa Claus 4 I cold 6 Christmas song 4 US, I original 6 exciting 4 F, A classic 6 bonding with friends 3 US, F, A familiar 6 no sleeping 3 US, F, A calories 5,5 keeps me going 3 US, F, A not boring 5 a N = 366; all countries sample; data were sorted primarily on median and secondarily on mean. b Medians close to 1, show the items very poorly indicated, while those close to 7 show that the items clearly indicated what each brand means to the respondents. c The Country initial indicates a country in which the item was evaluated as one of the top four. d The Country initial indicates a country in which the item was evaluated as one of the bottom four. The number of components was then increased to eight and nine, but the solutions were still not satisfactory as either many cross-loadings appeared, or the items did not load above .32 on either of the components.

145

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

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

Component

Item Refreshing Globally recognised Unhealthy ChristmasClassic Energising

Bonding with

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

146

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 ( . > .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 ( . < .70), but still relatively high ( . > .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.

147

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

Country Refreshing Globally recognised Unhealthy ChristmasClassic Energising

Bonding with

friends

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

As shown in table ap4-2, the total variance explained by the seven components was 59.3%. The Pearson’s correlation matrix is presented in table 6-8 for these components.

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

Component Refreshing Globally recognised Unhealthy Christmas a Classic Energising

Bonding with

friends a

Refreshing 1Globally recognised

.247** 1

Unhealthy -.165** .086 1Christmas -.169** -.095 .009 1Classic .135** .179** -.044 -.160** 1Energising .188** -.027 -.098 -.240** .076 1Bonding with friends

-.254** .047 .177** .143** -.054 -.309** 1

Minimum -4.80 -5.15 -3.57 -2.07 -3.88 -2.50 -3.04Maximum 2.11 1.47 1.96 1.93 2.52 2.85 2.97K-S z (sig.) 1.29 (.07) 2.18 (.00) .70 (.72) 1.46 (.03) 1.28 .74 (.64) .80 (.54) * p < .05 ** p < .01 a Smaller values indicate a higher emphasis

The Kolmogorov-Smirnov tests of normality revealed that Globally recognised (p = .00) and Christmas (p = .03) were not normally

148

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

Most of the components’ correlations were statistically significant, but relatively weak, with the strongest correlation existing between Energising and Bonding with friends (r = .31**).

The Brand Meanings for Google

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

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

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

149

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

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

Item Median b Countries c Item Median b Countries d

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

The obtained components could not be interpreted well, and as many as 11 communalities were less than .40. This was a sufficient condition for adding an additional component (Costello and Osborne 2005). Hence, the number of components was increased in each consecutive analysis until a

150

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

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

Table 6-10. The Pattern Matrix for Google

Component

Item Almighty Helpful Intrusive Inspired Fun Global Informative

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

151

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

Extraction Method: Principal Component Analysis; Rotation Method: Oblimin with Kaiser Normalization; Rotation converged in 22 iterations. To investigate whether the same items formed the same components in each of the studied countries, reliability tests were conducted. As the results in table 6-11 show, the Cronbach’s alphas were above the critical point of > .70 for five components in all the countries, which indicates consistency among their top items. For the Fun and Intrusive components, most of the Cronbach’s alphas were less than < .70.

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

152

the United States ( = .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 InformativeOverall .86 .85 .63 .83 .73 .81 .72 Austria .82 .86 .65 .82 .67 .82 .63 Finland .80 .84 .77 .74 .51 .82 .63 India . 84 .83 .42 .80 .78 .76 .80 United States .87 .83 .44 .81 .66 .82 .79

a Only items whose loadings were the highest on a particular component (in comparison to their loadings on other components) were included in each component’s reliability tests (Cronbach’s alphas).

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

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

Component Almighty Helpful a Intrusive Inspired Fun a Global a Informative

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

153

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

The Brand Meanings for Nike

To uncover which meanings were the most associated to Nike, the medians for each meaning were calculated and sorted from the highest to the lowest in table 6-13.

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

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

154

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

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

Item Median b Countries c Item Median b Countries d

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

155

Table 6-14. The Pattern Matrix for Nike

Component

Item Achiever Not weak Not socially responsible Pricey Sports

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

156

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

Country Achiever Not weak Not socially responsible Pricey Sports

Overall .94 .73 .67 .87 .88Austria .93 .72 .65 .87 .88Finland .93 .78 .80 .82 .83India .94 .77 .07 .75 .90United States .94 .58 .77 .91 .91a Only items whose loadings were the highest on a particular component (in comparison to their loadings on other components) were included in each component’s reliability tests s(Cronbach’s alphas). There was a problem in terms of consistency for the Not socially responsible component for the Indian data (table 6-15), for which the Cronbach’s alpha reached as low as .07. Since further analysis was conducted with normalised components, the low Cronbach’s alpha of the Not socially responsible component did not pose a major problem; nevertheless, more caution was necessary when interpreting the results for this meaning in India.

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

157

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

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

Component Achiever Not weak Not socially responsible Pricey a Sports a

Achiever 1Not weak .057 1Not socially responsible -.130* -.034 1

Pricey -.223** .015 .021 1Sports -.474** -.214** .049 .272** 1Minimum -3.05 -3.83 -2.70 -1.77 -1.45Maximum 2.16 2.18 2.53 2.87 3.92K-S z (sig.) .45 (.99) .56 (.92) .62 (.83) 1.26 (.08) 1.69 (.01)

* p < .05 ** p < .01 a Smaller values indicate higher emphasis

6.2.3.3. Types of Users

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

158

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

Table 6-17. The Distribution of the Respondents based on their Usage and Feelings for the Brands a

Feeling

Brand Type of user Love it Like it Indifferent/

mixed feelings Not like it Hate it Total

Users 291 1061 612 42 12 201 Non-users 23 153 633 223 33 105 Nike Total 31 121 124 26 4 306

Users 1354 1595 215 0 0 315 Non-users 0 15 15 0 15 3 Google Total 135 160 22 0 1 318

Users 366 1566 817 207 37 296 Non-users 0 68 308 288 108 74 Coca -cola Total 36 162 111 48 13 370

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

159

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

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

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

Differences in Coca-Cola’s Brand Meanings

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

The results of the ANOVA tests for the independent variable, Country, are displayed in table 6-18. Country had a statistically significant effect on the meanings: Refreshing (F (3, 354) = 3.69; p < .05), Globally recognised (F (3, 354) = 5.03; p < .01) Unhealthy (F (3, 354) = 8.71, p < .01), Christmas (F (3, 354) = 9.17; p < .01), Classic (F (3, 354) = 30.67; p < .01) and Bonding with friends (F (3, 354) = 30.35; p < .01). Size of effect was considered weak for Refreshing (partial 2 = .03) and Globally recognised

160

(partial 2 = .04), moderate for Unhealthy (partial 2 = .07) and Christmas (partial 2 = .07), and strong for Classic (partial 2 = .21) and Bonding with friends (partial 2 = .21), which indicated that between 3% and 21% of the variance in the meanings was explained by the variable Country.

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

Brand Meaning

Austria (1)

Finland (2)

India (3)

United States (4)

F (3, 354)

Partial 2

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

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

The Refreshing meaning was not supposed to be influenced by any of the cultural dimensions. Hence, it was not supposed to be differently

161

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

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

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

162

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

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

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

Christmas might have symbolised excessive consumerism – a

phenomenon contrary to what is propagated by Christmas. So, to express their non-agreement with such a view of Christmas, such respondents

163

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

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

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

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

164

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

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

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

165

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

Finally, the difference in the emphasis of the meaning Energising, as predicted, was not statistically significant among the countries.

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

166

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

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

Brand Meaning Engaged

users (1)

Disengaged users

(2)

Non-users (3)

F (2, 354)

Partial 2

Refreshing .378 (.768)2 3 -.324 (.994)1 -.523 (1.145)1 24.58** .121Globally recognised .088 (.898) -.089 (1.135) -.104 (1.043) 2.96 .016Unhealthy -.401 (.946)2,3 .244 (.829)1,3 .693 (.863)1,2 28.20** .137Christmas b -.138 (.954)3 .128 (1.013) .178 (1.058)1 4.83** .027Classic .148 (.903)2 -.306 (1.087)1 .041 ( 1.031) 6.97** .038Energising .202 (.929)3 -.020 (1.064)3 -.490 (.921)1,2 10.63** .057Bonding with friends b -164 (1.006) .119 (.953) .256 (.982) 2.98 .017a The values outside the brackets represent the countries’ means, whereas the values in the brackets represent the corresponding standard deviations. The indices indicate which groups are significantly different in statistical terms (Bonferroni post-hoc test). b The negative values represent high emphasis of the meaning. * p < .05; ** p < .01 Furthermore, Classic is a meaning for which the post-hoc tests revealed that the Engaged users (M = .15; SD = .90) had a statistically significantly higher result than the Disengaged users (M = .31; SD = 1.08). The Non-users (M = .04; SD = 1.03) did not significantly differ from either of the former two types of users. Finally, the results obtained in the post-hoc tests for the Christmas dependent variable revealed that the Engaged users emphasised the Christmas meaning statistically significantly more (M= -.14; SD = .95) than the Non-users (M = .18; SD = 1.06) did; whereas the Disengaged users (M = .13; SD = 1.01) occupied a place between the first two groups and showed no statistical difference from either.

167

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

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

To measure the interaction effects, (i.e. Country and Type of Coca-Cola user) the sample was divided into twelve subgroups. The interaction effects were not statistically significant for any of the dependent variables: Refreshing (p = .31), Globally recognised (p = .53), Unhealthy (p = .06), Christmas (p = .35), Classic (p = .23), Energising (p = .20) and Bonding with friends (p = .35).

Differences in Google’s Brand Meanings

In order to determine whether differences in Google’s meanings existed between the different countries and different types of Google users, 4 (countries) x 2 (type of Google user) two-way ANOVAs were conducted. The seven previously identified meanings were measured in continuous numerical values and each one represented a dependent variable in a separate ANOVA. The variances in the five meanings (i.e. Helpful, Intrusive, Inspired, Global, and Informative) were not homogeneous across the groups, so non-parametric tests were also conducted for these

168

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 2 = .03), Global (partial 2 = .05), and Informative (partial 2 = .05); moderate for Almighty (partial 2 = .10) and Inspired (partial 2 = .08); and strong for Fun (partial 2 = .14). This indicated that between 3% and 14% of the variance in the meanings was explained by the variable Country.

The post-hoc tests for the Almighty dependent variable revealed that the respondents from India (M = .61; SD = .74) and the United States (M = .37; SD = .88) emphasised the Almighty meaning more than the Austrian (M = -.46; SD = 1.01) and the Finnish respondents (M = -.20; SD = .92) and that the difference was statistically significant. Neither the former two, nor the latter two countries differed from one another in statistical terms.

It was hypothesised that the Almighty meaning would be affected by power distance. This is because high power-distance countries that view power as providing social order, relational harmony and role stability (House et al. 2004 p. 540) put more emphasis on this meaning. The results were in line with the Indian (PDI = 2.64) and American (PDUS = 2.85)

169

higher scores on power distance than the Austrian (UAA = 2.44) and Finish (UAF = 2.19) lower scores on this dimension.

Table 6-20. The Effects of Culture on the Google Meanings a

Brand Meaning

Austria (1)

Finland (2)

India (3)

United States (4)

F (3, 309)

Partial 2

Almighty -.462 (1.011)3 4 -.196 (.916)3 4 .612 (.743)1 2 .370 (.876)1 2 10.76** .095 Helpful b .188 (1.293) -.022 (.942) -.066 (.755) -.229 (.652) .596 .006 Intrusive .247 (1.032)2,4 -.156 (1.078)1 .135 (.837)4 -.392 (.852)1,3 3.674* .034 Inspired -.350 (1.080)3,4 -320 (.877)3,4 .475 (.695)1,2 .556 (.896)1,2 9.15** .082 Fun b .455 (1.007)3,4 .220 (.849)3,4 -.703 (.867)1,2,4 -.272 (772)1,2,3 17.18** .143 Global b -.199 (.986)2 .328 (1.082)1,3 -.229 (.785)2 .139 (.995) 4.829** .045 Informative .083 (1.042)2 -.534 (1.048)1,3,4 .345 (.681)2 .273 (.847)2 5.43** .050

a The values outside the brackets represent the countries’ means, whereas the values in the brackets represent the corresponding standard deviations. The indices indicate which groups are significantly different in statistical terms (Bonferroni post-hoc test). b The negative values represent a high emphasis of the meaning. * p < .05; ** p < .01 The Post-hoc tests for the Intrusive dependent variable revealed that the Austrians (M = .25; SD = 1.03) emphasised the Intrusive meaning statistically significantly more than the Finns (M = -.16; SD = 1.08) and Americans (M = -.39; SD = .85) did. The Indians’ (M = .14; SD = .84) emphasis of the Intrusive Google meaning was the second highest and statistically significantly different from the Americans’ lowest emphasis.

It was predicted that Intrusive would be influenced by in-group collectivism, because individuals in cultures low in in-group collectivism tend to look after themselves and view themselves as autonomous and independent of groups (House et al 2004, p. 454). Such cultures were thus also expected to consider Google’s private data gathering as a privacy infringement. The results for Intrusive were in line with the Austrian (CA = 5.27) lowest score, the Indian second lowest score; (CI = 5.32), and the American highest score (CUS = 5.77) on the cultural dimension of in-group collectivism.

170

Similarly, the post-hoc tests for the Inspired meaning revealed that the Indians (M= .48; SD = .70) and the Americans (M = .56; SD = .90) emphasised the Inspired meaning statistically significantly more than the Austrians (M= -.35; SD = 1.08) and the Finns (M = -.32; SD = .88), while the upper and lower subgroups did not differ statistically significantly between each other.

The Inspired meaning was predicted to be affected by uncertainty avoidance, assertiveness and performance orientation. The four studied countries all have almost the same results on performance orientation, so the influence of performance orientation had no influence on the results. Cultures low in uncertainty avoidance facilitate new product development, show less resistance to change and have a higher tolerance for breaking rules (House et al. 2004, p. 618); while those high in assertiveness value success, progress, taking initiative, and have a “can do” attitude (House et al. 2004, p. 405). Thus, countries low in uncertainty avoidance and high in assertiveness were both predicted to put a higher emphasis on Inspired meaning. Since the studied countries have a reverse score on these values, and since assertiveness has a stronger variance, it is also predicted to have a stronger effect (as already mentioned for the Unhealthy meaning of Coca-Cola). It can be claimed that the results for the Inspired meaning were in line with the Indian high score (AI = 4.76), the Finnish (AF = 3.68) and the United States (AUS = 4.32) moderate scores, and the Austrian (AA = 2.81) low score on assertiveness.

According to the results of the post-hoc tests for the Fun meaning, the Austrians (M = .46; SD = 1.00) emphasised the Fun meaning of Google statistically significantly less than the secondly positioned United States (M = -.27; SD = .77) and firstly positioned India (M = -.70; SD = .87). The third lowest emphasis of Fun displayed by Finland (M = .22; SD = .85) is also statistically significantly different from the score of the United

171

States and India. Finally, a difference in the emphasis of Fun was also statistically significant between the United States and India.

Fun was predicted to be influenced by human orientation because cultures that are less human oriented tend to value pleasure, comfort, and enjoyment more (House et al. 2004, p. 570). The results for Fun were thus in line with the prediction. In other words, India values human orientation (HOI = 5.28) the least, followed by the United States (HOUS = 5.53), which was also the order of the scores for the Fun meaning.

The post-hoc tests’ results for Global showed that the Finnish respondents (M = .33; SD = 1.08) emphasised the Global meaning of Google statistically significantly less than the Indian (M = -.23; SD = .79) and Austrian respondents (M = -.20; SD = .99) did. The American (M = .14; SD = 1.00) emphasis was slightly (not significantly) higher than the Finnish one.

The Global meaning was predicted to be affected by the cultural dimension of in-group collectivism, according to which cultures that have a relatively higher appreciation of in-group collectivism emphasise global relatedness and connectedness among members of the (Google) society (House et al. 2004, p. 454). However, the results were opposite from what was expected as Austria and India put the highest emphasis on the meaning Global, and at the same time value in-group collectivism the least (CI = 5.32; CA = 5.27).

The results of the post-hoc tests for the Informative dependent variable revealed that the Informative meaning was statistically significantly less emphasised in Finland (M = -.53; SD = 1.05) than in the other three countries, i.e. Austria (M = .08; SD = 1.04), the United States (M = .27; SD = .85), and India (M = .35; SD = .68). The latter three countries did not differ in their emphasis of the meaning Informative.

172

Informative was predicted to be affected by the uncertainty avoidance cultural dimension. In other words, cultures which greatly value uncertainty avoidance, appreciate order, formalised interactions, and rules also tend to keep experiences documented (House et al 2004, p. 618). Hence, it was also expected that such cultures would highly appreciate the Informative meaning. The results were partly confirmed. Since Austria (UAA = 3.66) has the lowest and Finland (UAF = 3.85) the second lowest score for uncertainty avoidance, the Austrian score for Informative should have also been statistically significantly lower than the scores of the United States and India.

Finally, there was no statistically significantly difference among the countries in their emphasis of the Helpful meaning. This result was the opposite from what was expected, since the Helpful meaning was supposed to be affected by low assertiveness and high human orientation.

The results of the ANOVA tests for the independent variable, Type of Google user, are displayed in table 6-21. The Type of Google user had a statistically significant effect on the meanings: Almighty (F (1, 309) = 30.21; p < .01), Helpful (F (1, 309) = 15.3; p < .01), Inspired (F (1, 309) = 20.3; p < .01), Fun (F (1, 309) = 6.36; p < .05), Global (F (1, 309) = 25.41; p < .01), and Informative (F (1, 309) = 15.0, p < .01). The emphasis of all the positive meanings (Almighty, Helpful, Inspired, Fun, Informative, and Global) was, as expected, higher among the Engaged users, as opposed to the Disengaged users.

Intrusive was the only identified and explored negative meaning of Google. The difference in emphasis of Intrusive between the Engaged users and the Disengaged users was not statistically significant. This was not in line with the prediction that the Disengaged users would emphasise

173

the Intrusive meaning statistically significantly more than the Engaged users.

The effect size was considered small for Helpful (partial 2 = .05), Fun (partial 2 = .02), and Informative (partial 2 = .05), while it was moderate for Almighty (partial 2 = .09), Inspired (partial 2 = .06), and Global (partial 2 = .08). The exposed effect sizes indicated that between 2% and 9% of the variance in the meanings was explained by the variable Type of Google user.

Table 6-21. The Effects of Type of Google User on the Google Meanings a

Brand Meaning

Engaged users (1)

Disengaged users (2)

F (1, 309)

Partial 2

Almighty .504 (.781) -.369 (.983) 30.21** .089Helpful b -.291 (.640) .213 (1.153) 15.3** .047Intrusive -069 (.928) .051 (1.049) .126 .000Inspired .461 (.856) -.337 (.964) 20.3** .062Fun b -.349 (.878) .255 (1.009) 6.36* .020Global b -303 (.719) .222 (1.114) 25.414** .076Informative .313 (.788) -.229 (1.076) 15.0** .046

a The values outside the brackets represent the countries’ means, whereas the values in the brackets represent the corresponding standard deviations. b The negative values represent high emphasis of the meaning. * p < .05; ** p < .01 In order to measure the interaction effects, the combination of the two independent variables, Country and Type of Google user, resulted in a total of eight subgroups. All of the subgroups had more than 20 observations, except for the group of the Disengaged American users. This group has 14 respondents and hence the results had to be interpreted with caution. The interaction effect was statistically significant for the Informative meaning (p < .01). Graph 6-1 shows that the Engaged Austrian users emphasised the Informative meaning less than the Disengaged Austrian users, while the Engaged users in the other three

174

countries emphasised the Informative meaning more than the Disengaged users. The interaction effect size is considered small (partial 2 = .05), i.e. 5% of variance in the meaning Informative was explained by this interaction.

This interaction effect was not predicted. However, this result shed more light on the results of the effect of Country on Informative reported previously. The uncertainty avoiding dimension predicted that the Austrians would emphasise Informative meaning the least. The results got distorted because of the inexplicably high score of the Disengaged Austrian users.

Graph 6-1. The interaction Effect on the Google Informative meaning

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).

United States

Disengaged users Engaged users

0,60

0,40

0,20

0,00

-0,20

-0,40

-0,60

-0,80

India Finland Austria

Country

Em

phas

is o

f the

Info

rmat

ive

Mea

ning

Type of Google user

175

Differences in Nike’s Brand Meanings

A set of 4 (country) x 3 (type of Nike user) two-way ANOVAs was conducted in order to determine whether differences in Nike’s meanings existed among the different countries and the different types of users. The five previously identified meanings were measured in continuous numerical values and each represented a dependent variable in a separate ANOVA. Variances of the meaning Sports were not homogeneous across the groups (ap4-34), so non-parametric tests were applied for this meaning. Country is a nominal independent variable which differentiates between four groups of respondents. Type of Nike user represents an ordinal numerical variable with three levels.

The results of the ANOVA tests for the independent variable Country are displayed in table 6-22. Country had a statistically significant effect on the following meanings: Achiever (F (3, 294) = 8.96; p < .01), Not socially responsible (F (3, 294) = 6.96; p < .01) and Pricey (F (3, 294) = 17.46, p < .01). The effect sizes were considered moderate for Achiever (partial 2 = .08) and Not socially responsible (partial 2 = .07), and strong for Pricey (partial 2 = .15), which indicated that between 7% and 15% of the variance in the meanings was explained by the variable Country.

The post-hoc tests for the Achiever dependent variable revealed that the Indians emphasised the Achiever meaning statistically significantly more (M= .54; SD = .86) than the Austrians (M = -.21; SD = 1.03) and Finns (M = -.22; SD = .96) did. The American score (M = .10; SD = .91) was centrally situated and not statistically significantly different from any of the other three scores.

The Achiever meaning was expected to be influenced by performance orientation and assertiveness. Highly performance-oriented and highly assertive cultures value competition, success and progress (House et al.

176

2004, p. 245, 405). Hence, countries high in these cultural dimensions were predicted to emphasise the Achiever meaning more than others. Since all four countries display a similar appreciation of performance orientation, the results were evaluated in terms of assertiveness and found to be as predicted. India had a high score (AI = 4.76), the United States (AUS = 4.32) a moderate score and Austria (AA = 2.81) and Finland (AF = 3.68) low scores on valuing assertiveness.

Table 6-22. The Effects of Culture on the Nike Meanings a

Brand Meaning

Austria (1)

Finland (2)

India (3)

United States (4)

F (3, 294)

Partial 2

Achiever -.207 (1.032)3 -.218 (.964)3 .539 (.858)1.2 .102 (.908) 8.96 ** .084 Not weak -.012 (1.076) .016 (.867) -.148 (1.117) .195 (.888) 1.971 .020 Not socially responsible .217 (1.008)2,3 -.140 (1.026)1,4 -.467 (.647)1,4 .439 (1.040)2,3 6.96** .066 Pricey b -.037 (.937)2,3 .581 (.864)1,3,4 -.634 (.753)1,2,4 -.136 (1.080)2,3 17.463** .151 Sports b .019 (1.085) .114 (.926) -.078 (.984) -.144 (.973) 2.235 .022

a The values outside the brackets represent the countries’ means, whereas the values in the brackets represent the corresponding standard deviations. The indices indicate which groups are significantly different in statistical terms (Bonferroni post-hoc test). b The negative values represent high emphasis of the meaning * p < .05; ** p < .01

The post-hoc tests for the Not socially responsible meaning revealed that there was a statistically significant difference between the Austrians (M = .22; SD = 1.01) and the Americans (M = .44; SD = 1.04) on one side, and the Indians (M = -.47; SD = .65) and the Finns (M = -.14; SD = 1.03) on the other side. The former two emphasise the Not socially responsible meaning more than the later two.

Not socially responsible was predicted to be influenced by two cultural dimensions; specifically, assertiveness and human orientation. Countries low in assertiveness have sympathy for the weak (House et al. 2004, p. 405); while those high in human orientation value altruism, benevolence and generosity (House et al. 2004, p. 570). Both sets of characteristics

177

belong to people who would emphasise the Not socially responsible meaning. In comparison to the other countries, India values assertiveness the most (AI = 4.76) and human orientation (HOI= 5.28) the least. This was in line with its very low emphasis of the Not socially responsible meaning for Nike. Austria, on the other hand, has the opposite results. It values assertiveness (AA = 2.81) the least and human orientation the most (HOA = 5.76). Consequently, Austria had the highest score on the Not socially responsible meaning emphasis. Finland (HOF = 5.81) is high in human orientation but moderate in assertiveness (AF = 3.68), which was also in line with its position between Austria and India in emphasising the Not socially responsible Nike meaning. However, the highest emphasis of the Not socially responsible Nike meaning in the United States could not be explained by its score on the afore-mentioned two cultural dimensions (AUS = 4.32; HOUS = 5.53).

The post-hoc tests for the Pricey meaning revealed that the Indians (M = -.63; SD = .75) emphasised this meaning statistically more than all the other studied countries. Austria (M = -.04; SD = .94) and the United States (M = -.14; SD = 1.08) followed. They did not statistically significantly differ from each other in their Pricey meaning emphasis, but they emphasised Pricey statistically significantly more than Finland did (M = .58; SD = .86).

Pricey was not predicted to be influenced by culture, so the results were not as predicted. However, the meaning Pricey was most probably not influenced by cultural dimensions. It seemed that in the case of Nike, India’s economic situation might have triggered the question of Nike’s affordability and affected differences in its Pricey emphasis.

As predicted Sports was equally emphasised across countries. Finally, the Not weak meaning was extracted in the qualitative phase as a result of

178

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 2 = .12), Not weak (partial 2 = .08), Not socially responsible (partial 2 = .12), and Sports (partial 2 = .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

Brand Meaning Engaged users (1)

Disengaged users (2)

Non-users (3)

F (2, 294)

Partial 2

Achiever .418 (.956)2,3 -.271 (.808)1 -.367 (.967)1 19.659** .117 Not weak .281 (.990)3 -.057 (.875) -.325 (.991)1 12.867** .080 Not socially responsible -.454 (.868)2,3 .314 (.970)1 .386 (.946)1 19.98** .120 Pricey b -.056 (.975) .123 (.960) -.005 (1.058) .869 .006 Sports b -.353 (.786)2,3 .156 (1.019)1 .356 (1.087)1 19.635** .118 a The values outside the brackets represent the countries’ means, whereas the values in the brackets standard deviations. The indices indicate which groups are statistically significantly different (Bonferroni post-hoc test). b The negative values represent high emphasis of the meaning * p < .05; ** p < .01 The post-hoc tests for the Achiever meaning revealed that the Engaged users emphasised the Achiever meaning statistically significantly more (M = .42; SD = .96) than the Disengaged users (M = -.27; SD = .81) and the

179

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

180

more than for the Engaged users. These results were in line with the prediction.

To measure the interaction effects, the combination of the two independent variables Country and Type of Nike user resulted in a total of twelve subgroups. The interaction effects were not statistically significant for any of the meanings as can be seen from the following p values: Achiever (p = .58), Not weak (p = .56), Not socially responsible (p = .14), Pricey (p = .56), and Sports (p = .22).

Overall, this section provided the necessary data for the evaluation of hypotheses 1, 2, and 3. The results were individually commented in regards to how they met the hypothesised predictions. Evaluation of hypotheses on an overall basis is conducted in the discussion in subchapter 6.2.4.

6.2.3.5. The Moderating Effect of Country

The data presented in the previous subchapters showed how the respondents’ countries and usage habits influenced the respondents’ emphasis of the brand meanings. For the purpose of this study, in particular hypothesis 4, it was also necessary to determine whether differences in the emphasis of the brand meanings among different user types were constant across cultures. It was also important to identify meanings that were more emphasised among the Engaged users than the other users in each of the countries for each of the brands. This was important as such data provide useful guidelines on how to increase the number of the Engaged users. For this analysis, the Type of user is observed as an interval variable and renamed Affection towards a brand. For each brand this variable ranges from the lowest affection for the Non-users to the highest for the Engaged users.

181

The results in table 6-24 show that the all-countries overall sample displayed a statistically significant correlation between the meanings Refreshing (r = .29; p < .01), Energising (r = .18; p < .01) and Unhealthy (r = -.42; p < .01) on one side and Affection towards Coca-Cola on the other side. Therefore, this indicates that the emphasis of the meanings Refreshing and Energising is positively related towards the respondents’ affection towards Coca-Cola.

Observing each country separately, Austria displayed the most similar pattern to the overall sample with statistically significant correlations for Refreshing (r = .25; p < .01), Energising (r = .19; p < .05) and Unhealthy (r = -.18; p < .01). In addition, Globally recognised (r = .21; p < .05) was also statistically significantly correlated with Affection towards Coca-Cola in Austria, indicating that Austrians who had more affection for Coca-cola placed a higher emphasis on the Globally recognised meanings than other Austrians who did not.

Table 6-24. The Partial Correlations a for Affection towards Coca-Cola b

Brand Meaning

Overall (df = 358)

Austria (df =103)

Finland (df = 98)

India (df = 77)

United States (df = 56)

Refreshing .289 (.000)** .252 (.010)** .280 (.005)** .219 (.052) .254 (.054)Globally recognised

.056 (.289) .211 (.031)* -.140 (.169) .146 (.200) .046 (.730)

Unhealthy -.417 (.000)** -.179 (.000)** -.404 (.000)** -.330 (.003)** -.337 (.010)** Christmas c -.052 (.328) -.056 (.569) -.197 (.049)* .090 (.430) .045 (.738)Classic .007 (.896) -.006 (.954) .143 (.157) -.087 (.443) .180 (.175)Energising .184 (.000)** .192 (.049)* .127 (.208) .195 (.085) .413 (.001)**Bonding with friends c

.022 (.677) .011 (.912) -.055 (.958) .051 (.655) .215 (.105)

a Each brand meanings’ correlations to Affection towards Coca-Cola were controlled for the effects of the other six brand meanings listed in the first column of the table. b The values outside the brackets represent the correlation coefficients, whereas the values in the brackets the significance levels. c The negative values represent high emphasis of the meaning. * p < .05; ** p < .01

182

In Finland, Refreshing (r = .28; p < .01) and Unhealthy (r = -.40; p < .01) also showed the same pattern as in the overall sample. In addition, the correlation between Christmas and Affection towards Coca-Cola, (r = -.20; p < .05) was also positive and statistically significant in Finland. This indicated that the Christmas Coca-Cola meaning was more emphasised by the Finns who have more affection for the brand, and less by those who do not.

In India, only Unhealthy (r = -.33; p < .01) was statistically significantly correlated to Affection towards Coca-Cola, while in the United States Unhealthy (r = -.34; p < .01) and Energising (r = .42; p < .01) repeated the same pattern as in the all-countries overall sample. Finally, when the critical significance level was set just slightly less rigorously, Refreshing also became significantly positively correlated with Affection towards Coca-Cola in both India (r = .22; p = .05) and the United States (r = .25; p = .05).

Table 6-25 shows the results for Google. The all-countries overall sample displayed statistically significant correlations between the meanings Almighty (r = .27; p < .05), Intrusive (r = -.12; p < .05), Inspired (r = .26; p < .01), Fun (r = -.15; p < .01) and Informative (r = .13; p < .05) on one side, and Affection towards Google on the other side. This indicated that the meanings Almighty, Inspired, Fun, and Informative were more emphasised by those respondents who were more affected towards Google. On the other hand, the meaning Intrusive (r = -.12; p < .05) had a negative direction of the correlation, so it was more emphasised by those respondents who were less affected towards Google.

Observing each country separately, correlations occurred for far fewer meanings in each of the countries. Finland displayed the most similar pattern to the overall sample, with statistically significant correlations for

183

Almighty (r = .23; p < .05), Inspired (r = .23; p < .05) and Informative (r = .24; p < .05). In Austria, only the meanings Almighty (r = .25; p < .05) and Intrusive (r = -.22; p < .05) showed the same pattern as in the overall sample, while all the other meanings did not correlate significantly to Affection towards Google. In India, no meaning was statistically correlated to Affection towards Google. Finally, in the United States only the meaning Informative (r = .41; p < .01) had a statistically significant positive correlation to Affection towards Google.

Table 6-25. The Partial Correlations a for Affection towards Google b

Brand Meaning

Overall (df = 309)

Austria (df = 93)

Finland (df = 82)

India (df = 65)

United States (df = 45)

Almighty .272 (.000)** .251 (.014)* .227 (.038)* .130 (.296) .136 (.362)Helpful c -.029 (.616) -.095 (.360) .029 (.792) .010 (.934) -.114 (.446)Intrusive -.124 (.028)* -.219 (.033)* .017 (.881) .023 (.856) -.198 (.183)Inspired .263 (.000)** .179 (.083) .227 (.038)* .081 (.515) .007 (.965)Fun c -.146 (.010)** -.074 (.477) -.062 (.578) -.014 (.908) -.259 (.179)Global c -.105 (.065) -.093 (.369) -.213 (.052) -.159 (.200) -.069 (.644)Informative .126 (.026)* -.139 (.178) .244 (.025)* .029 (.818) .410 (.004)**a Each brand meanings’ correlations to Affection towards Google were controlled for the effects of the other six brand meanings listed in the first column of the table. b The values outside the brackets represent the correlation coefficients, whereas the values in the brackets the significance levels. c The negative values represent high emphasis of the meaning. * p < .05; ** p < .01 The results in table 6-26 show that almost all of Nike’s correlations were statistically significant. In particular, Achiever (r = .24; p < .01), Not weak (r = .24; p < .01), Not socially responsible (r = -.37; p < .01), and Sports (r = -.16; p < .01) on one side, and Affection towards Nike on the other side were all statistically significantly correlated. Such results indicated that Achiever, Not weak and Sports were more emphasised by the respondents the more they felt affection towards Nike Conversely, Not socially responsible was more emphasised by the respondents who felt less affection towards Nike.

184

Table 6-26. The Partial Correlations a for Affection towards Nike b

Brand Meaning

Overall (df = 300)

Austria (df = 95)

Finland (df = 84)

India (df = 60)

United States(df = 43)

Achiever .241 (.000)** .283 (.005)** .232 (.002)** .014 (.913) .066 (.665)Not weak .236 (.000)** .293 (.004)** .190 (.080) .054 (.678) .388 (.008)**Not socially responsible

-.371 (.000)** -.338 (.001)** -.434 (.000)** -.089 (.494) -.368 (.013)*

Pricey c .086 (.135) .049 (.636) -.072 (.510) .051 (.695) .149 (.329)Sports c -.164 (.004)** -.075 (.467) -.218 (.043)* -.246 (.053) -.279 (.064)a Each brand meanings’ correlations to Affection towards Nike were controlled for the effects of the other four brand meanings listed in the first column of the table. b The values outside the brackets represent the correlation coefficients, whereas the values in the brackets the significance levels. c The negative values represent high emphasis of the meaning. * p < .05; ** p < .01 The Meaning Achiever had the same relation to Affection towards Nike only in Austria (r = .28; p < .01) and Finland (r = .23; p < .01), whereas in the other countries it did not correlate to Affection towards Nike. Not weak was also statistically significantly correlated to Affection towards Nike in Austria (r = .29; p < .01) and in the United States (r = .39; p < .01). In three of the studied countries: Austria (r = -.34; p < .01), Finland (r = -.43; p < .01) and the United States (r = -.37; p < .01), Not socially responsible was statistically significantly correlated to Affection towards Nike.

The emphasis of Pricey was not statistically significantly correlated to Affection towards Nike in any of the countries. Finally, Sports was a meaning that was statistically significantly correlated to Affection towards Nike in Finland (r = -.22; p < .05) (and with a slightly more loosely conditioned significance level in India (r = -.25; p = .06) and the United States (r = -.28; p = .06)).

Overall, the results showed that correlations between a brand meaning and affection towards a brand are not the same across cultures. A more

185

detailed discussion on the findings needed for the evaluation of hypothesis 4 is provided in subchapter 6.2.4.4.

6.2.4. Discussion

This chapter evaluates the hypotheses based on the summarised findings provided in the previous subchapters.

6.2.4.1. Hypothesis 1

Most of the brand meanings differed across cultures not only in their emphasis, but also in the importance of their contribution to the overall brand meaning. Such results provided plenty of evidence to support Hypothesis 1, which claimed that: The meaning of certain brands differs for consumers in different cultures. These results are in accordance with previous research (cf. Gram 2007; Phau and Lau 2000; Berthon, Pitt, and Campbell 2008).

6.2.4.2. Hypothesis 2

Table 6-27 presents summarised evidence for Hypothesis 2, which predicted that Brand meaning is affected by cultural dimensions. Out of the thirteen H2 sub-hypotheses, which each predicted the effect of a different cultural dimension, this research provided evidence to test eight of these hypotheses. This is the case because of the nature of the research, which allowed testing of only those particular meanings (and hypotheses) that were identified by the interviewees as the meanings relevant for the three studied brands.

As table 6-27 shows, three sub-hypotheses are fully supported (namely, H2b - which predicted the effect of low in-group collectivism; H2e - which predicted the effect of high power distance; and H2k - which predicted the effect of low human orientation.) Five additional hypotheses

186

are partly supported (namely, H2a – which predicted the effect of high collectivism; H2c - which predicted the effect of high uncertainty avoidance; H2g – which predicted the effect of high assertiveness; H2h – which predicted the effect of low assertiveness; and H2j - which predicted the effect of high human orientation). Overall, this gives substantial evidence to support hypothesis 2.

6.2.4.3. Hypothesis 3

Almost all of the positive meanings were found to be statistically more emphasised by the Engaged users than by the others, and all the negative meanings to be more emphasised by the Non-users than by the others. In addition, those meanings that did not show such patterns were not opposite to what was expected, but rather neutral (i.e. those meanings showed no significant difference in the meaning emphasis between the Engaged users and the Non-users). All of these arguments give a solid basis for accepting hypothesis 3, which predicted that Brand users (especially engaged ones) will emphasise positive brand meanings more than non-users, whereas non-users will emphasise negative meanings more than brand users (especially engaged ones).

187

Table 6-27. The Overall Effects of the Cultural Dimensions on the Brand Meanings

Cultural dimension Coca-Cola H Google H Nike H Conclusion

Collectivism+ Bonding with friends

S c Global NS H2a - partially supported

Collectivism- Intrusive S H2b - supported

Uncertainty avoidance +

Globally recognised

NS Informative PS H2c – partially supported

Uncertainty avoiding -

H2d – n.a.

Power distance +

Almighty S H2e – supported

Power distance -

H2f – n.a.

Assertiveness+ Inspired

S

Achiever

Not weak

S

NS a

H2g – partly supported

Assertiveness- Unhealthy

Classic

S

PS

Helpful b NS Not socially responsible

PS

H2h – partly supported

Performance orientation +

H2i – n.a.

Human orientation+

Christmas S c Not socially responsible

PS H2j – partially supported

Human orientation -

Fun S H2k – supported

Future orientation +

H2l – n.a.

Gender egalitariani +

H2m – n.a.

Not affected by culture

Refreshing

Energising

S d

S

Sports

Pricey

S

S d

a Predicted to be affected by assertiveness, but no effect was captured. b Predicted to be affected by assertiveness and human orientation, but no effect was found. c The results are distorted for forces other than culture d No cultural effect predicted - statistically significantly higher results for India are ascribed to environmental forces other than culture. S = supported, PS= partly supported, NS = not supported, n.a. = not applicable

188

6.2.4.4. Hypothesis 4

Negative meanings showed consistency across all the countries in terms of their different emphasis by Users vs. Non-users. In particular, such a pattern is demonstrated by the Not socially responsible Nike meaning and the Unhealthy Coca-Cola meaning. Apart from the negative meanings, consistency across all the countries is also demonstrated for the functional meanings Refreshing and Informative.

Except for the afore-mentioned meanings, the overall results differed among the countries and hence provided substantial evidence for the acceptance of hypothesis 4, which predicted that Differences in brand meaning emphasis between different types of users are not equal in each country.

189

7. Conclusion

7.1. Summary of the Findings

The aim of this research was twofold. Firstly, to identify what brands mean to consumers and specify the brand meaning concept, and secondly to explore whether differences in brand meanings exist among consumers of different cultures and whether they can be explained by differences in cultural dimensions.

Investigation of past studies gave an insight into the existing concepts in the world of brands. Although, or maybe even because, there are numerous branding concepts (Keller 2003a), it remains unclear what brands actually mean to consumers. Previous research analysis also revealed several established paradigms that served as a firm basis for constructing a relationship model between culture and brand meaning: brand meaning is created in consumers’ minds (Brown, Kozinets, and Sherry 2003; De Chernatony 2001, p. 19; Vargo and Lusch 2004); it is based on the existing personal knowledge (Berthon, Pitt, and Campbell 2008) and value systems (Phau and Lau 2000); and knowledge and value systems are culturally determined (Hofstede 2001).

To address the first aim of the paper, the research was conducted in the spirit of scientific discovery (Snow and Thomas 1994, p. 465). The empirical investigation was entered without a priori defined research propositions. A series of in-depth interviews was conducted with consumers in the four studied countries in order to understand what some of the carefully chosen global brands mean to them. Transcript and visual data were coded in three rounds until coherent groups of meanings were

190

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.

191

7.2. Managerial Implications

This research has several managerial implications as detailed below.

7.2.1. Brand Meaning is a set of Cognitive and Affective Experience-Based Brand Associations

The results of this study show that a brand becomes more meaningful the more cognitive and affective experience-based associations it has. In other words, with experience more types of meanings become associated to a brand, which results in higher awareness and a higher recall rate in various consumption situations. Concretely, Coca-Cola takes the top position on the Interbrand and Business Week (2008) list as the brand with the highest brand value, Google the tenth position and Nike “only” the 29th. Coca-Cola and Google are both richer in terms of the various positive meanings that consumers attach to them. Moreover, when it comes to types of meanings, symbolic meanings are becoming increasingly important to consumers. Nowadays, many brands are primarily consumed for their symbolic meanings (Veloutsou and Moutinho 2008). The importance of symbolic meanings is also visible in the case of Coca-Cola. It has the highest brand value and the highest number of symbolic meanings among the studied brands.

Therefore, brand managers are advised to invest their creative energy in creating as many experiences for their consumers and potential consumers as possible so that consumers can expand their brand association networks, become more connected to the brand and use it more often with great passion. This will result in increased profits for the companies which own the brands in question.

192

7.2.2. Consumer Understanding of Brands is Culturally Determined

The results of the research show that even the most global brands have different meanings for consumers in different cultures. This is because of the culturally embedded interpretation processes. Practitioners should thus firstly seek information on the cultural specifics in the target countries. Secondly, they should decide whether they consider unique positioning beneficial, or not. In case they desire consistency of meanings across cultures, they should, contrary to expectations, employ country-adapted communication and even consider adapting other marketing elements.

In case practitioners do not consider cross-cultural brand meaning consistency crucial and beneficial, they should adapt their brand messages in a way that they evoke particular associations and behaviours on the side of consumers.

7.2.3. Negative Brand Meanings Relate to Brand Usage

The results show that the emphasis of negative meanings between users and non-users across brands and countries differs the most. The research does not test for cause and effect, but the results give an important indication that a negative meaning might be the reason for some consumers’ avoidance of a particular brand, to a greater extent than positive meanings might be the reason for opting for the brand. Therefore, managers are advised to pay particular attention to the negative aspects of their brand’s meaning and give priority to fighting such meanings.

Furthermore, both Coca-Cola and Nike received their negative meanings when they failed to recognise consumer trends. Coca-Cola has been building its favourable iconic position since 1886 almost solely on one sweet and energising beverage. While this formula was successful for a century, the trend changed in the 1990s when consumers became more

193

fitness and wellness-oriented. Isdel, Coca-Cola’s CEO, admits that Coca-Cola did not understand the significance of the trend at the time (Theater 2004).

Similarly, since its beginnings in 1962, Nike has never owned its own facility and still houses its production facilities with Asian subcontractors (Riesenbeck and Perrey 2007). Observing this issue from the perspective of brand management, it is evident that Nike has neglected a consumer trend as well. That is, for decades nobody questioned Nike’s integrity, and nobody even related fancy high-tech shoes to the (lack of) working and living conditions in Nike’s subcontractors’ production facilities. Only, in the 1990s, as the consumers grew increasingly aware and sensitive to global societal problems, the Nike sweatshops scandal broke out and hurt its image.

Therefore, from the point of view of brand management, managers are advised to systematically and diligently observe consumer trends to avoid negative meanings becoming attached to their brands. This advice is aimed at market challengers, but even more so at market leaders who are more prone to ignoring the trends while being comfortable ahead of their competitors.

7.3. Scientific Implications

7.3.1. Major Scientific Insights of the Research

This research contributes scientifically to the fields of brand management, cross-cultural studies and consumer research. Each of the contributions will be further elaborated below.

194

7.3.1.1. Contribution to Brand Management Field

By taking an inductive approach to conceptualising brand meaning, this research contributes to the field of brand management. After having thoroughly analysed the existing branding literature, the need for a new concept named brand meaning has been identified. This is despite the many existing models and views on branding, and in line with the fact that it is still not clear why consumers prefer certain brands and which meanings those brands have for them. Previous literature has provided evidence that brands become strong, successful and preferred by consumers because of their appealing personalities (Phau and Lau 2000; Riesenbeck and Perrey 2007, p. 151); good resonance with the target consumers (Pringle and Thompson 1999, p. I); personal relationships with consumers (Fournier 1998; Sweeney and Brandon 2006); or benefits for consumers (Park, Jaworski, and Maclnnis 1986; Kapferer 2004, p. 23).

Because these are all related, but rather different concepts, this research took a step away from the existing literature by turning directly to consumers and asking them for the answers. The inductive approach was considered superior to its deductive counterpart because research in brand management field has reached a mature stage. Focusing on theory testing alone at the current mature stage of brand management research might hamper further knowledge creation (Montgomery, Wernerfelt and Balakrishnan 1989).

Because the field research was entered theory-free, the direct contact with consumers made it easier to understand what brands mean to consumers and whether brand meaning is a concept most linked to benefits, personalities, images, associations, relationships, and the like. The interpretation of the results revealed that the concept of brand meaning has

195

some unique aspects and a value of its own even though it shares many aspects of the just mentioned concepts.

7.3.1.2. Contribution to the Field of Cross-cultural Research

This research is embedded in the cross-cultural environment. Its contribution to the field is related to observing cultural dimensions as a priori predictors, which is an often neglected aspect of the research. A more common approach in cross-cultural studies is either identifying the studied phenomena’s differences among countries (e.g. in the studies by Costa and Pavia 1992 and Gram 2007), or at best using cultural dimensions as a posteriori constructs to explain the findings (e.g. in the studies by Foscht et al. 2008 and Pankhania, Lee, and Hooley 2007).

This research showed that cultural dimensions can and should be used as a priori predictors of certain consumer behaviours. Only when used a priori can cultural dimensions really serve the purpose they were defined for in the first place, and that is to help managers best meet the specifics of particular local markets.

7.3.1.3. Contribution to Methodology in Consumer Research

Because this research was conducted on three continents, without modern technologies its completion would have been much more complex, timely and costly. Although it has by now already become an established practice to conduct online surveys, online interviews conducted in written form have not been widely used among researchers. This research proved that for computer-literate respondents, online communication is not only just as good as face-to-face communication, but in some respects even superior to it.

The collected responses were compared to those of in-person conducted pilot interviews and no difference was discovered in either the depth or

196

breadth of the responses. As the results of this research proved, the feelings and moods of the respondents can also be expressed with the mood icons integrated into most of the online messengers, and potentially even better stored in the transcript. In addition, this mode of research allows the respondents to stay totally anonymous and remain in the comfort of their homes, with the possibility of even stopping the interview and resuming it at a later stage. All of these things contributed greatly to the validity of the responses. In addition, responses collected in written form are much easier to process, which leaves less space for errors in interpretation and coding.

Furthermore, although qualitative research usually serves to gain an initial understanding of the studied phenomenon, and quantitative research allows researchers to collect solid data to confirm hypotheses, this two-stage research discovered that the in-depth interviews already yielded results that greatly resembled those later obtained in the quantitative stage. Hence this research provided further evidence to support Coulter, Zaltman, and Coulter (2001) who claim that the ZMET method (that served as a basis for this research) can provide up to 90% of the information available from a larger set of classical interviews in just four to five in-depth interviews.

7.3.2. Limitations of the Research

There were several limitations to this study. As for the qualitative stage, the study was adjusted to an online setting. Although the Internet environment provided a number of advantages over the face-to-face interview setting, some of the association elicitation techniques such as image grouping and mind mapping could not have been conducted. Such techniques might have provided some additional relevant data on brand meanings.

197

Furthermore, interview transcripts were analysed by applying the hermeneutic approach. Such approach has an inherited problem of interpretation. In other words, any interpretation is heavily dependent on interpreter’s “personal history and cultural background” (Thompson, Pollio, and Locander 1994). Hence, although the coding of the responses was done with great care, the validity of the procedure could have been increased by employing multiple, preferably cross-cultural coders. Since the coded results were grouped and the groups of meanings were tested for reliability in the quantitative stage, this problem was diminished. The concern remains as to whether all the relevant meanings were recognised in the qualitative stage and taken into the quantitative.

When it comes to quantitative phase of the research, the study was conducted in English. Although, all the procedures for potential problem reduction were applied, the question remains whether conducting a survey in the respondents’ native languages would have yielded somewhat different results. Furthermore, although the overall sample was substantial, when divided into groups during the two-way ANOVA tests, some of the groups were relatively small. Thus, it would have been useful to collect more data. Finally, although cross-cultural research rarely investigates more than three cases in four cultures, expanding the research to include more cases and more cultures is a generally valid piece of advice.

7.3.3. Future Research

Firstly, it would be interesting to address the limitations of the current research. Expanding the study to more cultures and brands, as well as addressing the language issue of the research, (i.e conducting the study in the respondents’ native languages, and coding the responses with a team of multicultural researchers) would be beneficial.

198

In addition, the results call for further investigation of the relationship between brand meaning on one side and brand preference and loyalty on the other. More specifically, although this research revealed some initial understandings of which types of meanings are differentiated between Engaged users, Disengaged users and Non-users, the conclusions of this research would be strengthened with the results obtained in a study designed in a cause and effect mode. Furthermore, it would also be interesting to investigate how some associations in consumer minds evolve to represent meanings; and what, apart from emphasised consumer needs, makes certain meanings more relevant to consumer than others.

From a cross-cultural perspective, it would also be interesting to study whether consumers in different cultures form brand meaning groups differently. The results of the present study provide evidence of a rather consistent pattern in that matter. However, focusing on this issue in particular could provide even stronger conclusions. In addition, it would be beneficial to study several brands of the same product category to see whether the brand meanings of the preferred brand resemble the culturally determined value systems of the respondents more than the non-preferred brands.

Finally, during the research the question of whether the world is becoming globalised, Westernised or Americanised arose. Although such a topic would require a multidisciplinary perspective, from the point of view of brand management, it would be interesting to firstly investigate whether brands originating from the United States really convey American values and spread them around the world, or is it that they are truly global, in the sense that such brands drain different aspects of cultures from all parts of the world and actually represent global values. Additionally, it would also be interesting to analyse which role brands play in creating popular culture locally and whether their role is dominant in comparison to some other

199

forms of multicultural exposure, like media, art, literature, personal experiences and the like.

200

References

Aaker, David A. (1991), Managing Brand Equity: Capitalizing on the Value of a Brand Name. New York, NY: The Free Press.

------- (1994), “Building a Brand: The Saturn Story,” California Management Review, 36 (2), 114–33.

-------, David A. (1996), Building Strong Brands. New York: The FreePress.

-------, David A. and Erich Joachimsthaler (2002), Brand Leadership. London: Simon & Schuster.

-------, David A., V. Kumar, and George S. Day (2001), Marketing Research (7th ed.). New York: John Wiley & Sons, Inc.

Aaker, J.L. (1997), “Dimensions of Brand Personality,” Journal of Marketing Research, 34 (3), 347-56.

-------, Jennifer L. (2000), “Accessibility or Diagnosticity? Disentangling the Influence of Culture on Persuasion Processes and Attitude,” Journal of Consumer Research, 28 (4), 340-57.

Abou Aish, Ehab M., Christine T. Ennew, and Sally A. McKechnie (2003), “A Cross-Cultural Perspective on the Role of Branding in Financial Services: The Small Business Market,” Journal of Marketing Management, 19 (9/10), 1021-42.

Achenreiner, Gwen Bachmann and Deborah Roedder John (2003), “The Meaning of Brand Names to Children: A Developmental Investigation,” Journal of Consumer Psychology, 13 (3), 205-19.

201

AdAge (2008) “Top 10 Slogans of The Century,” (accessed 20th December 2008) [available at: http://adage.com/century/slogans.html].

Adigüzel, Feray and Michel Wedel (2008), “Split Questionnaire Design for Massive Surveys,” Journal of Marketing Research, 45 (5), 608-17.

Albert Noël, Dwight Merunka, and Pierre Valette-Florence, (2008), “When Consumers Love Their Brands: Exploring the Concept and its Dimensions,” Journal of Business Research, 61 (10) 1062–75.

Alden, Dana L., Jean-Benedict E. M. Steenkamp, and Rajeev Batra (1999), “Brand Positioning through Advertising in Asia, North America, and Europe: The Role of Global Consumer Culture,” Journal of Marketing, 63 (1), 75–87.

Allen, I. Elaine and Christopher A. Seaman (2007), “Likert Scales and Data Analyses,” Quality Progress, 40 (7), 64-65.

Ambler, Tim and Chris Styles (1997), “Brand Development versus New Product Development: Toward a Process Model of Extension Decisions,” Journal of Product & Brand Management, 6 (4), 222-34.

Apelbaum, Eidan, Eitan Gerstner, and Prasad Naik (2003), “The Effects of Expert Quality Evaluations versus Brand Name on Price Premiums,” Journal of Product & Brand Management, 12 (3), 154-65.

Aperia, Tony and Rolf Back (2004), Brand Relations Management. Copenhagen: Liber, Copenhagen Business School Press, Abstrakt.

Azoulay Audrey and Jean-Noel Kapferer (2003), “Do Brand Personality Scales Really Measure Brand Personality,” Journal of Brand Management, 11 (2), 143-55.

202

Banerjee Saikat (2008), “Dimensions of Indian Culture, Core Cultural Values and Marketing Implications: An Analysis,” Cross Cultural Management: An International Journal, 15 (4), 367-78.

Barr, Pamela S. and Mary Ann Glynn (2004), “Cultural Variations in Strategic Issue Interpretation: Relating Cultural Uncertainty Avoidance to Controllability in Discriminating Threat and Opportunity,” Strategic Management Journal, 25 (1), 59-67.

Baumgartner, Hans, Mita Sujan, and James R. Bettman (1992), “Autobiographical Memories, Affect, and Consumer Information Processing,” Journal of Consumer Psychology, 1 (1), 53-82.

Berthon, Pierre, Leyland F. Pitt, and Colin Campbell (2008), “Does Brand Meaning Exist in Similarity or Singularity?,” Journal of Business Research, In Press, Available online 3 August 2008

(doi:10.1016/j.jbusres.2008.05.015)

Bidney, David (1954), “Reviewed Work(s): Culture: A Critical Review of Concepts and Definitions by A. L. Kroeber; Clyde Kluckhohn,” The American Journal of Sociology, 59 (5), 488-89.

Bjerke, Rune and Rosemary Polegato (2006), “How Well Do Advertising Images of Health and Beauty Travel across Cultures? A Self-concept Perspective,” Psychology & Marketing, 23 (10), 865-84.

Black, Thomas R (1999), Doing Quantitative Research in The Social Sciences: An Integrated Approach to Research Design, Measurement and Statistics. London: Sage.

Blythe, Jim (2007), “Advertising Creatives and Brand Personality: A Grounded Theory Perspective,” Journal of Brand Management, 14 (4), 284-94.

203

Bodkin, Charles D., Christie Amato, and Cara Peters (2008), “The Role of Conflict, Culture, and Myth in Creating Attitudinal Commitment,” Journal of Business Research, In Press, Available online 20 June 2008 (doi:10.1016/j.jbusres.2008.05.005)

Brannen Mary Yoko, Carolina Gómez, Mark F. Peterson, Laurence Romani, Lilach Sagiv and Pei-Chuan Wu (2004), “People in Global Organizations: Culture, Personality, and Social Dynamics,” in The Blackwell Handbook of Global Management, Henry W. Lane, Martha L. Maznevski, Mark E. Mendenhall and Jeanne Mcnett, eds. Blackwell Publishing, 26 - 54.

Broderick, Amanda J. (2007), “A Cross-national Study of the Individual and National–cultural Nomological Network of Consumer Involvement,” Psychology & Marketing, 24 (4), 343-74.

Broniarczyk, Susan M. and Joseph W. Alba (1994), “The Importance of the Brands in Brand Extension,” Journal of Marketing Research, 31 (2), 214-29.

Brown, Stephen P., Robert V. Kozinets, John F. Sherry Jr. (2003), “Teaching Old Brands New Tricks: Retro Branding and the Revival of Brand Meaning,” Journal of Marketing, 67 (3), 19-33.

-------, Steven P. and Douglas M. Stayman (1992), “Antecedents and Consequences of Attitude toward the Ad: A Meta-analysis,” Journal of Consumer Research, 19 (June), 34-51.

Brown, Tom J., Peter A. Dacin, Michael G. Pratt, and David A. Whetten (2006), “Identity, Intended Image, Construed Image, and Reputation: An Interdisciplinary Framework and Suggested Terminology,” Journal of the Academy of Marketing Science, 34 (2), 99-106.

204

Bulmer, Ralph (1953), “Review: 211 Reviewed Work(s): Culture: A Critical Review of Concepts and Definitions by A. L. Kroeber; Clyde Kluckhohn,” Man, 53, 136-37.

Bulmer, Sandy and Margo Buchanan-Oliver (2006), “Advertising across Cultures: Interpretations of Visually Complex Advertising,” Journal of Current Issues & Research in Advertising, 28 (1), 57-71.

Burnkrant, Robert E. and H. Rao. Unnava (1995), “Effects of Self-Referencing on Persuasion,” Journal of Consumer Research, 22 (1), 17-26.

Calder, Bobby J. and Steven J. Reagan (2001), “Brand Design,” in Kellogg on Marketing, Dawn Iacobucci, ed. New York: John Wiley, 58-73.

Carlson, Brad D., Tracy A. Suter, and Tom J. Brown (2008), “Social versus Psychological Brand Community: The Role of Psychological Sense of Brand Community,” Journal of Business Research, 61 (4), 284-91.

Cattell, Raymond. B. (1966), “The scree test for the number of factors,” Multivariate Behavioral Research, 1 (2), 245-76.

Cavusgil, S. Tamer and Ajay Das (1997), “Methodological Issues in Empirical Cross-cultural Research: A Survey of the Management Literature and a Framework,” Management International Review, 37 (1), 71-96.

Chang, Pao-Long and Ming-Hua Chieng (2006), “Building Consumer–brand Relationship: A Cross-Cultural Experiential View,” Psychology & Marketing, 23 (11), 927-59.

205

Cho, Bongjin, Up Kwon, James W. Gentry, Sunkyu Jun; Fredric Kropp, (1999), “Cultural Values Reflected in Theme and Execution: A Comparative Study of U.S. and Korean Television Commercials,” Journal of Advertising, 28 (4), 59-73.

Christodoulides, George (2008), “Breaking free from the industrial age paradigm of branding,” Journal of Brand Management, 15 (4), 291-93.

Corain, Livio and Luigi Salmaso (2007), “A Critical Review and a Comparative Study on Conditional Permutation Tests for Two-Way ANOVA,” Communications in Statistics: Simulation & Computation, 36 (4), 791-805.

Costa, Janeen Arnold and Teresa M. Pavia (1992), “What it all adds up to: Culture and alpha-numeric brand names,” Advances in Consumer Research, 19 (1), 39-45.

Costello, Anna B. and Jason Osborne (2005). “Best Practices in Exploratory Factor Analysis: Four Recommendations for Getting the Most from Your Analysis,” Practical Assessment Research & Evaluation, 10 (7). [Available online: http://pareonline.net/getvn.asp?v=10&n=7].

Coulter, Robin Higie and Gerald Zaltman (1994), “Using The Zaltman Metaphor Elicitation Technique to Understand Brand Images” Advances in Consumer Research, 21 (1), 501-07.

-------, Robin Higie, Gerald Zaltman, and Keith S. Coulter (2001), “Interpreting Consumer Perceptions of Advertising: An Application of the Zaltman Metaphor Elicitation Technique,” Journal of Advertising, 30 (4), 1-21.

206

Cova, Bernard and Stefano Pace (2006), “Brand Community of Convenience Products: New Forms of Customer Empowerment - The Case “My Nutella the Community”,” European Journal of Marketing, 40 (9/10), 1087-2005.

Craig, C. Samuel and Susan P. Douglas (2000), International Marketing Research. New York: John Wiley and Sons.

Czerniawski, Richard D. and Michael W. Maloney (1999), Creating Brand Loyalty: The Management of Power Positioning and Really Great Advertising. New York: Amacom.

Davis, Teresa (2007), “Brand Meaning and Children: A Thematic Categorisation Task,” Journal of Brand Management, 14 (3), 255-66.

De Chernatony, Lesley (2001), From Brand Vision to Brand Evaluation. Oxford: Butterworth-Heinemann Ltd.

De Matos, Celso and Carlos Rossi (2008), “Word-of-mouth Communications in Marketing: A Meta-analytic Review of the Antecedents and Moderators,” Journal of the Academy of Marketing Science, 36 (4), 578-96.

Denzin, Norman K. and Yvonna S. Lincoln (1994), Handbook of Qualitative Research. Thousand Oaks & London: Sage.

Dillon, William R., Thomas J. Madden, Amna Kirmani, and Soumen Mukherjee, (2001), “Understanding What's in a Brand Rating: A Model for Assessing Brand and Attribute Effects and Their Relationship to Brand Equity,” Journal of Marketing Research, 38 (4), 415-29.

Dilts, Rober and DeLozier Judith (2000), Encyclopedia of Systemic NLP and NLP New Coding. Scotts Valley: NLP University Press.

207

Dittrich, Regina, Brian Francis, Reinhold Hatzinger, and Walter Katzenbeisser (2007), “A Paired Comparison Approach for the Analysis of Sets of Likert-scale Responses,” Statistical Modelling: An International Journal, 7 (1), 3-28.

Dolnicar, Sara and John R. Rossiter (2008), “The Low Stability of Brand-Attribute Associations is Partly Due to Market Research Methodology,” International Journal of Research in Marketing 25 (2), 104–08.

Edwards, Jeffrey. R. and Robert P. Bagozzi (2000), “On the Nature and Direction of Relationships Between Constructs and Measures,” Psychological Methods, 5 (2), 155–74.

Eisenhardt, Kathleen M. (1989), “Building Theories from Case Study Research,” Academy of Management Review, 14 (4), 532-32.

Erdem, Tuelin, Joffre Swait, and Ana Valenzuela (2006), “Brands as Signals: A Cross-Country Validation Study,” Journal of Marketing, 70 (1), 34-49.

Escalas, Jennifer Edson and James R. Bettman (2005), “Self-Construal, Reference Groups, and Brand Meaning,” Journal of Consumer Research, 32 (3), 378-89.

Fabrigar, Leandre. R., Duane. T. Wegener, Robert C. MacCallum, and Erin. J. Strahan (1999), “Evaluating the Use of Exploratory Factor Analysis in Psychological Research,” Psychological Methods, 4 (3), 272-99.

Feldwick, Paul (2002), What is Brand Equity Anyway?. Oxfordshire, UK: World Advertising Research Center.

208

Fine, Seymour H (1981), The Marketing of Ideas and Social Issues. New York: Praeger.

Fishbein, Martin and Susan Middlestadt (1995), “Noncognitive Effects on Attitude Formation and Change: Fact or Artifact?,” Journal of Consumer Psychology, 4 (2), 181-202.

Foscht, Thomas, Cesar Maloles III, Bernhard Swoboda, Dirk Morschett, and Indrajit Sinha (2008), “The Impact of Culture on Brand Perceptions: A Six-nation Study,” Journal of Product & Brand Management, 17 (3), 131-42.

Fournier, Susan (1998), “Consumers and Their Brands: Developing Relationship Theory in Consumer Research,” Journal of Consumer Research, 24 (4), 343-73.

Franzen, Giep and Margot Bouwman (2001), The Mental World of Brands: Mind, Memory and Brand Success. Oxfordshire: World Advertising Research Center.

Frith, Katherine T. And Barbara Mueller (2007), Advertising and Societies: Global Issues (3rd ed.). New York: Peter Lang, Digital Formations Vol. 14.

Furrer, Olivier, Ben Shaw-Ching Liu, and D. Sudharsan (2000), “The Relationships between Culture and Service Quality Perceptions: Basis for Cross-Cultural Market Segmentation and Resource Allocation,” Journal of Service Research, 2 (4), 355-71.

Gaál, Zoltán, Lajos Szabó and Zoltán Kovács (2007), “Culture, Competence, Competitiveness: Managing Diversity at Individual and Community Level,” International Journal of Diversity in Organisations, Communities & Nations, 7 (5), 131-41.

209

Gaines, Stanley O., William D. Marelich, Katrina L. Bledsoe, W. Neil Steers, Michael C. Henderson, and Cherlyn S. Granrose, Lisa Barajas, Diana Hicks, Michael Lyde, Yumi Takahashi, Nancy Yum, Diana I Rios, Ben R Garcia, Karlyn R. Farris, and Mary S. Page (1997), “Links between Race/Ethnicity and Cultural Values as Mediated by Racial/Ethnic Identity and Moderated by Gender,” Journal of Personality and Social Psychology, 72 (6), 1460-76.

Goldsmith, Ronald E. (2004), “Current and Future Trends in Marketing and their Implications for The Discipline,” Journal of Marketing Theory & Practice, 12 (4), 10-17.

Gram, Malene (2007), “Whiteness and Western Values in Global Advertisements: An Exploratory Study,” Journal of Marketing Communications, 13 (4), 291-309.

Gregory, Anne (2007), “Involving Stakeholders in Developing Corporate Brands: The Communication Dimension,” Journal of Marketing Management, 23 (1/2), 59-73.

Grier, Sonya and Anne M. Brumbaugh (1999), “Noticing Cultural Differences: Ad Meanings by Target and Non-target Markets,” Journal of Advertising, 28 (1), 79-99.

Griffin, Jill (1997), Customer Loyalty – How to Earn it, How to Keep it. San Francisco: Jossey-Bass Publishers.

Haig, Matt (2003), Brand Failures. London: Kogan Page.

Hall, Edward T. (1989), Beyond culture [Reprint]. New York: Anchor Books.

Hampden-Turner, Charles M. and Fons Trompenaars (1996), “A World Turned Upside Down: Doing Business in Asia,” in Managing

210

Across Cultures: Issues and Perspectives, Pat Joynt and Malcom Warner eds. London: International Thomson Business Press, 275–305.

Hankinson, Graham and Philippa Cowking (1996), The Reality of Global Brands: Cases and Strategies for Successful Management of International Brands. Columbus: McGraw-Hill.

Hatch, Mary Jo and James Rubin (2006), “The Hermeneutics of Branding,” Journal of Brand Management, 14 (1/2), 40-59.

Hauser, John R. and Frank S. Koppelman (1979), “Alternative Perceptual Mapping Techniques: Relative Accuracy and Usefulness,” Journal of Marketing Research, 16 (4), 495-506.

Heath, Robert, David Brandt, and Agnes Nairn (2006), “Brand Relationships: Strengthened by Emotion, Weakened by Attention” Journal of Advertising Research, 46 (4), 410-19.

Henderson, Geraldine R., Dawn Iacobucci, and Bobby J. Calder (1998), “Brand Diagnostics: Mapping Branding Effects Using Consumer Associative Networks,” European Journal of Operational Research, 111 (2), 306-27.

-------, Geraldine R., Dawn Iacobucci, and Bobby J. Calder (2002), “Using Network Analysis to Understand Brands,” Advances in Consumer Research, 29 (1), 397-405.

Henderson, Pamela W., Joseph A. Cote, Siew Meng Leong, Bernd Schmitt (2003), “Building Strong Brands in Asia: Selecting the Visual Components of Image to Maximize Brand Strength,” International Journal of Research in Marketing, 20 (4), 297-313.

211

Henry, Jim (2007), “Company's Quality Crusade Was Launched by an American,” Automotive News, 82, (2007, October 30) 58.

Herzberg, Frederick (1968), “One More Time: How Do You Motivate Employees?,” Harvard Business Review, 46 (1), 53-62.

Herzog, Regula A. and Jerald G. Bachman (1981), “Effects of Questionnaire Length on Response Quality,” Public Opinion Quarterly, 45 (4), 489–504.

Hickman, Craig and Christopher Raia (2002), “Incubating Innovation,” Journal of Business Strategy, 23 (3), 14-18.

Hofstede Geert (1997), “Cultural constraints in management theories” in Leadership: Understanding the dynamics of power and influence in organizations, Robert P. Vecchio, ed. Notre Dame, IN: University of Notre Dame Press, 465 – 83.

-------, Geert (1991), Cultures and Organization: Software of the Mind, London: McGraw-Hill.

-------, Geert (1980), “Motivation, Leadership, and Organization: Do American Theories Apply Abroad?,” Organizational Dynamics, 9 (1), 42-63.

-------, Geert (2001), Culture's Consequences: Comparing Values, Behaviors, Institutions, and Organizations across Nations (2nd ed.). Thousand Oaks, CA: Sage.

-------, Geert and Michael Harris Bond (1988), “The Confucius Connection: From Cultural Roots to Economic Growth,” Organizational Dynamics, 16 (4), 5-21.

212

Hollenbeck, Candice R., Cara Peters, and George M. Zinkhan (2008), “Retail Spectacles and Brand Meaning: Insights from a Brand Museum Case Study,” Journal of Retailing, 84 (3), 334–53.

Holt, Douglas B. (1995), “How Consumers Consume: A Typology of Consumption Practices,” Journal of Consumer Research, 22 (1), 1-16.

-------, Douglas B. (2002), “Why Do Brands Cause Trouble? A Dialectical Theory of Consumer Culture and Branding,” Journal of Consumer Research, 29 (1), 70-90.

-------, Douglas B. (2004), How Brands Become Icons: The Principles of Cultural Branding. Boston, MA: Harvard Business School Press.

Homack Susan R. (2001), “Understanding What ANOVA Post Hoc Tests Are, Really,” 1st February 2001, (ERIC Document Reproduction Service No. ED449222)

Homer, Pamela M. and Sun-Gil Yoon (1992), “Message framing and the

interrelationships among ad-based feelings, affect, and cognition,” Journal of

Advertising, 21 (1), 19-33.

House, Rober, Mansour Javidan, Paul Hanges, and Peter Dorfman (2002), “Understanding cultures and implicit leadership theories across the globe: an introduction to project GLOBE,” Journal of World Business, 39 (1), 3-10.

-------, Robert, Paul Hanges, Mansour Javidan, Peter Dorfman and Vipin Gupta (2004), Culture, leadership and organizations: The GLOBE Study of 62 Societies. Thousand Oaks, London, New Delhi: SAGE Publications.

213

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&type=desc&col=7&langid=1000].

Jamieson, Susan (2004), “Likert scales: How to (ab)use them,” Medical Education, 38 (12), 1217-18.

Jun, Jong Woo and Hyung-Seok Lee (2007), “Cultural Differences in Brand Designs and Tagline Appeals,” International Marketing Review, 24 (4), 474-91.

-------, Jong Woo, Chang-Hoan Cho, and Hyuck Joon Kwon (2008), “The Role of Affect and Cognition in Consumer Evaluations of Corporate Visual Identity: Perspectives from the United States and Korea,” Journal of Brand Management, 15 (6), 382-98.

Kapferer, Jean-Noel (2004), The New Strategic Brand Management (3rd ed.). London: Kogan Page.

Kates, Steven M. and Charlene Goh (2003), “Brand Morphing: Implications for Advertising Theory and Practice,” Journal of Advertising, 32 (1), 59-68.

Kay, Mark J. (2006), “Strong Brands and Corporate Brands,” European Journal of Marketing, 40 (7/8), 742-60.

Kaynak, Erdener, Gulberk Gultekin Salman, and Ekrem Tatoglu (2008), “An integrative framework linking brand associations and Brand Loyalty in Professional Sports,” Journal of Brand Management, 15 (5), 336-57.

Keller, Kevin L. (2008), Best Practice Cases in Branding: Lessons from the World’s Strongest Brands. Zagreb: Accent.

214

-------, Kevin L. (2003a), “Brand Synthesis: The Multidimensionality of Brand Knowledge,” Journal of Consumer Research, 29 (4), 595-600.

-------, Kevin Lane and Donald R Lehmann (2006), “Brands and Branding: Research Findings and Future Priorities,” Marketing Science, 25 (6), 740-59.

-------, Kevin L. (1993), “The Effects of Corporate Branding Strategies on Brand Equity,” Advances in Consumer Research, 20 (1), 27-27.

-------, Kevin L. (2003b), Strategic Brand Management: Building, Measuring and Managing Brand Equity (2nd ed.). Upper Saddle River, NJ: Pearson Education.

Kempf, Deanne S and Robert E. Smith (1998), “Consumer Processing of Product Trial and the Influence of Prior Advertising: A Structural Modeling Approach,” Journal of Marketing Research, 35 (3), 325-38.

Khalid, Halimahtun M. and Martin G. Helander (2004), “A Framework for Affective Customer Needs in Product Design,” Theoretical Issues in Ergonomics Science, 5 (1), 27-42.

Kim, Jooyoung, Jon D. Morris, and Joffre Swait (2008), “Antecedents of True Brand Loyalty,” Journal of Advertising, 37 (2), 99–117.

Kim, W. Chan and Renee Mauborgne (2000), “Knowing a Winning Business Idea When You See One,” Harvard Business Review, 78 (5), 129-138.

Kinra, Neelam (2006), “The Effect of Country-of-origin on Foreign Brand Names in the Indian Market,” Marketing Intelligence & Planning, 24 (1), 15-30.

Klein, Naomi (2000), No Logo: No Space, No Choice, No Jobs. London: Flamingo.

215

Kluckhohn, Clyde (1951), “The Study of Culture,” in The Policy Sciences, Daniel Lerner and Harold D. Lasswell, eds. Stanford, CA: Stanford University Press, 86-101.

Kluckhohn, R. Florence and Fred L. Strodtbeck (1961), Culture: A Critical Review of Concepts and Definitions. Cambridge, MA: Harvard University.

Krishnan, H.S. (1996), “Characteristics of Memory Associations: A Consumer-Based Brand Equity Perspective,” International Journal of Research in Marketing, 13 (October), 389-05.

Kumar, Archana, Hyun-Joo Lee, and Youn-Kyung Kim (2008), “Indian Consumers' Purchase Intention toward a United States versus Local Brand,” Journal of Business Research, In Press, Available online 24 September 2008 (doi:10.1016/j.jbusres.2008.06.018)

Lange, Fredrik and Micael Dahlen (2003), “Let's Be Strange: Brand Familiarity and Ad-Brand Incongruency,” Journal of Product & Brand Management, 12 (7), 449-461.

Langley, Ann (1999), “Strategies for Theorizing from Process Data,” Academy of Management Review, 24 (4), 691-710.

LeBel, Jordan L. and Nathalie Cooke (2008), “Branded Food Spokescharacters: Consumers' Contributions to the Narrative of Commerce,” Journal of Product & Brand Management, 2008, 17 (3), 143-53.

Lee, Eun-Jung and Eun-Young Rhee (2008), “Conceptual Framework of Within-Category Brand Personality Based on Consumers’ Perception (WCBP-CP): The Case of Men’s Apparel Category in South Korea,” Journal of Brand Management, 15 (6), 465-89.

216

Lee, Min-Young, Dee Knight, and Youn-Kyung Kim (2008), “Brand Analysis of a US Global Brand in Comparison with Domestic Brands in Mexico, Korea, and Japan,” Journal of Product & Brand Management, 17 (3), 163-74.

Leo, Cheryl, Rebekah Bennett, and Charmine E. J. Härtel (2005), “Cross-Cultural Differences in Consumer Decision-Making Styles,” An International Journal of Cross Cultural Management, 12 (13), 32-62.

LePla, Joseph F. and Lynn M. Parker (2002), Integrated branding: Becoming Brand-driven through Company-wide Action. London: Kogan Page.

Levitt, Theodor (1983), “The Globalization of Markets,” Harvard Business Review, 61 (3), 92-101.

Ligas, Mark and June Cotte (1999), “The Process of Negotiating Brand Meaning: A Symbolic Interactionist Perspective,” in Advances in Consumer Research, Vol. 26, no 1, Eric J. Arnould, and Linda M. Scott, eds. Provo, UT: Association for Consumer Research, 609-14.

Lim, Kenny and O. Cass Aron (2001), “Consumer Brand Classifications: An Assessment of Culture-of-Origin versus Country-of-Origin,” Journal of Product & Brand Management, 10 (2), 120-36.

Lindstrom, Martin (2005), Build Powerful Brands through Touch, Taste Smell, Sight and Sound. New York: Free Press.

Low, George S. and Charles W. Lamb Jr. (2000), “The Measurement and Dimensionality of Brand Associations,” Journal of Product & Brand Management, 9 (6), 350-68.

217

Luo, Xueming, K. Sivakumar and Sandra Liu (2005), “Globalization, Marketing Resources, and Performance: Evidence from China,” Journal of the Academy of Marketing Science, 33 (1), 50-65.

Martin, Ingrid, David Stewart, and Shashi Matta (2005), “Branding Strategies, Marketing Communication, and Perceived Brand Meaning: The Transfer of purposive, goal-oriented brand meaning to Brand Extensions,” Journal of the Academy of Marketing Science, 33 (3), 275-94.

Maslow, Abraham H. (1943) “A Theory of Human Motivation,” Psychological Review, 50, 370-96.

Maznevski, Martha L, Joseph J. DiStefano, Gomez, Carolina B., Noorderhaven, Niels G, and Pei-Chuan Wu (2002), “Cultural Dimensions at the Individual Level of Analysis,” International Journal of Cross Cultural Management, 2 (3), 275-95.

McAlexander, James H., John W. Schouten, and Harold F. Koenig (2002), “Building Brand Community,” Journal of Marketing, 66 (1), 38-54.

McClelland, David C. (1961), The Achieving Society. Princeton, NJ: Van Nostrand.

McCracken, Grant (1986) “Culture and Consumption: A Theoretical Account of the Structure and Movement of the Cultural Meaning of Consumer Goods,” Journal f Consumer Research, 13 (June), 71-84.

Miliopoulou, Zozeta (2007), “Brand as A Symbolic Token: A Theoretical Framework for the Consumer-Brand Relationship,” paper presented at 2007 Thought Leaders International Conference on Brand Management, Birmingham Business School, University of Birmingham (April 24-25).

218

Monga, Alokparna Basu and Deborah Roedder John (2007), “Cultural Differences in Brand Extension Evaluation: The Influence of Analytic versus Holistic Thinking,” Journal of Consumer Research, 33 (4) 529-36.

Montgomery, Cynthia A., Birger Wernerfelt, and Srinivasan Balakrishnan (1989), “Strategy Content and The Research Process: A Critique and Commentary,” Strategic Management Journal, 10 (2), 189-97.

Moore, David J. and Pamela Miles Homer (2008), “Self-brand Connections: The Role of Attitude Strength and Autobiographical Memory Primes,” Journal of Business Research, 61 (7), 707-14.

Morgan, Susan E. and Tom Reichert (1999), “The Message Is in the Metaphor: Assessing the Comprehension of Metaphors in Advertisements,” Journal of Advertising, 28 (4), 1-12.

Moriarty, Sandra E. and Thomas R. Duncan (1990), “Global Advertising: Issues and Practices,” Current Issues & Research in Advertising, 13 (2), 313-13.

Muniz Jr. Albert M. and Thomas C. O'Guinn (2001), “Brand Community,” Journal of Consumer Research, 27 (4), 412-12.

-------, Jr. Albert M. and Hope Jensen Schau (2005), “Religiosity in the Abandoned Apple Newton Brand Community,” Journal of Consumer Research, 31 (4), 737-47.

Nairn Agnes, Christine Griffin, and Patricia Gaya Wicks (2008), “Children’s Use of Brand Symbolism A Consumer Culture Theory Approach,” European Journal of Marketing, 42 (5/6), 627-40.

219

Nisbett, Richard E., Kaiping Peng, Incheol Choi, and Ara Norenzayan (2001), “Culture and Systems of Thought: Holistic versus Analytic Cognition,” Psychological Review, 108 (2), 291-310.

Nunnally, Jum C. (1978), Psychometric Theory. New York: McGraw-Hill Book Company.

Oakenfull, Gillian, Edward Blair, Betsy Gelb, and Peter Dacin (2000), “Measuring Brand Meaning,” Journal of Advertising Research, 40 (5), 43-53.

O'Reilly, Daragh (2005), “Cultural Brands / Branding Cultures,” Journal of Marketing Management, 21 (5/6), 573-88.

Orth, Ulrich R. and De Marchi, Renata (2007), “Understanding the Relationships Between Functional, Symbolic, and Experiential Brand Beliefs, Product Experiential Attributes, and Product Schema: Advertising-Trial Interactions Revisited,” Journal of Marketing Theory & Practice, 15 (3), 219-33.

-------, Ulrich R., Harold F. Koenig, and Zuzana Firbasova (2007), “Cross-national Differences in Consumer Response to the Framing of Advertising Messages An Exploratory Comparison from Central Europe,” European Journal of Marketing,. 41 (3/4), 327-48.

Packer, Martin J. (1985), “Hermeneutic Inquiry in the Study of Human Conduct,” American Psychologist, 40 (10), 1081-93.

Pankhania, Asha, Nick Lee, and Graham Hooley (2007), “Within-country Ethnic Differences and Product Positioning: a Comparison of the Perceptions of Two British sub-cultures,” Journal of Strategic Marketing, 15 (2/3), 121-38.

220

Park, C. Whan, Bernard J. Jaworski, and Deborah J. Maclnnis (1986), “Strategic Brand Concept-Image Management,” Journal of Marketing, 50 (4), 135-45.

-------, C. Whan, Sandra Milberg, and Robert Lawson (1991), “Evaluation of Brand Extensions: The Role of Product Feature Similarity and Brand Concept Consistency,” Journal of Consumer Research, 18 (2), 185-85.

-------, C. Whan and Deborah J Macinnis (2006), “What's In and What's Out: Questions on the Boundaries of the Attitude Construct,” Journal of Consumer Research, 33 (1), 16-18.

Pawle, John and Peter Cooper (2006), “Measuring Emotion -- Lovemarks, The Future Beyond Brands,” Journal of Advertising Research, 46 (1), 38-48.

Peter, Paul J. and Jerry C. Olson (2005), Consumer behaviour and Marketing strategy, (7th ed.). New York, NJ: McGrow Hill.

Peters, Rolf-Herbert (2008), The Puma Story: The Remarkable Turnaround of an Endangered Species into one of the World's Hottest Sportlifestyle Brands. London: Marshall Cavendish.

Phau, Ian I.P. and Kong Cheen K.L. Lau (2000), “Conceptualising Brand Personality: A Review and Research Propositions,” Journal of Targeting, Measurement and Analysis for Marketing, 9 (1), 52-69.

Plassmann, Hilke, Tim Ambler, Sven Braeutigam, and Peter Kenning (2007), “What can Advertisers Learn from Neuroscience?,” International Journal of Advertising, 26 (2), 151-75.

Podsakoff, Philip M., Scott B. MacKenzie, Jeong-Yeon Lee and Nathan P. Podsakoff (2003), “Common Method Biases in Behavioral

221

Research: A Critical Review of the Literature and Recommended Remedies,” Journal of Applied Psychology, 88 (5), 879-903.

Polegato, R. and R. Bjerke (2006), “The Link between Cross-Cultural Value Associations and Liking: The case of Benetton and its advertising,” Journal of Advertising Research, 46 (3), 263-73.

Pringle, Hamish and Marjorie Thompson (1999), Brand Spirit: How Cause Related Marketing Builds Brands. New York: John Wiley & Sons.

Quester, Pascale, Michael Beverland, and Francis Farrelly (2006), “Brand-Personal Values Fit and Brand Meanings: Exploring the Role Individual Values Play in Ongoing Brand Loyalty in Extreme Sports Subcultures” Advances in Consumer Research, 33 (1), 21-27.

Raggio, Randle D. and Robert P. Leone (2007), “The Theoretical Separation of Brand Equity and Brand Value: Managerial Implications for strategic planning,” Journal of Brand Management, 14 (5), 380–95.

Raiklin, Ernest and Bulent Uyar (1996), “On The Relativity of The Concepts of Needs, Wants, Scarcity and Opportunity Cost,” International Journal of Social Economics, 23 (7), 49-56.

Ramberg, Bjørn and Kristin Gjesdal (2009), “Hermeneutics,” in The Stanford Encyclopedia of Philosophy (Spring 2009 Edition), Edward N. Zalta, ed. (accessed 4th January 2008) [available at http://plato.stanford.edu/archives/spr2009/entries/hermeneutics/].

Raymond, Mary Anne, John D. Mittelstaedt, and Christopher D. Hopkins (2003), “When Is a Hierarchy Not a Hierarchy? Factors Associated with Different Perceptions of Needs, With Implications for

222

Standardization - Adaptation Decisions in Korea,” Journal of Marketing Theory & Practice, 11 (4), 12-25.

Reynolds, Thomas J. and Jonathan Gutman (1988), “Laddering Theory, Method, Analysis, and Interpretation,” Journal of Advertising Research, 28 (1), 11-31.

Richins, Marsha L. (1994), “Valuing Things: The Public and Private Meanings of Possessions,” Journal of Consumer Research, 21 (3), 504-04.

Riesenbeck Hajo and Jesko Perrey (2007), Power Brands. Weinheim: Wiley – VCH Verlag.

Rindfleisch, Aric, Alan J. Malter, Shankar Ganesan and Christine Moorman (2008), “Cross-Sectional versus Longitudinal Survey Research: Concepts, Findings, and Guidelines,” Journal of Marketing Research, 45 (3) 261-79.

Roehm, Michelle L.; Brady, Michael K. (2007), “Consumer Responses to Performance Failures by High-Equity Brands,” Journal of Consumer Research, 34 (4), 537-45.

Romani, Laurence (2004), “Culture in Management: the Measurement of Differences” in International Human Resource Management, Anne-Wil Harzing and Ruysseveldt, Eds. 2 ed. London: Sage Publications.

Ross, Jill and Rod Harradine (2004), “I'm Not Wearing That!: Branding and Young Children,” Journal of Fashion Marketing & Management, 8 (1), 11-26.

Roth, Martin S. (1995), “The Effects of Culture and Socioeconomics on The Performance of Global Brand Image Strategies,” Journal of Marketing Research, 32 (2), 163-63.

223

Ruigrok, Winfried, Michael Gibbert, and Barbara Kaes (2005), “In Search of Rigorous Case Studies: Patterns of Validity and Reliability across Ten Management Journals 1995-2000,” Working paper, University of St. Gallen.

Ryan M. Richard and Edward L. Deci (2000), “Self-Determination Theory and the Facilitation of Intrinsic Motivation, Social Development, and Well-Being,” American Psychologist, 55 (1), 68-78.

Sawyer, Alan G. and Dwayne, A. Ball (1981), “Statistical Power and Effect Size in Marketing Research,” Journal of Marketing Research, 18 (3), 275-90.

Schein, Edgar H. (1988), Organizational Culture and Leadership. San Francisco: Jossey-Bass.

Schimmack, Ulrich, Shigehiro Oishi, and Ed Diener (2005), “Individualism: A Valid and Important Dimension of Cultural Differences Between Nations,” Personality & Social Psychology Review, 9 (1), 17-31.

Schmitt, Bernd H. (1999), Experiential marketing: How To Get Customers To Sense, Feel, Think, Act and Relate to Your Company and Brands. New York: The Free Press.

Schönemann, Peter. H. (1990), “Facts, Fictions, and Common-Sense About Factors and Components,” Multivariate Behavioral Research, 25 (1), 47-51.

Schroeder, Jonathan E. (2005), “The Artist and The Brand,” European Journal of Marketing, 39 (11/12), 1291-305.

224

Schuh, Arnold (2007), “Brand Strategies of Western MNCs as Drivers of Globalization in Central and Eastern Europe,” European Journal of Marketing, 41 (3/4), 274-91.

Schwartz, Shalom H. (1999), “A Theory of Cultural Values and Some Implications for Work,” Applied Psychology: An International Review, 48 (1), 23-47.

Scitovsky, Tibor (1986), Human Desire and Economic Satisfaction. Brighton: Wheatsheat.

Sheth, Jagdish N., Bruce I. Newman, and Barbara L. Gross (1991), “Why We Buy What We Buy: A Theory of Consumption Values,” Journal of Business Research, 22 (2), 159-70.

Shook, Christopher L., David J. Ketchen Jr, Cynthia S. Cycyota, and Dilene Crockett (2003), “Data Analytic Trends and Training in Strategic Management,” Strategic Management Journal, 24 (12), 1231-37.

Smith, Robert E. and William R. Swinyard (1983), “Attitude-Behavior Consistency: The Impact of Product Trial versus Advertising,” Journal of Marketing Research, 20 (3), 257-67.

Snow, Charles C. and James B. Thomas (1994), “Field Research Methods in Strategic Management: Contributions to Theory Building and Testing,” Journal of Management Studies, 31 (4), 457-80.

Srinivasan, V. (1979), “Network Models for Estimating Brand-Specific Effects in Multi-Attribute Marketing Models,” Management Science, 25 (1), 11-21.

225

Steenkamp, Jan-Benedict E.M., Rajeev Batra, and Dana L. Alden (2003), “How Perceived Brand Globalness Creates Brand Value,” Journal of International Business Studies, 34 (1), 53-65.

Stern Barbara B. (2006), “What Does Brand Mean? Historical-Analysis Method and Construct Definition,” Journal of Academy of Marketing Science, 34 (2), 216-23.

Sung Youngjun and Spencer F. Tinkham (2005), “Brand Personality Structures in the United States and Korea: Common and Culture-Specific Factors,” Journal of Consumer Psychology, 15 (4), 334-50.

Sweeney C. Jillian and Brandon Carol (2006), “Brand Personality: Exploring the Potential to Move from Factor Analytical to Circumplex Models,” Psychology & Marketing, 23 (8), 639-63.

------- C. Jillian and Geoffrey N. Soutar (2001), “Consumer perceived value: The development of a multiple item scale,” Journal of Retailing, 77 (2), 203–220.

Synodinos, Nicolaos E. Charles F. Keown, Laurence W. Jacobs (1989), “Transnational Advertising Practices: A Survey of Leading Brand Advertisers in Fifteen Countries,” Journal of Advertising Research, 29 (2), 43-50.

Szymanski, David M., Sundar G. Bharadwaj, and Rajan P. Varadarajan (1993), “Standardization versus Adaptation of International Marketing Strategy: An Empirical Investigation,” Journal of Marketing, 57 (4), 1-17.

Tabachnick, Barbara G. and Linda S. Fidell (2001), Using Multivariate Statistics. Boston: Allyn and Bacon.

226

Taylor, Edward (1871), Primitive cultures. London: John Murray.

Teather, David (2004), “Coca-Cola Reduces Profit Targets,” The Guardian (November 12, 2004) (accessed 15th December 2008) [available at: http://www.indiaresource.org/news/2004/1053.html].

Thompson, Craig J. (2004), “Beyond Brand Image: Analyzing the Culture of Brands,” Advances in Consumer Research, 31 (1), 98-100.

-------, Craig J., Aric Rindfleisch, and Zeynep Arsel (2006), “Emotional Branding and the Strategic Value of the Doppelgaenger Brand Image,” Journal of Marketing, 70 (1), 50-64.

-------, Craig J., Howard R Pollio, and William B. Locander (1994), “The Spoken and the Unspoken: A Hermeneutic Approach to Understanding the Cultural Viewpoints That Underlie Consumers' Expressed Meanings,” Journal of Consumer Research, 21 (3), 432-52.

Thompson, Jim (2008), “Brandjunkies on the Influence of Brands: The Brandjunkie Survey Result,” Brandchannel (accessed 12th November, 2008) [available at http://www.brandchannel.com/features_effect.asp?pf_id=415]

Triandis, Harry C. and Michael J. Gelfand (1998), “Converging Measurement of Horizontal and Vertical Individualism and Collectivism,” Journal of Personality and Social Psychology, 74 (1), 118-28.

Trochim William and James P. Donnelly (2007), The Research Methods Knowledge Base (3rd ed.). Mason, OH: Atomic Dog Publishing.

Trompenaars, Fons (1994), Riding the Waves of Culture. London: Nicholas Brealey.

227

Vargo, Stephen L and Robert F. Lusch (2004), “Evolving to a New Dominant Logic for Marketing,” Journal of Marketing, 68 (1), 1-17.

Velicer, Wayne F. and Douglas N. Jackson (1990), “Component Analysis versus Common Factor-Analysis – Some Further Observations,” Multivariate Behavioral Research, 25 (1), 97-114.

Veloutsou, Cleopatra (2007), “Identifying the Dimensions of the Product-Brand and Consumer Relationship,” Journal of Marketing Management, 23 (1/2), 7-26.

-------, Cleopatra (2008), “Branding: A Constantly Developing concept,” Journal of Brand Management, 15 (5), 299-300.

-------, Cleopatra and Luiz Moutinho (2008), “Brand Relationships Through Brand Reputation and Brand Tribalism,” Journal of Business Research, In Press, Available online 6 August 2008

(doi:10.1016/j.jbusres.2008.05.010)

Verhage, Bronislaw J., Ugur Yavas, and Robert T. Green (1990), “Perceived Risk: A Cross-cultural Phenomenon,” International Journal of Research in Marketing, 7 (4), 297-303.

Ward, James C. and Amy L. Ostrom (2003), “The Internet as Information Minefield: An Analysis of the Source and Content of Brand Information Yielded by Net Searches,” Journal of Business Research, 56 (11) 907-14.

Watson, John, Steve Lysonski, Tamara Gillan, and Leslie Raymore (2002), “Cultural Values and Important Procession: A Cross-cultural Analysis,” Journal of Business Research, 55 (11), 923–31.

Weick, E. Karl (1995), “What Theory Is Not, Theorizing Is,” Administrative Science Quarterly, 40 (3), 385-90.

228

Wilcox, Rand (2002), “Methodological article: Understanding the Practical Advantages of Modern ANOVA Methods,” Journal of Clinical Child and Adolescent Psychology, 31 (3), 399-412.

Wipperfurth, Alex (2006), Brand Hijack: Marketing Without Marketing: New York: Portfolio Trade.

Yin, K. Robert Ed. (1998), The Abridged Version of Case Study Research. Thousand Oaks, CA: Sage.

Zaltman, Gerald (2003), How Customers Think: Essential Insights into the Mind of the Market. Boston: Harvard Business School Press.

Zhang, Yong and James P. Neelankavil (1997), “The Influence of Culture on Advertising Effectiveness in China and the USA,” European Journal of Marketing, 31 (1/2), 134-49.

Zhou, Lianxi, Lefa Teng, and Patrick S. Poon (2008), “Susceptibility to Global Consumer Culture: A Three-Dimensional Scale,” Psychology & Marketing, 25 (4), 336-51.

Zumpano, Anthony (2007), “Similar Search Results: Google Wins”, Brandchannel, (accessed 12th November, 2008) [available at http://www.brandchannel.com/features_effect.asp?pf_id=352]

I

Appendices

II

Table ap1-1. Visual Indications of What Coca-cola Means and Does Not Mean to the Interviewees Interviewee Images that indicate what

Coca-cola means to interviewees

Images that indicate what Coca-cola does not mean to interviewees

Hanna (Finland)

Minna (Finland)

Appendix 1. The Images Collected in the Interviews

III

Tiia (Finland)

Kari (Finland)

Hannu (Finland)

Anisha (India)

Praveen (India)

IV

Samyuktha (India)

Hemant (India)

Himanshu (India)

Jane (US)

Samantha (US)

V

Lucy (US)

Robert (US)

Peter (Austria)

VI

Michaela (Austria)

Christina (Austria)

Ronald (Austria)

Petra (Austria)

VII

Table ap1-2. Visual Indications of What Google Means and Does Not Mean to the Interviewees Interviewee Images that indicate what

Google means to interviewees Images that indicate what Google does not mean to interviewees

Hanna (Finland)

Minna (Finland)

Tiia (Finland)

Kari (Finland)

VIII

Hannu (Finland)

Anisha (India)

Praveen (India)

Samyuktha (India)

Hemant (India)

Himanshu (India)

IX

Jane (US)

Samantha (US)

Lucy (US)

Robert (US)

X

Peter (Austria)

Michaela (Austria)

Christina (Austria)

Ronald (Austria)

Petra (Austria)

XI

Table ap1-3. Visual Indication of what Nike Means and Does Not Mean to the Interviewees Interviewee Images that indicate what

Nike means to interviewees Images that indicate what Nike does not mean to interviewees

Hanna (Finland)

Minna (Finland)

Tiia (Finland)

XII

Kari (Finland)

Hannu (Finland)

Anisha (India)

Praveen (India)

XIII

Samyuktha (India)

Hemant (India)

Himanshu (India)

Jane (US)

XIV

Samantha (US)

Lucy (US)

Robert (US)

XV

Peter (Austria)

Michaela (Austria)

Christina (Austria)

Cheap prices

Ronald (Austria)

XVI

Petra (Austria)

XVII

Table ap1-4. Visual Indication of What Nokia Means and Does Not Mean to the Interviewees Interviewee Images that indicate what

Nokia means to interviewees Images that indicate what Nokia does not mean to interviewees

Hanna (Finland)

Minna (Finland)

Tiia (Finland)

Kari (Finland)

Hannu (Finland)

XVIII

Anisha (India)

Praveen (India)

Samyuktha (India)

XIX

Hemant (India)

Himanshu (India)

Jane (US)

Samantha (US)

Lucy (US)

XX

Robert (US)

n.a.

Peter (Austria)

Michaela (Austria)

Christina (Austria)

Ronald (Nokia)

Petra (Austria)

Table ap1-5. Visual Indication of What Toyota Means and Does Not Mean to the Interviewees Interviewee Images that indicate what

Toyota means to interviewees

Images that indicate what Toyota does not mean to 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

XXV

Table ap2-1. The Product Categories for Coca-cola, Google and Nike

Coca-cola Google Nike Mineral water Cakes Cigarettes Alcohol Toys Ice cream Energy bars Jeans Playing cards Artificial fertilizers

City guides Space shuttles Fashion items In class lectures on

global issues Life vests Books GPS navigation

gadgets Info stalls in the cities World sports games Virtual big brother

show Music Art

Sports TV Training on how to win presidential

campaign Sleepwear Spa centres Music Luxurious handbags Diamond mining in Africa High quality furniture Adventurous travels agency Sandals Cushions for chairs Innovative electrical appliances Jogging fields Sport utility vehicles (SUVs) Training people how to achieve what

they though was impossible Table ap2-2. The Images for Coca-cola

Table ap2-3. The Images for Google

Table ap2-4. The Images for Nike

Appendix 2. The Product Categories and Images Used in the Pilot Phase Survey

XXVI

Cross-cultural Branding I. Default Section

Thank you for participating in my research on differences in brand meaning among cultures. This questionnaire will take about 10 minutes of your time. The answers are anonymous. 1. Which University do you attend? ____________________________ 2. What is your major? ____________________________ 3. What type of degree are you studying for? a) BSc b) BA c) MSc d) MBA e) other 4. Which semester are you currently enrolled in? a) 1 b) 3 c) 5 d) 7 e) 9 f) other 5. How old are you? ____________________________ 6. Have you lived abroad for longer than 6 months at a stretch? a) yes b) no 7. If the answer is yes, please indicate when, where, and for how long. ________________________________________________________ 8. Do you use the following brands? Coca-Cola a) yes b) no Google a) yes b) no Nike a) yes b) no

II. Coca-Cola Coca-Cola is a drink we often encounter. Some of us like to drink it and some of us don’t. As you look at the following questions, please answer them by thinking about how it feels to experience Coca-Cola. 9. What does Coca-Cola mean to you? ____________________________ 10. To what extent do the associations listed below indicate what Coca-Cola means to you?

It means very little it means a lot American 1 2 3 4 5 6 7 traditional 1 2 3 4 5 6 7 worldwide 1 2 3 4 5 6 7 classic 1 2 3 4 5 6 7 warm summer 1 2 3 4 5 6 7

Appendix 3. The Questionnaire

XXVII

original 1 2 3 4 5 6 7 American lifestyle 1 2 3 4 5 6 7 Christmas 1 2 3 4 5 6 7 calories 1 2 3 4 5 6 7 drink it when tired 1 2 3 4 5 6 7 hot day 1 2 3 4 5 6 7 bonding with friends 1 2 3 4 5 6 7 global 1 2 3 4 5 6 7 healthy 1 2 3 4 5 6 7 unknown 1 2 3 4 5 6 7 recognisable 1 2 3 4 5 6 7 cold 1 2 3 4 5 6 7 fun 1 2 3 4 5 6 7 identifiable 1 2 3 4 5 6 7 obesity 1 2 3 4 5 6 7 sugar 1 2 3 4 5 6 7 Christmas song 1 2 3 4 5 6 7 unhealthy 1 2 3 4 5 6 7 energetic 1 2 3 4 5 6 7 freshness 1 2 3 4 5 6 7 Santa Claus 1 2 3 4 5 6 7 exciting 1 2 3 4 5 6 7 no sleeping 1 2 3 4 5 6 7 cooling 1 2 3 4 5 6 7 familiar 1 2 3 4 5 6 7 natural 1 2 3 4 5 6 7 parties 1 2 3 4 5 6 7 multinational 1 2 3 4 5 6 7 keeps me going 1 2 3 4 5 6 7 boring 1 2 3 4 5 6 7 cool 1 2 3 4 5 6 7 sweet 1 2 3 4 5 6 7 great brand 1 2 3 4 5 6 7 refreshing 1 2 3 4 5 6 7

10. What do you feel for Coca-Cola?

a) I love it b) I like it c) I am indifferent d) I don’t like it e) I hate it

III. Google

When something smells ‘fishy’ and you want to check whether it holds true, you will most likely hear a voice urging you to Google. With Google in mind, please answer the following questions.

11. What does Google mean to you? ____________________________

XXVIII

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

ld1 2 3 4 5 6 7

easy to 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

XXIX

satisfaction 1 2 3 4 5 6 7 multibillion industry 1 2 3 4 5 6 7

13. What do you feel for Google? a) I love it b) I like it c) I am indifferent d) I don’t like it e) I hate

it IV. Nike

Most of us have seen a commercial speaking about features and benefits a new Nike product offers. Whether you have worn Nike shoes or not, please answer the following questions.

14. What does Nike mean to you? ____________________________ 15. To what extent do the associations listed below indicate what Nike means to

you? It means very little it means a lot

sports 1 2 3 4 5 6 7 achiever 1 2 3 4 5 6 7 freedom 1 2 3 4 5 6 7 athletes 1 2 3 4 5 6 7 cool 1 2 3 4 5 6 7 expensive 1 2 3 4 5 6 7 sweat shop 1 2 3 4 5 6 7 good quality 1 2 3 4 5 6 7 lazy 1 2 3 4 5 6 7 life 1 2 3 4 5 6 7 crowd cheering 1 2 3 4 5 6 7 comfortable 1 2 3 4 5 6 7 jogging 1 2 3 4 5 6 7 Michael Jordan 1 2 3 4 5 6 7 powerful 1 2 3 4 5 6 7 self-assured 1 2 3 4 5 6 7 athletic 1 2 3 4 5 6 7 trendy 1 2 3 4 5 6 7 pricey 1 2 3 4 5 6 7 child labour 1 2 3 4 5 6 7 low quality 1 2 3 4 5 6 7 just do it 1 2 3 4 5 6 7 fitness 1 2 3 4 5 6 7 innovative 1 2 3 4 5 6 7 running 1 2 3 4 5 6 7

XXX

sporty 1 2 3 4 5 6 7 victory 1 2 3 4 5 6 7 failure 1 2 3 4 5 6 7 young 1 2 3 4 5 6 7 durable 1 2 3 4 5 6 7 air technology 1 2 3 4 5 6 7 socially responsible 1 2 3 4 5 6 7 innovation 1 2 3 4 5 6 7 winning 1 2 3 4 5 6 7 weak 1 2 3 4 5 6 7 Tiger Woods 1 2 3 4 5 6 7 positive thinking 1 2 3 4 5 6 7

16. What do you feel for Nike? a) I love it b) I like it c) I am indifferent d) I don’t like it e) I hate it 17. If you have any general comments, please write them in the space below: _____________________________________________________________ _____________________________________________________________ _____________________________________________________________

XXXI

393837363534333231302928272625242322212019181716151413121110987654321

Component Number

10

8

6

4

2

0

Eige

nval

ueScree Plot

Figure ap4-1. The Initial Scree plot for Coca-Cola Table ap4-1. The Communalities for Coca-Cola

Extraction Extraction American .475 sugar .628 traditional .720 Christmas song .795 worldwide .631 unhealthy .656 classic .667 energetic .551 warm summer .639 freshness .659 original .448 Santa Claus .767 Christmas .764 exciting .609 calories .626 no sleeping .448 drink it when tired .659 cooling .712 hot day .636 familiar .371 bonding with friends .717 not natural .421 global .710 parties .565 not healthy .466 multinational .649 not unknown .292 keeps me going .650 recognisable .476 cool .649 cold .537 sweet .457 fun .640 great brand .419 identifiable .437 refreshing .768 obesity .642

Extraction Method: Principal Component Analysis.

Appendix 4. The Statistics

XXXII

Table ap4-2. The Total Variance Explained for Coca-Cola

Initial Eigenvalues Extraction Sums of Squared Loadings

Rotation Sums of Squared Loadings a

Component Total % of Variance

Cumulative % Total

% of Variance

Cumulative % Total

1 8.339 22.539 22.539 8.339 22.539 22.539 5.6112 3.959 10.701 33.240 3.959 10.701 33.240 3.4473 2.827 7.639 40.879 2.827 7.639 40.879 3.9734 2.332 6.302 47.181 2.332 6.302 47.181 3.6165 1.620 4.378 51.559 1.620 4.378 51.559 2.8406 1.554 4.200 55.759 1.554 4.200 55.759 3.9457 1.326 3.584 59.343 1.326 3.584 59.343 4.3988 1.118 3.022 62.365 9 0.968 2.616 64.981

Extraction Method: Principal Component Analysis. a When components are correlated sums of squared loadings cannot be added to obtain a total variance. Table ap4-3. The Structure Matrix for Coca-Cola

Component 1 2 3 4 5 6 7 refreshing .818 .160 -.277 -.236 .228 .414 -.281 cooling .817 .148 -.201 -.123 .134 .273 -.385 cool .739 .207 -.204 -.228 .093 .339 -.454 cold .717 .176 -.048 -.101 .068 .194 -.287 freshness .685 .126 -.304 -.331 .214 .483 -.262 great brand .530 .353 -.312 -.214 .127 .131 -.216 familiar .451 .257 -.166 .065 .347 -.032 -.199 global .068 .796 .179 -.088 .218 .107 -.074 multinational .260 .781 .005 -.082 .057 .054 -.129 worldwide .057 .737 .082 -.044 .369 .004 .096 recognisable .294 .618 .074 -.136 .236 -.108 .193 identifiable .427 .507 -.019 -.228 .240 -.112 .070 not unknown .184 .418 .051 .050 -.047 -.222 .247 unhealthy -.165 .060 .801 -.003 .017 -.102 .240 sugar -.109 .097 .775 -.120 .042 -.100 .200 obesity -.239 .014 .774 .080 -.041 -.023 .019 calories -.102 .047 .759 -.048 -.033 -.079 -.071 sweet .286 .270 .491 -.196 -.083 .111 .082 not natural -.235 .080 .448 .174 -.348 -.195 .391 not healthy -.077 .022 .441 .332 -.110 -.395 .431 Christmas song .086 .092 .091 -.881 .081 .137 -.065

XXXIII

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.

XXXIV

42

41

40

39

38

37

36

35

34

33

32

31

30

29

28

27

26

25

24

23

22

21

20

19

18

17

16

15

14

13

12

11

10

987654321

Component Number

14

12

10

8

6

4

2

0

Eige

nval

ue

Scree Plot

Figure ap4-2. The Initial Scree Plot for Google Table ap4-4. The Communalities for Google

Extraction Extraction fun .585 trustful .571 fast .568 all knowing .585 not old .407 high speed .518 world .627 global .660 life saver .544 almighty .564 no privacy .657 very important .667 useful .720 no independency .515 entertainment .639 no problem .497 happiness .734 informative .672 huge .520 knowledge .631 connecting with the world .572 observes people's habits .593 easy to use .642 encyclopaedia .608 provides data .389 efficient .494 worldwide .735 young .677 basic necessity .562 dominant .604 freedom .587 innovative .708 helpful .729 creative .679 smart .636 satisfaction .603 convenient .639 multibillion industry .521

Extraction Method: Principal Component Analysis.

XXXV

Table ap4-5. The Total Variance Explained for Google

Initial Eigenvalues Extraction Sums of Squared

Loadings

Rotation Sums of Squared Loadings a

Component Total

% of Variance

Cumulative % Total

% of Variance

Cumulative % Total

1.000 11.898 31.310 31.310 11.898 31.310 31.310 6.5752.000 3.178 8.364 39.674 3.178 8.364 39.674 6.7403.000 2.390 6.290 45.963 2.390 6.290 45.963 2.2534.000 1.565 4.118 50.082 1.565 4.118 50.082 6.1285.000 1.423 3.744 53.826 1.423 3.744 53.826 4.5576.000 1.283 3.375 57.201 1.283 3.375 57.201 5.9097.000 1.123 2.954 60.156 1.123 2.954 60.156 3.6898.000 0.999 2.629 62.785

Extraction Method: Principal Component Analysis. a When components are correlated sums of squared loadings cannot be added to obtain a total variance. Table ap4-6. The Structure Matrix for Google Component 1 2 3 4 5 6 7very important .766 -.425 .059 .295 -.191 -.305 .323almighty .685 -.106 .296 .329 -.296 -.321 .172all knowing .648 -.221 .031 .343 -.327 -.381 .460life saver .640 -.270 -.006 .328 -.530 -.194 .252trustful .630 -.322 -.270 .367 -.337 -.144 .307basic necessity .629 -.492 .133 .222 -.375 -.249 .176satisfaction .618 -.410 -.023 .551 -.421 -.312 .345no problem .594 -.280 -.029 .358 -.426 -.200 .396dominant .509 -.145 .393 .434 -.203 -.347 -.232helpful .353 -.836 -.092 .296 -.158 -.395 .312useful .205 -.815 -.028 .033 -.156 -.343 .224easy to use .149 -.773 -.046 .273 -.169 -.456 .177convenient .285 -.760 -.102 .300 -.088 -.402 .410fast .235 -.725 -.019 .323 -.070 -.424 .141provides data .203 -.526 -.014 .321 -.044 -.426 .340no privacy -.195 .104 .762 -.090 .014 -.103 -.143observes people's habits .164 .021 .743 .201 -.020 -.165 .003no independency .158 .024 .703 .080 -.118 -.052 -.021innovative .360 -.206 .115 .828 -.300 -.229 .228creative .319 -.252 .099 .812 -.300 -.213 .235young .281 -.034 .102 .735 -.450 -.282 .127smart .400 -.536 .057 .664 -.348 -.228 .310

XXXVI

efficient .446 -.462 -.029 .545 -.230 -.272 .400multibillion industry .284 -.315 .270 .536 -.169 -.468 -.110high speed .483 -.476 -.015 .492 -.026 -.412 .283happiness .471 -.189 .018 .369 -.809 -.292 .215entertainment .238 -.205 .105 .278 -.778 -.214 .179fun .278 -.230 -.014 .407 -.707 -.166 .225freedom .565 -.285 .040 .482 -.605 -.264 .153not old -.173 -.288 -.229 .243 .348 -.017 .265worldwide .176 -.489 .120 .211 -.113 -.837 .139global .264 -.321 .066 .295 -.048 -.792 .199world .127 -.334 .119 .166 -.162 -.773 .243connecting with the world .182 -.472 .137 .276 -.355 -.646 .314huge .378 -.358 .117 .155 -.304 -.630 -.026knowledge .405 -.386 -.030 .332 -.224 -.433 .690informative .385 -.473 .005 .348 -.142 -.490 .686encyclopaedia .393 -.233 .151 .272 -.348 -.234 .682Extraction Method: Principal Component Analysis. Rotation Method: Oblimin with Kaiser Normalization.

XXXVII

37363534333231302928272625242322212019181716151413121110987654321

Component Number

12,5

10,0

7,5

5,0

2,5

0,0

Eige

nval

ueScree Plot

Figure ap4-3. The Initial Scree Plot for Nike Table. Ap4-7. The Communalities for Nike

Extraction Extraction sports .397 pricey .812 achiever .532 child labour .636 freedom .576 just do it .382 athletes .515 fitness .714 cool .618 innovative .654 expensive .825 running .737 sweat shop .599 sporty .732 good quality .557 victory .713 not lazy .554 not failure .657 life .475 young .407 crowd cheering .532 durable .512 comfortable .446 air technology .422 jogging .558 not socially responsible .512 Michael Jordan .503 innovation .656 powerful .515 winning .699 self-assured .526 not weak .598 athletic .628 Tiger Woods .411 trendy .558 positive thinking .637

Extraction Method: Principal Component Analysis.

XXXVIII

Table ap4-8. The Total Variance Explained for Nike

Initial Eigenvalues Extraction Sums of Squared

Loadings

Rotation Sums of Squared Loadings a Compo

nent Total

% of Variance

Cumulative % Total

% of Variance

Cumulative % Total

1 12.130 33.695 33.695 12.130 33.695 33.695 10.7232 2.853 7.926 41.621 2.853 7.926 41.621 2.6993 2.617 7.270 48.891 2.617 7.270 48.891 2.7214 1.797 4.991 53.882 1.797 4.991 53.882 3.3325 1.410 3.915 57.797 1.410 3.915 57.797 7.7866 1.306 3.627 61.424 7 1.056 2.934 64.358 8 .987 2.742 67.100 Extraction Method: Principal Component Analysis. a When components are correlated sums of squared loadings cannot be added to obtain a total variance. Table ap4-9. The Structure Matrix for Nike

Component 1 2 3 4 5 victory .825 .151 -.073 -.265 -.520 winning .823 .135 -.034 -.237 -.477 innovation .773 .048 -.319 -.264 -.406 positive thinking .772 .133 -.230 -.044 -.371 innovative .761 .017 -.340 -.252 -.438 cool .725 .072 -.320 -.248 -.510 self-assured .724 .040 -.120 -.193 -.363 freedom .714 .009 -.320 -.115 -.409 crowd cheering .709 -.060 .012 -.256 -.345 achiever .706 .123 -.120 -.232 -.482 powerful .699 .041 -.073 -.280 -.439 life .670 -.085 -.182 -.112 -.288 air technology .615 -.024 .036 -.027 -.361 trendy .610 -.017 -.326 -.457 -.425 durable .599 .081 -.339 -.383 -.439 young .596 .042 -.203 -.314 -.300 Tiger Woods .557 .119 .196 -.039 -.180 Michael Jordan .504 .086 .418 -.176 -.242 just do it .504 .226 .176 -.102 -.416 not failure .127 .787 -.199 .053 -.171 not weak .095 .761 -.071 -.087 -.265 not lazy -.015 .738 -.061 .073 -.161

XXXIX

child labour -.128 -.250 .755 -.045 -.008 sweat shop .003 -.287 .675 -.241 -.076 not socially

ibl-.419 .287 .530 .085 .103

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.

XL

Table ap4-10. The Levene's Tests of Equality of Error Variances for the Coca-Cola’s Meanings

Dependent variable F df1 df2 Sig. Refreshing 2.429 11 354 .006

Globally recognised 1.847 11 354 .045 Unhealthy 1.729 11 354 .066 Christmas .545 11 354 .872

Classic 3.993 11 354 .000 Energising 1.298 11 354 .224

Bonding with friends 1.066 11 354 .388 Table ap4-11. The Non-parametric Kruskal-Wallis Test for the Dependent Variable Refreshing

Independent v. Country Type of Coca-Cola user Chi-Square 23.339 56.977 df 3 2 Asymp. Sig. .000 .000

Table ap4-12. The Non-parametric Mann-Whitney Tests for the Dependent Variable Refreshing and the Independent Variable Country

Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4 Mann-Whitney U 4737.0 2988.0 2966.0 3288.0 3312.0 2061.0Wilcoxon W 10953. 9204.000 9182.0 8959.0 5392.0 4141.0Z -2.479 -4.394 -1.815 -3.205 -.257 -2.527Asymp. Sig. (2-tailed) .013 .000 .069 .001 .797 .0111 = Austria; 2 = Finland; 3 = India; 4 = United States Table ap4-13. The Non-parametric Mann-Whitney Tests for the Dependent Variable Refreshing and the Independent Variable Type of Coca-Cola User

Independent v. 1 vs. 2 1 vs. 3 2 vs. 3 Mann-Whitney U 5703.0 3463.0 3412.0 Wilcoxon W 10956.0 6238.0 6187.0 Z -5.796 -6.401 -1.085 Asymp. Sig. (2-tailed) .000 .000 .278

1 = Engaged users; 2 = Disengaged users; 3 = Non-users Table ap4-14. The Non-parametric Kruskal-Wallis Tests for the Dependent Variable Globally Recognised

Independent v. Country Type of Coca-Cola user Chi-Square 10.730 2.240 df 3 2 Asymp. Sig. .013 .326

XLI

Table ap4-15. The Non-parametric Mann-Whitney Tests for the Dependent Variable Globally recognised and the Independent Variable Country Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4 Mann-Whitney U 5798.0 3920.0 2694.0 3790.0 2591.0 2549.0 Wilcoxon W 11469.0 7575.0 4774.0 7445.0 4671.0 4629.0 Z -.184 -2.026 -2.658 -1.883 -2.576 -.656 Asymp. Sig. (2-tailed) .854 .043 .008 .060 .010 .512 1 = Austria; 2 = Finland; 3 = India; 4 = United States Table ap4-16. The Non-parametric Kruskal-Wallis Tests for the Dependent Variable Classic

Independent v. Country Type of Coca-Cola Chi-Square 100.296 11.887 df 3 2 Asymp. Sig. .000 .003

Table ap4-17. The Non-parametric Mann-Whitney Tests for the Dependent Variable Classic and the Independent Variable Country Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4 Mann-Whitney U 2575.0 3983.0 1642.0 1510.0 3213.0 959.0 Wilcoxon W 8791.0 7638.0 7858.0 5165.0 5293.0 4614.0 Z -7.155 -1.866 -5.917 -7.888 -.576 -6.753 Asymp. Sig. (2-tailed) .000 .062 .000 .000 .565 .000 1 = Austria; 2 = Finland; 3 = India; 4 = United States Table ap4-18. The Non-parametric Mann-Whitney Tests for the Dependent Variable Classic and the Independent Variable Type of Coca-Cola User

Independent v. 1 vs. 2 1 vs. 3 2 vs. 3 Mann-Whitney U 7304.0 6582.0 3126.0 Wilcoxon W 12557.0 9357.0 8379.0 Z -3.469 -.804 -1.942 Asymp. Sig. (2-tailed) .001 .421 .052

1 = Engaged users; 2 = Disengaged users; 3 = Non-users

XLII

Table ap4-19. The Levene's Tests of Equality of Error Variances for Google’s meanings

Dependent variable F df1 df2 Sig. Almighty 1.776 7 309 .092

Helpful 3.503 7 309 .001 Intrusive 2.330 7 309 .025 Inspired 3.347 7 309 .002

Fun 1.193 7 309 .306 Global 4.238 7 309 .000

Informative 3.331 7 309 .002 Table ap4-20. The Non-parametric Kruskal-Wallis Test for the Dependent Variable Helpful and the Independent Variable Country

Independent v. Country Chi-Square 4.597 df 3 Asymp. Sig. .204

Table ap4-21. The Non-parametric Mann-Whitney Test for the Dependent Variable Helpful and the Independent Variable Type of Google User

Independent v. Type of Google user Mann-Whitney U 8254.000 Wilcoxon W 17299.000 Z -4.971 Asymp. Sig. (2-tailed) .000

Table ap4-22. The Non-parametric Kruskal-Wallis Test for the Dependent Variable Intrusive and the Independent Variable Country

Independent v. Country Chi-Square 16.204 df 3 Asymp. Sig. .001

Table ap4-23. The Non-parametric Mann-Whitney Tests for the Dependent Variable Intrusive and the Independent Variable Country

Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4 Mann-Whitney U 3635.0 3470.0 1758.0 2716.0 2079.0 1291.0 Wilcoxon W 7730.0 6171.0 3189.0 6811.0 3510.0 2722.0 Z -2.386 -.660 -3.493 -1.899 -1.279 -3.180 Asymp. Sig. (2-tailed) .017 .509 .000 .058 .201 .001

1 = Austria; 2 = Finland; 3 = India; 4 = United States

XLIII

Table ap4-24. The Non-parametric Mann-Whitney Test for the Dependent Variable Intrusive and the Independent Variable Type of Google User

Independent v. Type of Google user Mann-Whitney U 11787.000 Wilcoxon W 20832.000 Z -.588 Asymp. Sig. (2-tailed) .557

Table ap4-25. The Non-parametric Kruskal-Wallis Test for the Dependent Variable Inspired and the Independent Variable Country

Independent v. Country Chi-Square 56.020 df 3 Asymp. Sig. .000

Table ap4-26. The Non-parametric Mann-Whitney Tests for the Dependent Variable Inspired and the Independent Variable Country

Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4 Mann-Whitney U 4536.0 1996.0 1392.0 1557.0 1133.0 1667.0 Wilcoxon W 9687.0 7147.0 6543.0 5652.0 5228.0 4368.0 Z -.024 -5.155 -4.885 -5.767 -5.233 -1.322 Asymp. Sig. (2-tailed) .981 .000 .000 .000 .000 .186

1 = Austria; 2 = Finland; 3 = India; 4 = United States Table ap4-27. The Non-parametric Mann-Whitney Test for the Dependent Variable Inspired and the Independent Variable Type of Google User

Independent v. Type of Google user Mann-Whitney U 6547.000 Wilcoxon W 23383.000 Z -7.088 Asymp. Sig. (2-tailed) .000

Table ap4-28. The Non-parametric Kruskal-Wallis Test for the Dependent Variable Global and the Independent Variable Country

Independent v. Country Chi-Square 20.723 df 3 Asymp. Sig. .000

XLIV

Table ap4-29. The Non-parametric Mann-Whitney Tests for the Dependent Variable Global and the Independent Variable Country

Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4 Mann-Whitney U 3030. 3635.0 2141.0 2175.0 2155.0 1516.0 Wilcoxon W 8181. 8786.0 7292.0 4876.0 3586.0 4217.0 Z -3.973 -.157 -2.037 -3.704 -.961 -2.068 Asymp. Sig. (2-tailed) .000 .875 .042 .000 .336 .039

1 = Austria; 2 = Finland; 3 = India; 4 = United States Table ap4-30. The Non-parametric Mann-Whitney Test for the Dependent Variable Global and the Independent Variable Type of Google User

Independent v. Type of Google user Mann-Whitney U 8772.000 Wilcoxon W 17817.000 Z -4.328 Asymp. Sig. (2-tailed) .000

Table ap4-31. The Non-parametric Kruskal-Wallis Test for the Dependent Variable Informative and the Independent Variable Country

Independent v. Country Chi-Square 39.565 df 3 Asymp. Sig. .000

Table ap4-32. The Non-parametric Mann-Whitney Test for the Dependent Variable Informative and the Independent Variable Country

Independent v. 1 vs. 2 1 vs. 3 1 vs. 4 2 vs. 3 2 vs. 4 3 vs. 4 Mann-Whitney U 2996.0 3158.0 2377.0 1524.0 1270.0 1924.0 Wilcoxon W 7091.0 8309.0 7528.0 5619.0 5365.0 3355.0 Z -4.062 -1.612 -1.139 -5.877 -4.660 -.052 Asymp. Sig. (2-tailed) .000 .107 .255 .000 .000 .959

1 = Austria; 2 = Finland; 3 = India; 4 = United States Table ap4-33. The Non-parametric Mann-Whitney Test for the Dependent Variable Informative and the Independent Variable Type of Google User

Independent v. Type of Google user Mann-Whitney U 8319.000 Wilcoxon W 25155.000 Z -4.890 Asymp. Sig. (2-tailed) .000

XLV

Table ap4-34. The Levene's Test of Equality of Error Variances for Nike’s meanings

Dependent variable F df1 df2 Sig. Achiever 1.477 11 294 .139

Weak 1.471 11 294 .142 Not socially 1.750 11 294 .062

Pricey 1.601 11 294 .098 Sports 4.056 11 294 .000

Table ap4-35. The Non-parametric Kruskal-Wallis Test for the Dependent Variable Sports

Independent v. Country Type of Nike user Chi-Square 4.317 32.152 df 3 2 Asymp. Sig. .229 .000

Table ap4-36. The Non-parametric Mann-Whitney Tests for the Dependent Variable Sports and the Independent Variable Type of Nike user

Independent v. 1 vs. 2 1 vs. 3 2 vs. 3 Mann-Whitney U 3016.0 4253.0 3051.0 Wilcoxon W 12196.0 13433.0 5262.0 Z -3.716 -5.312 -1.314 Asymp. Sig. (2-tailed) .000 .000 .189

1 = Engaged users; 2 = Disengaged users; 3 = Non-users

XLVI

Curriculum Vitae Last Name: First First Name: Ivana Born: March 17, 1978 in Rijeka, Croatia

Education

04/05-09/09 University of St. Gallen, Doctoral Programme in Multicultural Management, St. Gallen, Switzerland

10/02-09/04 University of St. Gallen, M.Sc. Programme in International Management, St. Gallen, Switzerland

08/03-05/05 University of St. Gallen, CEMS M.A. in International Management, St. Gallen, Switzerland

08/03-12/03 Stockholm School of Economics, Exchange Semester within M.Sc. and CEMS M.A. Programmes, Stockholm, Sweden

10/96-06/01 Faculty of Economics in Rijeka, B.Sc. in Economics, Rijeka, Croatia

Working Experience 12/01-... University of Rijeka, Faculty of Economics, Rijeka, Croatia

Research and Teaching Assistant

06/04-09/04 Siemens VDO Automotive AG, Frankfurt Germany Intern in HVAC business unit

08/01-11/01 Beling d.o.o., Rijeka, Croatia Kaba-Elzett and other security products manager

01/00-05/01 Silca-Unican-Elzett Rt., Budapest, Hungary

AIESEC intern in marketing department Contact Information E-mail: ifirst@efri.hr