The influence of ethnicity on Country-of-Origin Image

Master Thesis, Msc BA Marketing Management

The influence of ethnicity on Country-of-Origin Image

University of Groningen Faculty of Economics and Business

Department of Marketing Nettelbosje 2 9747 AE Groningen Author Jorrit Bosma Student number: 1586157 Tuinbouwstraat 23A 9717 JA Groingen +31610097345 jorritb@hotmail.com Supervisors:

2

Management Summary

The starting point for this research was to investigate whether subcultural differences influence the Country-of-Origin Image. The problem statement of this research is: How does ethnicity affect the Country-of-Origin Image, and is this effect moderated by personal values?

Previous research already showed that Country-of-Origin Image differs between people from different cultures. However, most of these studies defined culture by the national country boundaries. Because of the existence of different subcultures within the same country, this research used ethnicity as proxy for subcultural to find differences between subcultures within the same country. A Country-of-Origin study is used to compare the different perceptions of the ethnic groups on a foreign country, its people, and its products. To see whether this relation between ethnicity and Country-of-Origin Image would be influenced by personal values and beliefs, the Individuals-Collectivism dimension is tested for its moderating effect. Through an online questionnaire the data needed to provide an answer on the problem statement was collected. Field of application of this study was Malaysia, where three main ethnic groups (Malay, Chinese, and Indians) life together. The results show that Malay people are more positive than Indians about China and the Chinese people in general which indicates that ethnicity has some influence on elements of Country-of-Origin Image. Also a positive relation between the country and its people in general (GCA) and its products in general (GPA) is found. No significant moderating effect of personal values is found. Neither did a cultural linkage with the Country-of-Origin effect the Country-of-Origin Image.

A better understanding of the need for (sub)cultural segmentation is useful for international manager to find the right balance between standardization and adaptation in their marketing strategy. Besides, insights in Country-of-Origin Image effects can be useful for companies, industries, or even governments to stimulate export activities. Whereas companies can make use of a positive country image, governments can consider country branding to use the positive halo effect of country image on industries, companies, and products.

3

Word of thanks

This thesis represents the final chapter of my life as a student and is the last step in obtaining my Master’s degree in Business Administration, specialization Marketing Management. I can look back on a fantastic student time in Groningen, Argentina, and Malaysia. A special word of thanks goes out to my parents, without their support I would never have gotten to where I am. They have always undoubtedly supported and motivated me in my choices and initiatives.

Also, I would like to thank all my friends for their welcome distractions which contributed to my delayed graduation.

For the facilitation, help and support of my graduation and my inspiring time in Malaysia I would like to thank all my colleagues at DHL Malaysia, especially, Dennis v/d Meijs, Cynthia Tan, Melissa Ting, Thijs Prijkel, and Joanna Samuel.

I would like to thank my housemate Rama for his ‘cultural lectures’ about Malaysia which resulted in the choice for this thesis topic.

Of course, I would like to thank Liane for all her time and effort she invested in my thesis. Her excellent guidance, enthusiasm, support ensured a smooth and pleasant cooperation during the whole process. Finally, muchas gracias to Janny for her ‘open door’ and support ever since the fantastic IBR Argentina experience.

4

Preface

During a four month internship period in Malaysia, I noticed that it was a very unique and special country, especially with regard to its culture. Malaysia is a country where three main ethnic groups live together; the Malays, the Chinese, and the Indian inhabitants. Although many people emphasize the harmony among the three ethnicities, the differences are definitely observable. Where most of the travel guides characterize Malaysia as a perfect example of a ‘cultural melting pot’ I personally experienced it more as a ‘cultural salad bowl’. Malaysian people do mix during work, school, and sports. However, in their free-time they mostly hang out with people who have the same ethnical background. Even though the different ethnic groups do have respect for each other, I observed that they still have their own religion, food, bars, neighborhoods, and talk their own language where possible. However, now and then you hear the mixed language Manglish. It is an English based mixed language in Malaysia with words originating from English, Malay, Hokkien, Mandarin, Cantonese, and Tamil.

5 Table of Contents Management Summary... 2 Word of thanks ... 3 Preface……… ... 4 1. Introduction ... 7 1.2 Problem statement ... 9 1.3 Research questions ... 9

1.4 Relevance of the thesis ... 9

1.4.1 Academic relevance ... 9

1.4.2 Managerial relevance ... 10

1.5 Outline ... 11

Chapter 2. Conceptual Framework ... 12

2.1 Country-of-Origin Image (COI) ... 12

2.1.1 Conceptualization of COI ... 13

2.2 Relation between cultural background and COO ... 17

2.3 The use of ethnicity as proxy for subculture ... 18

2.3.1 Conceptualization of (sub)culture ... 18 2.3.2 Operationalization of (sub)culture ... 18 2.3.3 Conceptualization of ethnicity... 19 2.4 Individualism-Collectivism ... 20 2.4.1 Hofstede dimensions ... 20 2.4.2 Conceptualization Individualism-Collectivism ... 21 2.5 Conceptual Model ... 23

Chapter 3. Research Design ... 25

3.1 Field of application ... 25 3.2 Research Method ... 25 3.2.1 Population ... 27 3.2.2 Sample ... 28 3.2 Operationalization ... 29 3.2.1 Operationalization of COI ... 29

3.2.2 Operationalization of the Individualism-Collectivism dimension ... 31

3.3 Plan of Analysis ... 33

3.3.1 Recoding of items ... 33

3.3.2 Computing new variables ... 33

3.3.3 Multicollinearity ... 34

3.3.4 Use of three models ... 35

6

3.3.6 Influence of cultural linkage on COI ... 37

Chapter 4. Results ... 38

4.1 Regression Analysis ... 38

4.2 Relation between cultural linkage and COI ... 40

4.3 Conceptual model ... 41 Chapter 5. Discussion ... 42 5.1 Conclusions ... 42 5.2 Implications ... 43 5.3 Limitations ... 43 5.4 Future research ... 45 Literature…….. ... 47 Appendixes ... 53

Appendix 1: Scores on Individualism ... 53

Appendix 2: Recoded scale items... 54

Appendix 3: Multicollinearity precaution ... 55

Appendix 4: Mail sent to DHL colleagues ... 56

Appendix 5: Flyers used to find respondents ... 57

7

1.

Introduction

Global companies always have to find the right balance in the extent they will adapt their strategy to the local, regional or cultural circumstances. This question seems to become even more relevant in the last decades (Theodosiou & Leondiou, 2003). Due to increasing liberalization of trade policies, growing stability in monetary transactions; creation of regional integrations; uninterrupted flow of goods due to relatively peaceful world conditions; and revolutionary advances in transportation, communication, and information technologies a dramatic globalization in the international business scene can be witnessed in the last decades (e.g. Czinkota & Ronkainen, 2001; Keegan, 1999). Theodosiou and Leonidou (2003) state that as a result, issues relating to the design of sound international marketing strategies to compete effectively and efficiently in this new business environment have been the focus of a sizeable stream of research. This has particularly concentrated on whether firms, irrespective of the foreign market entry mode chosen, should standardize or adapt their marketing strategy in overseas markets.

Advocates of the adaptation approach argue that, despite increasing globalization tendencies, variations between countries in such dimensions as consumer needs, use conditions, purchasing power, commercial infrastructure, culture and traditions, laws and regulations, and technological development are still too great, thus necessitating the adjustment of the firm’s marketing strategy to the different circumstances of each foreign market (Terpstra & Sarathy, 2000). These advocates also stress the fact that the ultimate objective of the firm is not cost reduction through standardization, but long-term profitability through higher sales accrued from a better exploitation of the different consumer needs across countries (Onkvisit & Shaw, 1990; Rosen, 1990; Whitelock & Pimblett, 1997).

Influence of (sub)culture on consumer needs

Cross-cultural studies are conducted to find a better understand of differences among countries, cultures, and their consumer needs. According to Heslop et al. (1998) many of those studies have dealt extensively with the influence of culture on behavior in general and with behavioral differences across nations, but relatively little is known about these issues within national boundaries and from an international perspective. Although many studies are described as “cross-cultural”, most are in fact cross-national, and the two terms are often used interchangeably. The problem with the implicit assumption that consumers within nations are homogeneous has been identified by several authors (e.g. Padmanabhan, 1988).

8

1998). Within a culture, subcultures also serve to shape attitudes and behavior and can be defined as segments of society that share distinguishing values and behavior patterns that differ from those of the overall culture (Peter & Olson, 1987).

Country-of Origin Image

In order to find possible differences in consumer perceptions among ethnic groups within the same national boundaries, a Country-of-Origin study can be performed. Country-of-Origin (COO) is the producing country of a specific product. In many occasions it is used as the “made in ….” label. The perception of this Country-of-Origin that is hold by consumers is called Country-of-Origin Image (COI). The COI construct is built on the idea that people possess stereotypical perceptions to other people and countries (Nagashima, 1970; Maheswaran, 1994). The image of a COO has a significant impact on consumers’ judgements of product quality and willingness to buy a product (e.g. Bilkey and Nes, 1982; Han and Terpstra, 1988; Maheswaran, 1994). COI serves as an extrinsic information cue –like price, brand name and retailer reputation- to infer product quality or as a product attribute that can provide certain benefits (Lin and Sternquist, 1994). Botschen and Hemettsberger (1998) reported that consumers link COI not only to product quality, but also to feelings of national pride and memories of past visits.

Personal values and beliefs

9

1.2 Problem statement

The main objective of this research is to gather insights in the way ethnicity can be used as a predictor for COI and to what extent personal values and beliefs moderate this effect. As previously mentioned, there are still many gaps in the literature regarding the antecedents of COI. So, hopefully the insights based on the outcome of this research will contribute to the larger dilemma about standardization versus adaptation.

The following research objective is formulated:

How does ethnicity affect the Country-of-Origin Image, and is this effect moderated by personal values?

1.3 Research questions

In order to reach this objective the following research questions are formulated: 1. Does ethnicity affect COI?

2. Does cultural linkage of an ethnic group with the COO influence its COI? 3. Do personal values and believes moderate the effect of ethnicity on COI?

1.4 Relevance of the thesis

This section will explain the relevance of this thesis regarding its academic and managerial value.

1.4.1 Academic relevance

While international marketing research on the COI made significant contributions, numerous gaps remain, not the least being the lack of knowledge about subcultural influences on the image and evaluations of domestic and foreign products (Heslop et al., 1998). Even though Kaynak and Cavusgil already suggested in 1983 that “future studies should attempt to include regional and subcultural differences as these are major sources of variations”.

Academic research has recognized the importance and consequences of the COI effect in a variety of domains. However, the antecedents have received surprisingly little attention. The limited research on COI determinants focuses on the characteristics of the COO and the demographic characteristics of consumers. Balabanis et al. (2002) state that cultural differences seem to be a fruitful area to understand variations in COI. However, very little is known on how culture influences COI. Chattalas et al. (2008) support this finding and are also wondering why there is such a lack of investigation whereas COI has a strong impact on the marketplace.

10

and effects of cultural links on products and country evaluations. Some researchers (e.g. Heslop et al., 1998) are more specific in their suggestions. They recommend future research in the countries Belgium, Switzerland, Malaysia, and India because of their large and distinct subcultures whose members’ attitudes and behavior may differ so much from national norms that such norms might well be considered as meaningless as a basis for marketing action.

Therefore, this study wants to contribute to the existing literature and follows the recommendations of Heslop et al. (1998) by testing the influence of ethnicity on the COI in Malaysia. The presence of three dominant ethnic groups (the Malay, the Chinese, and the Indians) makes Malaysia a perfect field of application for this study. Thereby, this study is in line with the current shift in studies on the COI effect. Whereas conventional COI studies allow researchers to analyze if consumers prefer products or brands from one country in comparison to another, emphasis on the perceived images of the countries involved enables scholars to analyze why this is the case (Roth and Diamantopoulos, 2009).

1.4.2 Managerial relevance

In today’s markets, diverse customer needs and increasingly sophisticated marketing techniques have led to fewer situations where a mass marketing approach (standardization) is feasible (Wedel & Kamakura, 2000). Marketers should pay attention to specific segmentation and their profitability. Many multinationals, such as Coca-Cola, McDonald’s, IBM, and others, derive significant revenues from international operations. Therefore, the development of effective marketing strategies that are sensitive to cultural differences across countries is of considerable importance for success in the global marketplace (Gürhan-Canli & Maheswaran, 2000).

This research will contribute to the analyses of those subcultural differences that international marketers should perform before expanding into targeted countries. Subcultural differences should be taken into account in the development and implementation of marketing strategies. Because it is crucial not to underestimate the importance and potential buying power of different ethnic groups (Lindrigde and Dibb, 2002).

Understanding how (sub)culture may influence consumers’ evaluations of a country and its products will help practitioners, first, to better understand the contributing factors underlying a COI, and second, to better shape or manage the use of a product’s COI as a marketing tool (Balabanis et al., 2002). Especially, since foreign markets emerge, knowledge about COI in those markets will enable marketers to make wiser decisions (Zhang, 1996).

11

1.5 Outline

12

Chapter 2. Conceptual Framework

This chapter provides the theoretical background of the constructs and concepts that are used in this research. The background will help to create a better understanding of the constructs, variables and the relation between them.

2.1 Country-of-Origin Image (COI)

This study searches for insights regarding possible differences among ethnic groups in their perception towards a Country-of-Origin. The perception of this Country-of-Origin that is hold by consumers is called Country-of-Origin Image (COI). Nagashima (1970) defined COI as:

“the picture, the representation, the stereotype that businessmen and consumers attach to products of a specific country. This image is created by such variables as representative products, national characteristics, economic and political background, history and tradition”.

An example by provided Bikley and Nes (1982) is about a Puerto Rican shoe manufacturer who shipped his entire production to New York City and back and advertised those shoes as being originally from New York. Experience (and the COI effect) had convinced him that Puerto Ricans would buy this shoes more readily when they were perceived as made in New York rather than Puerto Rico.

13

In their meta-analysis Balabanis et al. (2002) summarize some of the previous research on COI. This overview in table 1.1 will create the context of this research.

Table 2.1 Findings of previous research on COI

Topic Authors Findings

Single-cue versus multi-cue COO effects

Wall et al., 1991; Johansson et al., 1985; Agrawal and Kamakura, 1999

The use of COO is mainly researched from an informational processing perspective; COO can be used to make inferences regarding quality. When COO is just cue among many, its effect on consumer attitudes will decrease.

Multiple COO cues (hybrid) effect

Hui and Zhou, 2003; Srinivasan et al., 2004

They found that country of brand origin accounting for the highest impact. Even more than country of manufacturing.

Product type effect Roth and Romeo, 1992; Kaynak andCavusgil, 1983

A number of studies found that COO effects on evaluation vary by type.

Consuming country and cultural orientation effects

Narayana, 1981; Heslop and Papadopoulos, 1993; Gurhan-Canli and Maheswaran, 2000

Due to different stereotypes among different countries, COO effects differ across consuming countries. They also found that collectivistic cultures have the tendency to consistently favour domestic over foreign products.

Consumer involvement / motivation effects

Gurhan-Canli and Maheswaran, 2000

When involvement is high, the motivation to engage in information processing is also high. They demonstrated that the COO is reduced under high-consumer involvement.

Consumer ethnocentrism effects

Shimp and Sharma, 1987 The literature suggests that consumers high in ethnocentrism generally pay more attention to the COO cue, often perceiving the consumption of imported products as socially undesirable and unpatriotic. National and country

stereotypes

Verlengh and Steenkamp, 1999 COO links a product to an associative network: -cognitive: COO as cue for quality

-affective: COO links products with emotional associations with a particular nation.

-normative: consumers may hold socially desirable behavioural norms linked to COO cues.

Development of the consumers

Lin and Sternquist, 1994 In less developed countries consumers have less abundant information and purchasing experience with foreign products. Therefore, the effect of COI play a less significant role in their attitude and behaviour.

2.1.1 Conceptualization of COI

In this section the academic conceptualization of COI will be discussed. Different definitions will be reviewed and the COI model that will be used in this research will be explained.

COI as psychological cue

14

Parameswaran et al., 1994) suggest that the COI construct should comprise (1) a cognitive component, which includes consumers’ beliefs about a particular country, (2) an affective component that describes the country’s emotional value to the consumer, and (3) a conative component, capturing consumers’ behavioral intentions with regard to the sourcing country. Some other researchers like Obermiller and Spagenberg (1998) replace the conative component by the normative dimension which emphasizes consumers’ perceived proximity to the norms and values associated with a COO.

Different categories of COI

Many researchers use different levels of aggregation in defining COI. As shown in figure 2.1 Roth and Diamantopoulos (2009) divided the definitions into three main categories. However, they do not suggest any interaction between the categories.

Figure 2.1 COI construct Roth & Diamantopoulos (2009)

The construct consists of the following three categories;

1. General Country Attributes (GCA); definitions of the (general) image of countries

The first group of definitions views country image as a generic construct consisting of generalized images created not only by representative products but also by the degree of economic and political maturity, historical events and relationships, culture and traditions, and the degree of technological virtuosity and industrialization (Allred et al., 1999; Bannister and Saunders, 1978; Desborde, 1990). While all of these factors refer to cognitive beliefs about a particular country, Askegaard and Ger (1998) and Verlegh (2001) are among the few researchers who explicitly also mention an affective component of country image which captures emotions and feelings about a particular country.

GCA

COI

15

2. General Product Attributes (GPA); definitions of the image of countries and their products The next category of definitions focuses on the image of countries in their role as origins of products. For example, Li et al. (1997) define country image as “consumers’ image of different countries and of products made in these countries.” This definitions implies that, first, country image and product image are two distinct (but related) concepts, and, second, that country images affect the images of products from that country.

3. Specific Product Attributes (SPA); definitions of the images of specific products from a country The last group of definitions focuses exclusively on the images of the products of a country and dates back to Nagashima (1970). However, although using the term country to specify the image object, Nagashima’s (1970) definition actually refers to the products of a particular country (Papadopoulos and Heslop, 2003). Thus, it is product image rather than country image that is actually captured by the definitional domain of the construct. Because the focus within this category is on the product rather than the country aspects, this category will not be relevant for this research and therefore excluded.

Parameswaran and Pisharodi (2002) designed a construct (figure 2.2) which also uses the three categories, however they suggest some interaction and dependability among the categories. The different categories of definitions are actually used as variables. This research model will be used in this research as starting point to test research statement.

Figure 2.2: Theoretical Structure of Country of Origin Image (Parameswaran & Pisharodi, 2002)

According to Papadopoulos and Heslop’s review, the vast majority of extant COI studies focuses on SPA and the number of studies who have in fact included country measures is extremely small. On the other hand, the meta-analysis of Roth and Diamantopoulos (2009) shows that the majority of the studies they reviewed employed ‘global’ rather than product category-specific measures.

GCA

IP SPA

GPA

16

Interaction between variables in the COI-construct

The basic theoretical model of COI (Figure 2.2) used in this research and in the research of Parameswaran & Pisharodi (2002) is actually based on the work of Yaprak and Parameswaran (1986). According to their conceptualization, the dependent variable, intention to purchase, is directly influenced by the specific product attributes (SPA) of a brand. Consumer purchase intention is also influenced by consumers’ general perceptions of products from a country (GPA) as well as perceptions of the source and its people (GCA). However, Parameswaran & Pisharodi (2002) state that the influences of GCA and GPA on IP are primary through their influences on consumers’ perception of the attributes of a particular product or brand (SPA).

To explain the working of the COI process I will use the example provided by Parameswaran & Pisharodi (2002); “According to this conceptual model, a US consumer’s purchase of a BMW car is influenced by how he or she perceives the features of the BMW car, which in turn, is affected by how he or she views German products as well as Germany and its people. The perception of German products (GPA), in turn, is influenced by perceptions of Germany and its people (GCA)”.

Exclusion of SPA in this research

Previous studies suggested that the impact of COI effect varies among different products (Zhang, 1996). For example, Kaynak and Cavusgil (1983) showed that COI biases existed between product classes such as electronics, food items, fashion merchandise and household goods. According to Eroglu and Machleit (1989) this fluctuation in effect is caused by the level of expertise with the product category. For example, Han (1989) found when familiarity with a particular country’s products is low, national stereotypes operate as a halo, allowing consumers to evaluate any unfamiliar product associated with that nation. This implicates that someone’s personal level of expertise regarding a specific product will cause some blur for the overall perception in this research towards the GCA and GPA of a country.

17

2.2 Relation between cultural background and COO

According to Nagashima (1970) COI is created by such variables as representative products, national characteristics, economic and political background, history and traditions. Undoubtedly, the characteristics of the COO are important determinants to its image, however Bar Tal (1997) added to the list the relations between the COO and the perceiver country as an important factor.

Bar Tal (1997) states that information about a country and its people can be obtained either through direct contact with the country through what he called transmitting mechanisms (TV, radio, magazines, newspapers, art, theatrical plays, school books film, etc.). Information about a COO coming from the perceiver country sources is usually tinted by ideology, beliefs and goals and, in most cases, reflects the prevailing stereotypes about the country (Bar Tal, 1997).

Direct contact (through visits) provides the opportunity for first-hand experience with the country and expands the local reality to which a consumer is exposed. This may help consumers to reconsider some of the myths surrounding the country and see countries and their products under a different light. Thus, the higher the level of direct contact, the more objective consumer perceptions about a country will be.

Fluency in a COO’s language can result in more objective evaluations of the country by enabling access to additional information sources residing outside one’s country and by facilitating direct contact. Language fluency may amplify the effects of direct contact on consumers’ perception

For example, it has been demonstrated that consumers’ willingness to purchase a product is related to economic, political, and cultural characteristics to the product’s country of origin (Parameswaran & Pisharodi, 1994). Additionally, COI is affected by the consumer’s perception of similarity between his or her own country and the origin country’s political and cultural climate and belief systems (Han, 1990).

18 2.3 The use of ethnicity as proxy for subculture

Whereas many researches lack a clear definition of culture in its use as variable or antecedent, this research tries to overcome this limitation by the provision of a thorough definition and precise conceptualization and operationalization.

2.3.1 Conceptualization of (sub)culture

Poortinga (1992) defined culture “as a set of shared constraints that limit the behaviour repertoire available to members of a certain group”. He identified two types of constraints:

-external constraints: which refer to ecological, social, economic, historical and political conditions of a group, and

-internal constraints: which refer to culturally transmitted values and beliefs.

A general definition of subculture will be provided by Lenartowicz and Roth (2001); “A subculture has been defined as a subdivision of a national culture, composed of a combination of social situations such as class status, regional, rural, or urban residence, religious affiliation and ethnic background, that together from a functional unity which has an integrated impact on participating individual. The distinct shared history and values of a subcultural group may influence their consuming patterns and behavior (Laroche et al., 2002).

There is a large debate regarding the appropriate definition of (sub)culture, and many social researchers have recognized the complexity and multi-dimensionality of the concept (Balbanis et al., 2002).

2.3.2 Operationalization of (sub)culture

Nationality is drawing some boundaries around individuals belonging to a large group. It is related to the operational nature of culture. It is likely that, historically, shared culture has been a fundamental building block in the progressive construction of modern nation-states. Nevertheless an attempt by Usunier (1993) to equate culture directly with the nation-state, or country would be misguided for a number of reasons:

-A country culture can only be defined by reference to other country cultures. India is a country culture in comparison with Italy or Germany, but the Indian subcontinent is made up of highly diversified ethnic and religious groups and languages.

19

-Political decisions, especially during the last century, have imposed the formation of new states, particularly through the processes of colonization and decolonization. The fixing of borders of these new states often showed little respect for cultural realities.

These examples also show a clear link with many cross-cultural research which pretends to do research among different cultures, where in fact they conduct a cross-national study based on national borders.

National/cultural territories with specific, recognized borders are rarely fully homogeneous. The transition from one to another is often facilitated by ‘cross-border’ cultures.

The national element seems not always the (only) source of culture when regarded from ‘operational culture’ perspective. Usunier (1993) identified many different sources of cultural background at the level of an individual: language, nationality, education, profession, ethnicity, religion, family, sex, social class, or corporate culture. The list of sources implicates that (sub)culture seems very hard to grasp and to measure in research. Therefore, researchers should be very precise and careful in their choice for a proxy of culture used in their study. In this research ethnicity will be used as source and proxy for subculture.

2.3.3 Conceptualization of ethnicity

Ethnicity has been best defined within cultural anthropology, but it has been a debated topic which results in the lack of a single definition or theory of how ethnic groups are formed (Baumann, 2004). This research will use the six main features provided by Hutchinson and Smith (1996) which results in the distinction between ethnic groups:

1. a common proper name, to identify and express the ‘essence’ of the community. 2. a myth of common ancestry that includes the idea of common origin in time and place. 3. shared historical memories, or better, shared memories of common past.

4. one or more elements of common culture 5. a link with a homeland

20 2.4 Individualism-Collectivism

2.4.1 Hofstede dimensions

Many scholars and researchers discuss the choice of dimensions most appropriate for conceptualizing and operationalizing culture. However, Hofstede’s framework is the most widely used national cultural framework in psychology, sociology, marketing or management studies. Hofstede used 116.000 questionnaires from over 60.000 respondents in seventy countries in his empirical study. He created six dimensions, and linked the dimensions with demographic, geographic, economic, and political aspects of society, a feature unmatched by other frameworks (Soares et al. 2007). Although the original framework exists of four dimensions, in 1991 a fifth and in 2010 a sixth dimension were added. The six dimensions developed by Hofstede will now be shortly described in table 2.2 based on the description Hofstede provides on his website (geert-hofstede.com).

Table 2.2: Hofstede’s six cultural dimensions

Dimension Description

Individualism vs. collectivism Individualism can be defined as a preference for a loosely-knit social framework in which individuals are expected to take care of themselves and their immediate families only. Collectivism represents a preference for a tightly-knit framework in society in which individuals can expect their relatives or members of a particular in-group to look after them in exchange for unquestioning loyalty.

Uncertainty avoidance The uncertainty avoidance dimension expresses the degree to which the members of a society feel uncomfortable with uncertainty and ambiguity. The fundamental issue here is how a society deals with the fact that the future can never be known: should we try to control the future or just let it happen?

Power Distance This dimension expresses the degree to which the less powerful members of a society accept and expect that power is distributed unequally. The fundamental issue here is how a society handles inequalities among people.

Masculinity vs. Femininity The masculinity side of this dimension represents a preference in society for achievement, heroism, assertiveness and material reward. Femininity stands for a preference for cooperation, modesty, caring for the weak and quality of life. Long-term vs. short-term orientation

(added in 1991)

Societies with a short-term orientation generally have a strong concern with establishing the absolute truth. They exhibit great respect for traditions, a relatively small propensity to save for the future and a focus on achieving quick results. Long-term oriented societies believe that the truth depends very much on situation, context and time. They show an ability to adapt traditions to changed conditions, a strong propensity to save and invest, thriftiness, and perseverance in achieving results.

Indulgence vs. restraints (added in 2010)

21

2.4.2 Conceptualization Individualism-Collectivism

Among all Hofstede’s dimensions the individualism-collectivism dimension is crucial to the cross-cultural understanding of consumer behavior (Maheswaran and Gudykunst & Ting-Toomey, 1988; Triandis et al., 1988) and is perhaps one of the most significant ways in which societies and cultures differ. In fact, more than one third of published studies in cross-cultural research cited individualism-collectivism as at least a partial explanation of observed differences across cultures (Hui & Yee, 1994).

The discussion of individualism can be traced in Sophists’ teaching (1776) and in the ideas of Adam Smith (1949), whereas the collective themes can be found in Plato’s Republic (1772) and in Jean-Jacques Rousseau’s Du Contrat Social (1954). In more recent social science literature, credit for introducing the concept to the study of cross-cultural differences is usually assigned to Hofstede (1980), who identified Individualism as one of four (since 2010 six) primary dimensions of cultural variation in organizational values and practices (Brewer & Chen, 2007).

Hofstede (1980) described individualism-collectivism as the relationship between the individual and collectivity that prevails in a given society. In individualistic cultures, individuals tend to prefer independent relationships to others and to subordinate ingroup goals to their personal goals. Individualist societies emphasize “I” consciousness, autonomy, emotional independence, individual initiative, right to privacy, pleasure seeking, financial security, need for specific friendship, and universalism. In collectivistic cultures, on the other hand, people are more likely to have interdependent relationships to their ingroup goals. Collectivist societies stress “we” consciousness, collective identity, emotional dependence, group solidarity, sharing, duties and obligations, need for stable and predetermined friendship, group decision, and particularism (Hofstede, 1980; Han & Shavit, 1994).

22

scores around the world. The differences between individualists and collectivists could also be a partial explanation of my personal experience of fellow students cooperating with Asian students. One of the frustrations among western students is that Asian students are very reserved in expressing their feelings or sharing their opinion. This seems to be in line with the findings of Kramer et al. (2007) that collectivistic (vs. individualistic) people are more likely to act in accordance with the expectations of others to express their interdependence and connectedness, rather than in accordance with their individual preferences.

Furthermore, Iyengar and Leper (1999) find that Asian-American children have greater intrinsic motivation and perform best in situations in which choices are made for them by relevant in-group members, but for Anglo American children motivation and performance are best when they make their own choices.

According to Hofstede’s measurements Malaysia is a collectivistic country. For example, when we compare Malaysia with the United States; United States is highly individualistic with a score of 91 on a 100-point individualism scale, whereas Malaysia is clearly on the collectivistic side with a score of 26. In addition to variations across cultures, people within a given culture may vary in the extent to which they are collectivist or individualist in their orientation (Gürhan-Canli & Maheswaran, 2000). So, besides possible differences between COI among the ethnical groups, this relationship possibly be moderated by the influence of differences between I-C within these groups.

23

2.5 Conceptual Model

Based on the literature review a conceptual model is constructed, which will be tested in Malaysia (see section 3.2). As shown in figure 2.3 the scope of this research mainly focuses on the antecedents of GCA and its indirect effect on GPA. Ethnicity will be used as proxy for subculture as predictor for GCA. I-C will function as moderator in this relationship.

Also age and gender are added as variables to check their predicting value on GCA. Wall and Heslop (1986), Good and Huddleston (1995) and Lawrence et al. (1992), for example, established that gender plays an important role in the perception of ‘made-in’ images with females having a higher COO bias against foreign products and in favour of domestic ones. With regard to age, Schooler (1971) found that older consumers evaluate foreign products less favourably than younger ones.

Figure 2.3: Conceptual Model

In the set of hypotheses related to this conceptual model, a distinction will be made between two types of hypotheses:

1. Hypotheses that are used in order to prove the existence relationship regardless of the direction of this relation. For those hypotheses the direction of the relation is not relevant or the direction cannot be predicted based on any literature or logic.

-H1a: Ethnicity has an effect on the perceived GCA.

This hypothesis will be used to find the existence of the relation between ethnicity and GCA. Therefore the direction of the relationship or the ethnic group which will have the most positive GCA is not relevant. Ethnicity _GCA I-C Gender Age VC HC VI HI GPA GCA: General Country Attributes GPA: General Product Attributes SPA: Specific Product Attributes IP: Intention to Purchase I-C: Individualism vs. collectivism

H 2d H 2a H 2b H 2c

H 1a, b _{H 5}

24

-H3: Gender has an effect on the perceived GCA. -H4: Age has an effect on the perceived GCA.

Although literature suggests that women and younger people possess a more positive COO bias than men or older people, this relation is only tested in comparison with a domestic country (ethnocentrism). Therefore, no assumptions can be made with respect to the direction of the relation in the setting of only a single COO.

2. Hypotheses that are used to find the existence of a relationship and its direction. In contrast to the first type of hypotheses, grounded assumptions can be made about the direction of the relationship.

-H1b: Cultural linkage with the COO will have a positive influence on the perceived GCA.

Previous research showed that people who have a cultural linkage with the COO hold a more positive bias towards its people and products.

-H2a: Horizontal-Individualism has a negative influence on the relation between ethnicity and GCA. -H2b: Vertical-Individualism has a negative influence on the relation between ethnicity and GCA. -H2c: Horizontal-Collectivism has positive influence on the relation between ethnicity and GCA. -H2d: Vertical-Collectivism has a positive influence on the relation between ethnicity and GCA. These hypotheses are based on the studies that indicate that individualistic oriented people are more open and positive towards foreign countries, cultures and people.

H5: GCA will have positive effect on GPA.

25

Chapter 3. Research Design

This chapter outlines the research methods that are used to provide an answer on the problem statement and the related research questions. The research design and data collection will be outlined below. Finally, the plan of analysis will be stated.

3.1 Field of application

To test the conceptual model and search for possible cultural differences Malaysia is chosen as field of application. In Malaysia three ethnic groups (Malay, Chinese, and Indians) live together within the same national boundaries.

The Malays form the largest community (50.4%)1 and play a dominant political role. They make up about half of the total population. By constitutional definition, Malays are Muslims who practice Malay customs and culture. The second largest ethnic group is the Chinese who make up 24.6% of the total population. They have been dominant in trade and business since the early 20th century. The Indian community in Malaysia is the smallest of the three main ethnic groups, comprising 7.1% of the population. Indians began migrating to Malaysia in the early 19th century. Besides these three main ethnic groups, a minority of Malaysians do not fit into these broader ethnic groups. A small population exists of people from European and Middle Eastern descent. Most of these ‘Eurasian’ Malaysians trace their ancestry to British, Dutch and/or Portuguese colonists. Although, this mixed minority is part of the Malaysian society, this group is small and therefore, will not fall within the scope of this research.

Choice of Country-of-Origin

Since the Chinese are one of the ethnic groups in this research and China is one of Malaysia’s biggest import partners (www.economywatch.com), China will be used as COO. A comparison can be made between the three Malaysian ethnicities in their perception towards China and its products.

The use of ethnicity as proxy for subculture in Malaysia

Based on personal observation ethnicity seems one of the binding factors in Malaysia. For example the Chinese, the majority speaks the same language, loves the same food, practices the same religion and most of the time hangs around with Chinese friends.

So, there existing many sources that can be used as base for operationalization of subculture. In this research one of the external constraints will be used for the conceptualization of subculture.

1

26

Ethnicity will be the demographic characteristic that will be used as representation of the external constraints. In other words, ethnicity will be used as the proxy for subculture in this research. Here I will follow the most common ethnic division in Malaysia, that is also used by governmental institutions; Malay, Chinese and Indian. The choice to take ethnicity is supported by several arguments. First, people in Malaysia are used to be asked about their ethnicity. For example, on government documents ethnicity is a standard demographic question. Therefore, asking for someone’s ethnicity is no issue at all. A second argument is my observation that people with the same ethnicity generally practice the same religion, hang out with friends of the same ethnicity and eat food which is related to their ethnical background. Therefore, ethnicity seems to be some kind of an overall ‘binding factor’. The third reason is more practical, for handing out the questionnaire it is easy to recognize people based on their ethnicity. This was necessary in order to get an equal amount of respondents for each ethnical group.

3.2 Research Method

The type of research that is conducted to answer the problem statement is characterized by Malhotra (2007) as a descriptive conclusive research. A conclusive research is based on a large, representative samples, and the data obtained are subjected to quantitative analysis. Moreover, the analysis in such research designs should be based on quantitative data and the results of the analysis will lead to draw conclusions and recommendations (Malhotra, 2007). Descriptive research is characterized by the prior formulation of specific research questions.

Online questionnaire

For the collection of quantitative data an online questionnaire is used. This questionnaire consisted of three main parts; demographics, COI towards China, and the personal values. In appendix 5 an overview of the total lay-out of the questionnaire can be found. All the questions regarding the COI and personal values are taken from the constructs as will be described in sections 3.2.1 and 3.2.2 about the operationalization of the variables.

For the design of the questionnaire www.thesistools.com is used. By using an online questionnaire people could fill in the questionnaire at a moment that was most convenient for them. An incentive is used in order to encourage people filling out the online questionnaire. Every respondent had a chance to win 100 Malaysian Ringgit (RM) (€25,-) after submitting the questionnaire.

27

3.2.1 Population

The target group of this research consists of all the Malaysian inhabitants that belong to one of the three ethnic groups. Although the requests for filling out the questionnaire is done by e-mail and hard copy flyers, all the respondents could only participate in the questionnaire by using the online link.

Because of the limited time and financial resources the judgemental sampling method is used in order to collect as much data as possible. Judgemental sampling is a form of convenience sampling in which the population elements (ethnicity in this case) are selected based on the judgement of the researcher (Malhotra, 2007). For this research respondents that represented one of the three ethnic groups in Malaysia were requested. Therefore a strategy was set up to find enough people to represent each ethnic group. Two methods are used for the distribution of the questionnaire: sending e-mails and distributing flyers.

E-mail

- sending out an e-mail with a link to the questionnaire to my colleagues at DHL: most of them were Chinese. Many of them also asked some relatives to help me. (appendix 4)

Flyers

Flyers (appendix 5) are distributed on different strategic places.

- handing out flyers at the entrance and canteen of a state owned company: most of the employees of state owned companies in Malaysia are Malay.

- handing out flyers at the university campus: most of the students of universities in Malaysia are Malay.

- handing out flyers at the entrance of an apartment complex: the apartment was in an Indian neighbourhood. So, many of the Indian respondents did use this flyer.

28

3.2.2 Sample

Different methods are used in order to find enough respondents to fill out the questionnaire. Like already mentioned the data was collected by sending emails and handing-out flyers with a link to the online questionnaire. In total 500 flyers were handed out in Kuala Lumpur and around 60 e-mails are sent to my colleagues at DHL. In total 144 people responded. This leads to a response rate of 25,7%. After studying the results the data showed that 3 of the respondents do not have the Malaysian nationality and 10 of the respondents did not complete the whole questionnaire. So, 13 respondents are excluded, which remains 131 respondents for the analyses and a netto response rate of 23,4%. Table 3.1 provides an overview of the demographic characteristics of the respondents.

29

3.2 Operationalization

This section will elaborate on the theoretical constructs that are used for the operationalization of GCA, GPA, and I-C.

The main scaling technique used for the online questionnaire is a 5-point Likert scale, ranging from “strongly agree” to “strongly disagree”. In addition to the items mentioned above, two variables in the questionnaire have a nominal measurement scale (ethnicity and gender) and for age a ordinal scale is used.

For the measurement of the three main variables (GCA, GPA, and IC) standard constructs from the literature are used. Even though the constructs are test in previous research, different results may occur based on cross-cultural differences in response style (Chen et al. 1995). The majority of constructs in the literature is based on the results of tests in Western countries. Therefore, a Cronbach Alpha test is performed on all the three constructs to assess the internal consistency among the items of the constructs. The coefficients will vary from 0 to 1, and a value of 0,6 or less generally indicates unsatisfactory internal consistency reliability (Malhotra et al., 2010)

3.2.1 Operationalization of COI

Parameswaran and Pisharodi (1994) showed that subsequent empirical studies determined that several of the hypothesized environmental factors, spanning hundreds of items, had negligible to minimal effect on product perceptions. Those findings led to the current conceptualization of COI. A review of the literature indicates that a variety of scales, differing greatly in complexity, have been used to capture a country’s image. We conducted a study to clarify the underlying facets of COI scale that was first articulated by Yaprak and Parameswaran (1986) to measure a three faceted COI based on extant theory. Although, the their construct measures three elements (GCA, GPA, and SPA) of COI, this study will only use two of them (GCA and GPA).

30

product, marketing, and firm-goodwill-related attributes). As mentioned before and shown in figure 2.1 these three facets (GCA, GPA, and SPA) represent the current conceptualization of COI.

Table 3.2: Scale items used for measurement COI

Perceptual constructs Items

Alpha if item deleted A. General Country Attitudes

(GCA) 1. People in China are well educated ,786

2. China emphasis on technical training ,776

3. People in China are hard working ,769

4. People in China are creative ,781

5. People in China are friendly and likeable ,785

6. Technical skills of the Chinese workforce are high ,768 7. China is friendly toward Malaysia in international affairs. ,780 8. People in China are motivated to raise living standards ,746 9. People in China are proud to achieve high standards ,775 B. General Product Attitudes

(GPA) 1. China produces highly technical products ,561

** 2. Products from China are unreasonably expensive ,706

3. China is known for ‘luxury’ products ,600

4. Chinese products are made with precise workmanship ,543 5. Chinese products are imitations, not innovations ,636 ** 6. Chinese products are distributed worldwide ,657 7. Products from China are concerned with performance, not appearance ,582

8. Chinese products need frequent repairs ,625

9. Chinese products are marked in a wide range of styles ,629

10. Products from China are long-lasting (durable) ,546

11. Advertising of Chinese products is informative ,596

12. Products from China are hard to service here in Malaysia ,624 13. Chinese products are cheaply made consumer items ,587 ** Items deleted

Papadopoulos, Heslop, and Beracs (1990) claim that inclusion of scale items relating to origin countries and their people is rare in country image studies and should be viewed as an improvement. According to Roth and Diamantopoulos (2009) the best way to conceptualize the COI construct is from an attitude-theoretic perspective. As such, “country of origin not merely a cognitive cue for product quality, but also relates to emotions, identity, pride and autobiographical memories” (Verlegh et al., 1999). Table 3.2 shows the scale of Parameswaran and Yaprak (1987) that is used for the operationalization of GCA and GPA in this research. The scale consists of a mix of cognitive and affective emotional items.

Reliability

31

Computing the Cronbach Alpha for the General Product Attributes without deleting any of the items results in a score of 0,63. Although this alpha is more than the required 0,6 deleting two items will increases the value to 0,714. The items that were deleted are:

-Products from China are unreasonably expensive.

A possible explanation for the low consistency of this statement in relation to the other items can be the word ‘unreasonably’. Respondents maybe assess this formulation as too extreme.

-Chinese products are distributed worldwide.

This is a quite neutral statement. It seems not very clear whether this statement has a positive or negative contribution to the overall score.

3.2.2 Operationalization of the Individualism-Collectivism dimension

32 Table 3.3: Output Factor Analysis I-C dimensions & Cronbach Alpha’s

Rotated Component Matrix

Component Alpha

Items 1 2 3 4

HI 1. I'd rather depend on myself than on others _{,832 ,025 ,178 ,010}

,835 HI 2. I rely on myself most of the time: I rarely rely on others _{,883 ,088 ,047 ,021}

HI 3. I often do 'my own thing' _{,830 ,057 -,122 ,056}

HI 4. My personal identity, independent of others, is very

important ,660 ,108 ,013 ,374

VI 5. It is important that I do my job better than others _{,546 ,150 ,120 ,568}

,685

VI 6. Winning is everything _{,229 -,130 ,160 ,715}

VI 7. Competition is the law of nature _{,003 ,038 ,125 ,817}

VI 8. When another person does better than I do, I get tense

and aroused ,014 -,012 -,134 ,646

HC 9. If a co-worker gets a prize, I would feel proud _{,006 ,792 ,066 -,058}

,776 HC 10. The well-being of my co-workers is important for me _{-,026 ,847 ,170 -,040}

HC 11. To me, pleasure is spending time with others _{,217 ,621 ,210 -,022} HC 12. I feel good when I cooperate with others _{,078 ,692 ,350 ,025} VC 13. Parents and children must stay together as much as

possible -,016 ,208 ,815 ,030

,786 VC 14. It is my duty to take care of my family, even when I

have to sacrifice what I want ,053 ,306 ,816 ,125

VC 15. Family members should stick together, no matter

what sacrifices are required ,061 ,261 ,698 -,008

VC 16. It is important to me that I respect the decisions made

by my groups ,126 ,616 ,401 ,147

Summarizing; horizontal individualists (HI) want to be unique and distinct from groups. Vertical individualists (VI) often want to become distinguished an acquire status, and they do this in individual competitions with others. Horizontal collectivists (HC) see themselves as being similar to others and emphasize common goals with others, interdependence, and sociability, but they do not submit easily to authority. Vertical collectivists (VC) emphasize the integrity of the in-group, are willing to sacrifice their personal goals for the sake of group goals, and support competitions of their in-group with out-in-groups.

To test their original scale of 27 items they conducted studies; in the US and Korea. They performed factor analyses and table 3.3 shows the final scale with the highest loadings. Because the construct of Triandis and Gelfand (1998) is also tested in Asia, it seems even more suitable as construct in this research.

Reliability

33

Cronbach Alpha’s for each sub-dimension is calculated. All this to be sure the scale of Triandis and Gelfand (1998) would be a reliable measure for this research.

Table 3.3 shows the output of the Factor Analysis and the Cronbach Alpha scores. Both show enough internal consistency reliability, whereas all the Alpha coefficients are higher than 0,6. Most of the factor loadings show a clear correlation between the factors (item) and the variables (sub-dimension). Despite some single high cross-loadings the original construct will not be changed.

3.3 Plan of Analysis

In order to test the conceptual model two regression analyses will be performed. However, before the data is ready for the analysis some preparations need to be done.

3.3.1 Recoding of items

Before the individual items can be transformed into one variable some of the items have to be recoded. Not all the items are formulated in the same direction yet. Therefore, items B 2, 5, 8, 12, and 13 (appendix 2) are recoded into the opposite direction. Therefore, for all items a higher rating means the more positive people are towards Chinese products.

3.3.2 Computing new variables

After recoding and deleting some of the items, the variables (GCA,GPA, and IC) that will be needed can be computed for testing the conceptual model and performing the linear regression. The values of the variables are calculated by taking the mean of all the remaining items per construct for every respondent.

Dummy variables

Because some of the variables (ethnicity, gender, and age) in the conceptual model are categorical they should be transformed into dummies before they can be inserted in the regression analyses. -Ethnicity (Malay, Chinese, and Indian) is transformed into two (K-1) dummy variables with the Malay group as the reference group.

-Gender (male and female) changes into one (K-1) dummy with the male group as reference.

34

Interaction / Moderation

Besides the transformation of the existing variables, some new variables are created in order to test the interaction effect in the conceptual model. The interaction variables are easily computed by multiplying the dummy variables of ethnicity with the IC sub-dimensions (see table 3.6).

3.3.3 Multicollinearity

The overall conceptual model will be tested by using a linear regression. Before we will focus on the results the model and related variables are tested for multicollinearity. Malhotra et al. (2012) state that virtually all multiple regression analyses done in marketing research involve independent variables that are related. Multicollinearity, however, arises when intercorrelations among the independent variables are very high. One of the most relevant problems that is caused by multicollinearity is the difficulty to assess the relative importance of the independent variable in explaining the variation in the dependent variable. Appendix 3 shows the scores of the tolerance and VIF which indicate the multicollinearity.

The variance inflation factor (VIF) provides an index that measures how much the variance of an estimated regression is increased because of collinearity. In a tolerance of less than 0,5 or less indicated a multicollinearity problem, a tolerance of 0,30 or less indicates a serious multicollinearity problem (Janssens et al. 2008).

A VIF of 5 or 10 above indicates a multicollinearity problem. The columns with the unstandardized tolerance and VIF scores show that especially the ethnicity dummies and the interaction variables cause multicollinearity.

35

3.3.4 Use of three models

Table 3.6 shows the three models that will be used in the regression. By adding new variables in every model a contribution can be discovered. The first model only uses the most important variables, Ethnicity and IC sub-dimensions. The second model adds the two Interaction variables, and in the third model the Gender and Age dummies are added.

Table 3.4: Deviation into three models

Models Variable In regression formula 1 2 3 (Z)Eth1Dummy_Malay _X1a X X X (Z)Eth2Dymmy_Chinese _X1b X X X (Z)HI X2a X X X (Z)VI _X2b X X X (Z)HC _X2c X X X (Z)VC X2d X X X Interaction_(Z)Malay_(Z)HI _X1a * X2a X X Interaction_(Z)Malay_(Z)VI _X1a * X2b X X Interaction_(Z)Malay_(Z)HC X1a * X2c X X Interaction_(Z)Malay_(Z)VC _X1a* X2d X X Interaction_(Z)Chinese_(Z)HI _X1b * X2a X X Interaction_(Z)Chinese_(Z)VI _X1b * X2b X X Interaction_(Z)Chinese_(Z)HC X1b * X2c X X Interaction_(Z)Chinese_(Z)VC _X1b * X2d X X (Z)GenderDummy_male _X3 X (Z)AgeDummy_young_19_34 X4 X

36 X1a X1b X2a X2b X2c X2d

3.3.5 Use of regression analysis

The first analysis will consist of simple linear regressions, or bivariate regression, in order to measure the effect of the independent variable on the dependent variable. In the first simple regression analysis the dependent variable is GCA and the independent variables are ethnicity, age, and gender. The moderator that will be tested is IC. The second regression will test the relation between GCA as independent variable and GPA as dependent variable.

As outlined in the conceptual model, the construct is subdivided into separate steps, which will also be analyzed through bivariate linear regressions. The equation models for these regressions are formulated as follows;

Ŷ (General Country Attributes) =

α + β

∗

X

+ ε

α + β

∗

X

+ ε

α = regression intercept βi = regression coefficient

Xi = independent variable

εi = error term

Ŷ (General Country Attributes)

= α + β

∗ + β

∗ + β

∗ + β

∗

X

+ β

∗

X

+ ε

Table 3.4 already provided an overview of which variables are included per model.

Ŷ (General Product Attributes)

=

α + β

∗

X

In this formula

X

Based on the outcome of the regressions the significant relations will be used in chapter 4 to construct the final regression formulas.

37

3.3.6 Influence of cultural linkage on COI

38

Chapter 4. Results

4.1 Regression Analysis

In this chapter the results of the empirical research will be discussed. Section 4.1 will describe the characteristics of the sample group. Then section 4.2 will continue with the explanation of the data preparation. Finally, section 4.3 will conclude with testing the conceptual model.

Table 4.1: Variable overview

Variable Operationalization Function

GPA (Z)GPA dependent variable

GCA (Z)GCA (in)dependent variable

Ethnicity (Z)Eth1Dummy_Malay independent variable

(Z)Eth2Dummy_Chinese independent variable

I-C

(Z)HI moderator

(Z)VI moderator

(Z)HC moderator

(Z)VC moderator

Gender (Z)GenderDummy_male independent variable

Age (Z)Age_Dummy_young_19_34 Independent variable

Interaction (Z)Interaction1_Eth1_IC interaction (Z)Interaction2_Eth2_IC interaction

39

The R-square is the squared correlation index that indicates the proportion of variance of the optimally scaled data. It varies between 0 and 1 and signifies the proportion of the total variation in Y that is accounted for by the variation in X. Table 3.6 shows that the independent variables have a relatively low predictive value for the dependent variable.

The results in table 4.2 show that only effect of the (Z)Eth1Dummy_Malay show a significant result of a 0.10 significance level. This implicates that Malay people compared to Indian have a less positive General Country Image towards China. Furthermore, there does not seems to be any significant influence ofI-C, Gender or Age.

Table 4.3: Output regression GPA as dependent variable

Variable In regression formula Beta Constant ,007 (Z)GCA _X1a *** -,217 Level of significance *=.90, **=.95, ***=,99

Based on the outcome in table 4.3 there seems to be significant relation between GCA as a predictor for GPA. Malaysian people who have a positive image towards China and its inhabitants (GCA) will also hold this positive image for products in general from China (GPA). This relation supports the findings (figure 2.2) of Parameswaran & Pisharodi (2002).

Based on the output of the regression analyses in the tables 4.3 and 4.4 two equations can be formulated. Only the independent variables that are significant (.10) can be included.

Y (General Country Attributes) = 0,18 + -0,213 *

X

40 4.2 Relation between cultural linkage and COI

By performing an ANOVA possible evidence can be found that will support previous research that indicates that there exists a positive relation between ethnicity and COO. In this research this would suggest that Chinese people in Malaysia will have a more positive image towards China based on their cultural linkage.

Table 4.4: Differences among ethnical groups

GCA GPA I-C HI VI HC VC total sample (n=131) 2,688 3,441 2,262 2,617 1,827 1,715 Malay (n=38) 2,548 3,460 2,368 2,572 1,783 1,691 Chinese (n=53) 2,683 3,340 2,236 2,577 1,880 1,779 Indian (n=40) 2,830 3,556 2,194 2,713 1,81 1,656 Significance level 0,074* 0,103 ,571 ,603 ,718 ,638 Level of significance *=.90, **=.95, ***=,99

The output in table 4.4 shows a significant difference between the ethnic groups in their scores on GCA. People belonging to the Malay ethnic group have the most positive image towards China and their inhabitants. The differences between the scores on GPA (which are almost significant) indicate that the Chinese in Malaysia value the product from China in general the most.

41

Ethnicity

Malay, Chinese, Indian

GCA China I-C Gender Age VC HC VI HI GPA China

GCA: General Country Attributes GPA: General Product Attributes SPA: Specific Product Attributes IP: Intention to Purchase I-C: Individualism vs. collectivism

H 2d - H 2a + H 2b + H 2c - H 1a H 1b + H 5 ₊ H 4 H 3 4.3 Conceptual model

The conceptual model as showed in figure 4.1 is tested by using a linear regression. Like mentioned in chapter 3 the analysis will take place in two steps. First, the regression analysis with GCA as the dependent variable will be performed. Second, an analysis with GCA functioning as the independent and GPA as the dependent variable.

Figure 4.1: Conceptual model

-H1a: Ethnicity has an effect on the perceived GCA.

-H1b: Cultural linkage with the COO will have a positive influence on the perceived GCA.

-H2a: Horizontal-Individualism has a negative influence on the relation between ethnicity and GCA. -H2b: Vertical-Individualism has a negative influence on the relation between ethnicity and GCA. -H2c: Horizontal-Collectivism has positive influence on the relation between ethnicity and GCA. -H2d: Vertical-Collectivism has a positive influence on the relation between ethnicity and GCA. -H3: Gender has an effect on the perceived GCA.

-H4: Age has an effect on the perceived GCA. -H5: GCA will have positive effect on GPA.