Does the country of origin effect apply to brands originating in
one of the BRICS countries?
A quantitative study focuses on the influence of the country of origin effect on brand equity of brands originating in one of the BRICS countries
Master Thesis
University of Amsterdam – Economics & Business MSc. Business Administration – International Management 1st supervisor: E. Dirksen MSc.
2nd supervisor: Dr. J.P. Lindeque
Student: Silke de Ruiter
Student number: 10982574
1 Statement of originality
This document is written by Silke de Ruiter who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
2 Table of contents
ABSTRACT ... 4
1. INTRODUCTION ... 5
2. LITERATURE REVIEW ... 8
2.1. COUNTRY OF ORIGIN EFFECT ... 8
2.2. DIMENSIONS OF THE COUNTRY OF ORIGIN EFFECT ... 11
2.2.1. BRAND ORIGIN ... 12
2.2.2. COUNTRY OF MANUFACTURE ... 13
2.2.3. MACRO AND MICRO COUNTRY IMAGES ... 13
2.3. BRAND EQUITY ... 14
2.4. THE INFLUENCE OF COUNTRY OF ORIGIN EFFECT ON BRAND EQUITY ... 17
2.4.1. THE COUNTRY OF MANUFACTURE ON PERCEIVED BRAND QUALITY ... 17
2.4.2. BRAND ORIGIN ON PERCEIVED BRAND QUALITY ... 19
2.4.3. BRAND ORIGIN ON BRAND IMAGE ... 21
2.5. THE COUNTRY OF ORIGIN EFFECT ON BRANDS ORIGINATING FROM BRICS COUNTRIES .... 22
2.6. BRAND TYPICALITY ... 25
3. RESEARCH GAP ... 27
4. DATA AND METHODOLOGY ... 29
4.1. RESEARCH ... 29 4.1.1. DATA COLLECTION ... 29 4.1.2. SAMPLE ... 30 4.1.3. SELECTED BRANDS ... 30 4.2. MEASURES ... 31 4.2.1. DEPENDENT VARIABLES ... 32 4.2.2. INDEPENDENT VARIABLES ... 32 4.2.3. MODERATING VARIABLE ... 32 4.2.4. CONTROL VARIABLE ... 33 5. RESULTS ... 34 5.1.1. DATA PREPARATION ... 34 5.1.2. DEMOGRAPHIC STATISTICS ... 34 5.1.3. RELIABILITY ANALYSIS ... 36 5.1.4. CORRELATION ANALYSIS ... 37
5.2. HYPOTHESIS TESTING –HIERARCHICAL REGRESSION ... 38
5.2.1. HYPOTHESIS TESTING –HYPOTHESIS 1A ... 38
5.2.2. HYPOTHESIS TESTING –HYPOTHESIS 1B ... 46
5.2.2. HYPOTHESIS TESTING –HYPOTHESIS 2 ... 51
5.3. HYPOTHESIS TESTING –MODERATOR ... 56
5.3.1. HYPOTHESIS TESTING –HYPOTHESIS 3A ... 57
5.3.2. HYPOTHESIS TESTING –HYPOTHESIS 3B ... 58
5.4. SUMMARY RESULTS ... 59
6. DISCUSSION ... 61
6.1. FINDINGS ... 61
6.2. THEORETICAL CONTRIBUTIONS ... 66
3 6.4. LIMITATIONS ... 68 6.5. FUTURE RESEARCH ... 69 7. CONCLUSION ... 71 8. BIBLIOGRAPHY ... 73 9. APPENDIX ... 79 9..1 Constructs ... 79
9.2. Most important output... 81
9.2.1. Correlation matrix ... 81
9.2.2. Hierarchical regression analysis ... 83
9.2.2.1 Hypothesis 1a overall model ... 83
9.2.2.2. Hypothesis 1a individual brands ... 84
9.2.2.3. Hypothesis 1b overall model ... 93
9.2.2.4. Hypothesis 1b individual brands ... 94
9.2.2.5. Hypothesis 2 overall model ... 100
9.2.2.6. Hypothesis 2 individual brands ... 101
9.2.3. Moderating analysis ... 107
9.2.3.1. Hypothesis 3a ... 107
4 Abstract
Despite a large body of research, the country of origin effect is still not completely understood. Based on the model of Hamzaoui-Essoussi, Merunk and Bartikowski, (2011), the
country of origin effect in relation
with brand equity is examined. This is done in order to find out if the findings of Hamzaoui-Essoussi et al. (2011) also apply to the brands originating in one of the BRICS countries. In a quantitative analysis, the influence of the brand origin and the country of manufacture image on two dimensions of brand equity is examined. In addition, brand typicality is measured as moderator on these relations. Results tell that overall seen, support is found for brand origin and country of manufacture image positively influencing brand equity for brands originating in one of the BRICS countries. However, zooming in at the six separate brands tells that it depends on the kind of brand whenever the hypotheses are supported or not. Furthermore, no statistical significant evidence is found for brand typicality as moderator. This study contributes to existing country of origin literature by examining the country of origin effect on brand equity of brands originating in one of the BRICS countries. This thesis concludes with suggestions for future research on the country of origin effect. Specifically, more research is needed for brands of one of the BRICS countries that have high brand awareness
5 1. Introduction
When a foreigner is asked to name the first brand that pops up when he or she thinks of Brazil, they would probably say Havaianas. Very few brands around the world have managed to remodel themselves and develop such a strong international known brand without having to change the essence of their product like Havaianas has (Huffington post Brazil, 2015). Even though Brazil has dealt with several economic and political crises last years, Havaianas did not suffer from this and even used her country of origin as a marketing tool to create brand equity.
This occurrence is a perfect example of the country of origin effect; the country image or representation that customers link to products of that particular nation (Nagashima, 1970). As said above, consumers strongly relate the sunny flip-flops of the brand Havaianas to the white beaches of Brazil and based thereon, Havaianas developed a strong internationally recognized brand.
The country of origin effect is not a new concept and has been the subject of dozens of studies in the last decades. Most studies focused on the country of origin effect in relation to the quality perception of customers. Many authors found a positive influence of the country of origin effect on the quality perception of customers (Steenkamp, 1990). A distinction was made between developed and developing countries; developed countries tend to be perceived as producing a high quality product and developing countries suffered from a low perception of quality (Cordell, 1992; Pappu, Quester and Cooksey, 2007). Other authors stated that the country of origin is not only a factor that influences the perception of product quality but also a way to influence customers’ emotions and feelings towards a brand (Fournier, 1998). Therefore, the country of origin can influence brand equity as a whole rather than solely brand quality (Fournier, 1998, Hamzaoui-Essoussi, Merunk and Bartikowski, 2011).
6 Hamzaoui-Essoussi et al. (2011) were one of the first to develop a framework of different dimensions of the country of origin effect in relation with different dimensions of brand equity instead of solely brand quality. They distinguished country of origin in brand origin (BO) and country of manufacture (COM). They brought these dimensions in relation with brand equity, which they divided in brand quality and brand image. They found significant evidence for their hypothesis and so they found evidence for the fact that the country of origin effect has a positive influence on brand equity (Hamzaoui-Essoussi et al., 2011).
Although the framework of Hamzaoui-Essoussi et al. (2011) is an extension of previous studies on the country of origin effect, it does not take the changing globalization into account. The growth and success of emerging countries like Brazil, Russia, India, China and South Africa (also known as the BRICS countries) could have an impact on the framework of Hamzaoui-Essoussi et al. (2011). Several authors have found significant evidence for consumers preferring brands from developed countries over brands of developing countries (Batra, Ramaswamy, Alden, and Steenkamp, 1999). The BRICS countries developed tremendously in the last decade and therefore it might be possible that brands originating in of the BRICS countries could benefit from the country image of their brand of origin. Therefore, it would be interesting to examine if the framework of Hamzaoui-Essoussi et al. (2011) also applies to brands originating in one of the BRICS countries. This leads to the following research question:
’To what extent does the country of origin effect influence brand equity of brands originating in one of the BRICS countries?’’
This study is an explanatory quantitative study based on online structured surveys. The used sample contained Dutch consumers between 18 and 65 years old. Six well-known brands, which originate from three of the six BRICS countries (Brazil. Russia and China), were used in the survey: Havaianas, Smirnoff, Xiaomi, Lenovo, Huawei and Embraer.
7 This study contributes to existing literature by examining the influence of the country of origin effect on the brand equity of brands from the BRICS countries. The outcome of this study will extend existing country of origin literature by examining if proven models of the country of origin effect also apply to brands from the new “booming” economies like Brazil, Russia and China. Managers of brands of the BRICS countries as well as managers of Western brands can use these findings to make strategic choices in the field of brand building.
This thesis starts with the literature review, which entails the theory in the field of country of origin effects, the influence of the country of origin effect on brand equity, and ends with the relation between the country of origin effect and the BRICS countries on which five hypotheses are based. This is followed by the discussion of the research gap and the research question. After that, the research method and the results are discussed where after the discussion and conclusion follows. This thesis concludes with the bibliography and appendix in which the used questionnaire and the most important SPSS output are presented.
8 2. Literature review
2.1. Country of origin effect
The country of origin effect is one of the most extended topics in international marketing and has been studied since 1962 (Brijs, Bloemer and Kasper, 2011). The study of Dichter (1962) formed the starting point of the country of origin effect. His study examined the success or failure of a brand due to the country in which the product was manufactured (Dichter, 1962). Not long after, Schooler (1965) did the first empirical test and agreed with Dichter by finding differences between product evaluations of similar products with different country of origins. Nagashima came up with the first official definition and Nagashima (1970, p. 68) defined the country of origin effect as: ‘’the picture, the reputation, the stereotype that businesses and consumers attach to products of a specific country’’. This reputation is created by factors like economic growth, political background, historical stories and culture (Nagashima, 1970). Checchinato, Disegna and Vescovi (2003, p. 411) extended this definition by stating that ‘’the country of origin effect is the way in which consumers perceive products originating from a particular nation’’. So, the country of origin effect can be seen as an information cue which has a strong effect on consumer perception and sometimes even buying behavior (Verlegh and Steenkamp 1999).
Despite the fact that the country of origin effect has been researched tremendously, there are still a lot of contradictions going on and so the country of origin effect is still not clearly understood (Verlegh and Steenkamp, 1999). Due to several contradicting research outcomes, two parties have emerged who believe in two contradicting theories. Magnusson and Zdravkovic (2011) divided these two parties in the traditional theories and the new theories. The traditional theory claims that the country of origin effect does affect product evaluations (Magnusson and Zdravkovic, 2011; Verlegh and Steenkamp, 1999 and Pharr, 2005). This claim is based on different kind of studies that are done with different products, brands,
9 companies and countries (Magnusson and Zdravkovic, 2011). However, the supporters of the new theory are skeptical towards these findings. This new theory discusses that due to the globalization of today’s economy, consumers do not rely anymore on the country of origin of products. As multinationals operate all around the world and manufacture in different countries, there is no need to base your buying process on the fact in which country a product comes from or in in which country a product is manufactured (Amdt and Liefeld, 2004). Sammie, Shimp and Sharma (2005) and Balabanis and Diamantopoulos (2008) supported this statement by finding empirical evidence for consumers only recognizing one-third of country of origins with the right brands. So, this tells that most of the consumers do not even know in which country a brand produces or where a brand originally originates from. This makes the country of origin effect less interesting for marketers as consumers do not have the knowledge where a brand operates of originates from and so do not use this knowledge in their buying process or in the development of their attitude towards the brand. Magnusses and Zdravokovic (2011) capitalized on the formation of these contradicting sides in the country of origin theory and did a study which served as a bridge between the new and traditional perspectives. They conducted a quantitative study in which they examined the country of origin effect on brand attitude of American consumers towards thirty-three brands belonging to three product categories: televisions, automobile and the fashion industry. The brand of origin of these brands were among others South Korea, the Netherlands, Japan, Germany, Sweden, UK, Italy and France. They included the beliefs of the new theory by measuring the accuracy of the American consumers in terms of the brand of origins of these brands. The statistical significant results of Magnusses and Zdavokovic (2011) showed that the country of origin effect remains important, regardless of the objective accuracy of brand of origin perceptions. So, the brand attitude of consumers can either being influenced when somebody thinks that Dior originated in the USA as if that same person believes that Dior originated in
10 France. The study of Magnusson and Zdravokovic (2011, pp. 454) concluded that telling consumers the truth about the brand of origin changes attitudes and even strengthens the perception of brand. So when the consumer of the latter example gets to know the truth about the brand origin of Dior, their attitude can even be reinforced.
In the same year, Hamzaoui-Essoussi, Merunka and Bartikowski (2011) agreed with Magnusson and Zdravokovic (2011) on the remaining importance of the country of origin effect. They extended the country of origin literature by measuring merely dimensions of the country of origin effect in relation with merely dimensions of brand equity. As one can see in figure 1, Hamzaoui-Essoussi et al. (2011) developed a framework in which they tested the influence of the country of origin effect divided in brand of origin and country of manufacture, which again were divided in micro and macro image. The influence of the country of origin effect was examined on two dimensions of brand equity: brand quality and brand image. They found significant evidence for brand origin positively influencing both brand quality and brand image. Also the influence of the country of manufacture on perceived brand quality was confirmed. So, it may be clear that the country of origin still is relevant for international marketing and is an important factor for creating brand equity. As the model of Hamzaoui-Essoussi et al. (2011) was the first to combine multiple dimensions of the country of origin effect with multiple dimensions of brand equity, this model will be the leitmotif on which this study is based. In the following chapter the different dimensions of the country of origin effect will be discussed.
11 Figure 1. Country of origin model.
Source: Hamzaoui-Essoussi et al. (2011).
2.2. Dimensions of the country of origin effect
As said before, the country of origin effect has been influenced by the globalization of today’s economy. Firms all over the world have expanded their businesses abroad in order to gain several advantages with asset advantages and transaction advantages being the most important reasons to expand (Birkinshaw, Hood and Jonsson, 1998). This led to companies operating all over the world whereby the countries where companies produce often differ from the country in which the company originates from. Therefore, many products are called bi-national or ‘’hybrid’’ which means that the production, design or origin happens across national boundaries (Iyer and Kalita, 1997).
That is why a distinction has been made in the country of origin literature between the country in which the brand is founded (country of product design), the country in which the majority of the product’s final assembly occurred (country of assembly) and the country where most of
Brand Typicality COM Macro Image COM Micro Image BO Micro Image BO Macro Image Brand Quality Brand Image
12 the materials come from and/or the components are made (country of manufacture) (Insch and MCBride, 2004). Hamzaoui-Essoussi et al. (2011) used these dimensions in their framework but combined country of assembly and country of manufacture. The latter authors also made a distinction between the micro and macro image of brand of origin and country of manufacture.
2.2.1. Brand origin
To start with the country in which the brand is founded: brand origin, abbreviated as BO. Hamzaoui-Essoussi et al. (2010, p. 973) defined brand origin as ‘’the country where the brand originates from, which belongs to the nationality of the brand’’. For instance, Thakor and Kohli (1996, p. 26) defined brand origin as ‘’the place, region or country where a brand is perceived to belong to its target customers’’. Martin and Cervino (2011, p. 532) extended these definitions by stating that ‘’the brand of origin forms the identity and the personality of the brand’. Although the globalization of the economy has forced multinationals to expand parts of their firms abroad, the brand origin cannot be moved to foreign countries as this country has been determined from the day on which the brand established. Therefore, brand origin is known as a permanent characteristic of a brand (Brouthers and Hennart, 2014; Martin and Cervino, 2011). As a result, brand origin is seen as a more reliable indicator of the country of origin effect than the country of manufacture as the latter can change from time to time (Thakor and Kohli, 1996; Lim and O’cass, 2001; Kinra, 2006). The brand origin of a brand belongs to the long-term memory of the consumer as consumers keep linking the brand to the country in which it is established (Keller, 1993). A clear example of this is Volkswagen; although the production department of this brand has been moved to several foreign countries like France, Spain, Brazil and Japan, consumers still link this company to her brand origin Germany.
13
2.2.2. Country of manufacture
The second dimension Hamzaoui-Essoussi et al. (2011) used in their framework is the country of manufacture, abbreviated as COM. The country of manufacture used by Hamzaoui-Essoussi et al. (2011) includes both the country of assembly and the country of manufacture mentioned by Insch and McBride (2004, p. 974) as they defined the country of manufacture as ‘’the country that produces or assembles the branded product’’. Relocating departments of multinationals such as manufacturing departments has been the trend of the last decades. In particular, manufacture departments have been expanded to countries with low production costs in order to obtain cost benefits (Barner-Rasmussen, Piekkari and Bjorkman, 2007). This led to the consequence that the country of manufacture is often changing and therefore is seen to be more unstable and unpredictable than brand origin. ‘’As country of manufacture is more uncertain and has often been exposed to more change, country of manufacture seems to be a weaker brand association than brand origin’’ (Hamzaoui-Essousi et al. 2011, p. 974). A much-discussed distinction is made between the country of manufacture as a developed or developing country. Many studies have shown that consumers prefer brands producing in developed countries over brands producing in developing countries (Cordell, 1992). Therefore, it seems to be beneficial for brands to choose developed countries for their country of manufacture over developing countries. Studies have showed that the economic level is a determinant of the country of origin effect and so consumers prefer developing countries over developed countries. They tend to reason that product quality correlates with the economic level of the country of manufacture (Cordell, 1992).
2.2.3. Macro and micro country images
The last two dimensions Hamzaoui-Essoussi et al. (2011) used in their framework are macro and micro country images. As Hamzaoui-Essoussi et al. (2011) found only partially
14 significant evidence for their hypotheses of micro brand origin and micro country of manufacture, the decision is made to only use the macro image of brand origin and country of manufacture.
The macro country image contains the worldwide level and this includes a broad variety of associations linked to a specific country (Hooley, Shipley, Krieger, 1998; Laroche, Papadopoulos and Heslop, 2003). Examples of these are national symbols, economic and political levels, levels of industrialization and cultural dimensions (Hamzaoui-Essoussi et al. 2011). The micro image is about the relation between the image and a specific product category (Hamzaoui-Essoussi et al. 2011). An example of the COO micro image is red wine from France. France is known for her ability of producing high quality red wine, therefore any red wine made in France will profit from the positive image of the red wine industry of France in general. As the micro country image will not be used as this study proceeds, the definitions macro and micro will not be used and one can assume that when the terms brand origin and country of manufacture are used, the macro images of these variables are meant.
2.3. Brand equity
Hamzaoui-Essoussi et al. (2011) measured the influence of the just discussed dimensions of the country of origin effect on two dimensions of brand equity: perceived brand quality and brand image. Brand equity is seen as one of the most valuable assets for any firm as it is ‘’an intangible asset that represents any brand’’ (Keller and Lehman, 2003, p. 27). Brand equity has been the subject of hundreds of studies and different brand equity models are developed. In general sense, brand equity is defined as ‘’the marketing effects uniquely attributable to the brand – for example, when certain outcomes result from the marketing of a product or service because of its brand name that would not occur if the same product or service did not have that name’’ (Keller 1993, p. 1). Nam, Ekinci and Whyatt (2011) extended this definition by
15 stating that brand equity consists of brand image, perception of service or product quality and brand familiarity. Two perspectives have been used to look at brand equity: the financial perspective and the strategy-based perspective. The financial perspective is used to measure brand equity in straight figures and is in general done for managerial accounting (Keller 1993). From the financial perspective, Simon and Sullivan (1990) defined brand equity as the future cash flow of a brand with a brand name in comparison with the future cash flow of brand without a brand name. The second perspective, which is the strategy-based perspective, focuses on increasing marketing productivity (Keller, 1993). As competition increases, costs are rising and the power of consumers strengthens, brands need a better understanding of consumer behavior as a basis for making better strategic decisions in the field of their marketing strategy (Keller, 1993). The purpose of the strategy-based perspective is to get an understanding of the needs and wants of the consumer and based on these develop a marketing strategy. So, the latter perspective focuses on the individual consumer instead of the figures and financial valuation. As this study examines the influence of the country of origin effect on consumer behavior, the strategy-based perspective will be used from now on.
Keller (1993) developed a framework based on the strategy-based perspective: the customer-based brand equity framework. He defined customer-customer-based brand equity as ‘’the differential effect of brand knowledge on consumer response to the marketing of the brand. A brand is said to have positive customer-based brand equity if customers react more favorably to the product, price, promotion, or distribution of the brand than they do to the same marketing mix element when it is attributed to a unnamed version of the product or service’’ (Keller 1993, p. 8). The newest version of this framework is the customer-based brand equity pyramid and therefore, this latter pyramid will be used as this study proceeds. This framework focuses primarily on the fit between the brand and the needs of the customer. The pyramid consists of
16 six brand-building blocks which are needed for successful brand development. As figure 2 shows, Keller (2001) linked these six blocks to four steps which are guidelines for the brand development process (identity, meaning, response and relationships). Building a strong brand starts with the bottom building block salience, which entails the awareness of the brand. In other words, people need to know that the brand exists and need to be able to recognize the brand. Step 2 in the process exists of two building blocks of the model, namely performance and imagery. Performance examines the way in which the product meets the needs of the customer in the field of product reliability, durability and serviceability (Keller, 2001). Imagery does the same but focuses on the social and psychological department. The response to performance and imagery can been found in step 3: judgments and feelings. Judgments are based upon quality, credibility, consideration and superiority and comprise the way in which consumers think the brand represents qualitative products. Hamzaoui-Essoussi et al. (2011) used this latter building block in their measures for brand equity as they divided brand equity in brand image and perceived brand quality (see figure 1). The second building block of step 3 is also used by Hamzaoui-Essoussi et al. (2011): feelings. Feelings represents the way in which consumers feel about the brand, this can be warmth, fun, excitement, security, social approval, and self-respect (Keller, 2001). The last step in the framework of Keller (2011) is the most desirable level of brand equity: brand resonance. When customers show consumer brand resonance, they do not only have positive accessible reactions to the brand but they also show intense active loyalty to the brand (Keller, 2001). An example of brand resonance could be a customer of Volkswagen who not only really loves the brand but also actively discusses the newest cars of Volkswagen in forums or in online communities.
17 Figure 2. Customer based-equity pyramid.
Source: Keller (2001).
2.4. The influence of country of origin effect on brand equity
As all variables have been defined and discussed, the following chapter will discuss literature about the influence of the two dimensions of country of origin effect (brand origin and country of manufacture) on two dimensions of brand equity (brand image and perceived brand quality). First the country of manufacture image on perceived brand quality will be discussed where after brand origin will be discussed in relation with the same brand equity dimension: perceived brand quality. This chapter will end with the influence of the brand origin on brand image.
2.4.1. The country of manufacture on perceived brand quality
The country of origin effect influencing perceived brand quality is one of the most studied relations in the field of the country of origin effect. Several authors found significant evidence for the fact that consumers use the country of origin effect as a cognitive cue, and use
18 information about the country where a company produces her products to create a perception about the quality of the product (Verlegh and Steenkamp, 1999). According to the latter authors, the country of origin effect is an extrinsic cue and belongs to other extrinsic cues like price and name or reputation of the brand (Steenkamp, 1990). Consumers use these extrinsic cues to develop an opinion about the quality of the product. This quality perception is done through product-country images, which are linked to the culture, national habits, symbols and products of a nation (Verlegh and Steenkamp, 1999, p. 525). Hamzaoui-Essoussi et al. (2011) only tested the country of manufacture in relation with perceived brand quality (and not brand image) as previous studies have shown that the country of manufacture dimension has the most impact on brand quality and less on brand image (Chao, 1998; Srinvisan, Jain and Sikand, 2004). As the country of manufacture is less stable and is often changing due to globalization, consumers tend to link this information more to perceived brand quality than brand image as this is the country where the products are produced and not where the identity of the brand comes from (Chao, 1998).
Previous studies already showed that the country of manufacture positively influences perceived brand quality. Most of these studies focused on the difference between products made in developed and developing countries. For instance, Johannsson and Nebezahl (1986) found in their study that cars produced in developed countries were perceived as products with higher quality than cars produced in developing countries. Han and Terpstra (1988) and Jian and Guogun (2007) found similar evidence by stating that consumers prefer products made in developed countries over products made in developing countries. Iyer and Kalita (1997) agreed with the latter authors; their study entailed four product categories: sneakers, jeans, portable stereos and watches and the product quality was rated higher when the country of manufacture was a developed country instead of a developing country. Srinivasan et al. (2004) extended previous finding by using both extrinsic and intrinsic cues in contrast to the
19 previous studies that used the country of origin as the only information cue. Srinvisan, et al. (2004) recognized the importance of more information cues for consumer to base their evaluation on and used a multi-cue setting to measure the influence of the country of origin effect. Srinvisan et al. (2004) came up with the same conclusions as previous studies as they found that when the country of manufacture is a developed country, the quality ratings were higher than when the country of manufacture was a developing country. Besides that, they also concluded that when a country of manufacture is changing from a developing country to a developed country, the quality evaluation increases (Srinvisan et al. 2004). In addition to their study, Hamzaoui-Essoussi et al. (2011) found significant support for the country of manufacture influencing brand quality positively. The study of Hamzaoui-Essoussi et al. (2011) confirmed earlier findings by concluding that brand quality of developed countries of manufacture of televisions and cars are rated higher than these of developing countries.
So, it might be clear that brand quality of products made in developed countries is rated higher than these produced in developing countries. Product category also plays a role in these findings. Iyer and Kalita (1997) stated that when nontechnical, fashion products (for example Chanel bags) are used to measure the influence of the country of manufacture on brand quality, this relation is stronger than when low-technology technical products are used. Low-technology technical products contain products which are technical but do not require complicated technological production. These products have more standardization in their production processes and consequently consumers tend to believe that these production skills can easily be transferred to developing countries.
2.4.2. Brand origin on perceived brand quality
To continue with the brand origin on perceived brand quality. Namely, Hamzaoui-Essoussi et al. (2011) found significant support for brand origin positively influencing perceived brand
20 quality. Several elderly studies found already statistical significant evidence for this result. These studies examined brand origin image on perceived brand quality in different kind of settings and using different kind of products. They found support for products in general (Nagashima, 1970; Anderson and Cunningham, 1972; Bannister and Saunders, 1987), for specific product categories (Gaedeke, 1973; Schooler, 1965), and for different kind of brands (Gaedeke, 1973), and so it has been proven that brand origin image does influence perceived brand quality in different kind of product categories and brands (Bilkey and Nes, 1982). Just like the studies about the country of manufacture on perceived brand quality, several studies made a distinction between developed and developing countries. Also here, consumers tend to prefer brands that originates from developed countries over brands that originate from developing countries (Bilkey and Nes, 1982). One of these study was the one of Insch and McBride (2004) who examined the influence of the country of manufacture and brand origin on perceived brand quality. They used inhabitants of the USA and Mexico as respondents and found a difference between functional and fashion products. For consumers in the USA, both the country of manufacture and brand origin positively influenced the perceived brand quality. For consumers in Mexico, only statistical significant evidence was found for country of manufacture influencing perceived brand quality. Insch and McBride (2004) stated that brand origin was less an important cue for solely functional products in Mexico whereas brand origin was found to be important for fashion-oriented products in Mexico. They concluded that brand origin has a stronger impact on the respondents in the USA than the Mexicans. This could be because consumers of the USA attach more importance to the style of the brand whereas Mexicans attach more importance to the functionality of the products (as they have less wealth) (Insch and McBride, 2004). This is in line with the previous discussed findings of Iyer and Kalita (1997) who examined besides the influence of the country of manufacture on perceived brand quality also the brand origin on perceived brand quality.
21 They concluded that the brand origin of fashion products is more important to consumers than brand origin of low technology technical products. An example of this is the popular fashion brand Chanel. This brand is famous for her exclusive French haut-couture designs and consumers believe that the design skills really originate from that particular developed country. Chanel would completely lose her dignity when they would move the complete company to a developing country.
2.4.3. Brand origin on brand image
Now that the influence of the country of origin effect on perceived brand quality is discussed, the influence of the country of origin effect on brand image will be reviewed. Several studies have stated that the country of origin effect has more meaning just being an extrinsic cue for product quality (Li and Wyer, 1994). According to Li and Wyer (1994) country of origin has also an emotional meaning to consumers by anticipating on emotions like pride, love and solidarity for a particular nation. The study of Fournier (1998) agreed with Li and Wyer by concluding that the country of origin effect may lead to consumers relating products to national identities. ‘’This phenomenon can lead to strong emotional attachment to certain brands or products’’ (Verlegh and Steenkamp, 1999, p. 523). Experiences such as holidays or education may influence the perception of customers about a brand (Verlegh and Steenkamp, 1999). An example of this is a study done by Batra, Ramaswamy, Alden, Steenkamp and Ramachander (1999) who found that Indian customers prefer products that originate from western countries because they admire the politics, economics and culture of western culture. This example makes clear that these Indian people do not only use information about the country to develop a perception towards product quality but also include their feeling for the country (western countries) in their preference for particular brands which can be related to the building block ‘’feelings’’ of the customer brand equity pyramid of Keller.
Hamzaoui-
22 Esoussi et al. (2011) confirmed these conclusions by finding significant support for the brand origin of televisions and cars positively influencing the brand image of these products.
Different product categories seem to be important for the country of origin effect. Li and Wyer (1994) noted that brand image increases due to the country of origin effect in the case of high-involvement products. Lia and Wyer (1994) explained that the high-involvement products are products on which the purchase decision is quite important. As this purchase decision becomes more elaborate, the importance of the country of origin effect increases. These products tend to be more luxury goods (Chanel bag) than necessity goods (grocery) as consumers buy groceries without thinking while buying a luxury good takes a lot of consideration. Piron’s (2000) findings are in line with Li and Wyer (1994) by concluding that the country of origin effect is stronger for consumed luxury goods than consumed necessity goods. Lastly, Sirgly et al. (1991) mentioned that the country of origin effect is linked to the image consumers create about them self. So, the country of origin should be higher for publicly seen products than private products (Piron, 2000). This is in line with Iyer and Kalita (1997) who stated that consumers tend to be more sensitive for the country of origin effect for products with symbolic status (for examples cars). So it can be concluded that brand origin does positively influence brand image, although this depends on the extent of developed/developing brand origin and the product category.
2.5. The country of origin effect on brands originating from BRICS countries
So as all dimensions and relations of the country of origin effect have been discussed, it has become clear that consumers prefer brands originating from and/or producing in developed countries over originating from and/or producing in developing countries. Although many studies have focused on the country of origin effect in relation with developed versus
23 developing countries, there is limited research on the country of origin effect in relation with strongly upcoming economies like Brazil, China or Russia (Cordell, 1992; Pappu et al, 2007).
Jim O’Neill developed the term BRIC countries with which he recognized the growing importance of Brazil, Russia, India and China as embracers of global capitalism (Cheng, Gutierrez, Mahajan, Shachmurove and Shahrohki, 2007). Later, the S was added to the phenomenon BRIC in order to involve the largest economy of African continent: South Africa (Gammeltoft, 2008). The BRICS countries have developed tremendously in the last decades and are therefore seen as important new players in the world economy (Biggerman and Fam, 2011). The BRICS countries are responsible for a considerable part of the goods and services consumed globally and are trading intensely amongst each other with an average growth in excess of 5% annually (Biggerman and Fam, 2011; p. 5). Predicted is that China and India will become dominant suppliers of manufactured goods and services, while Brazil and Russia would become suppliers of raw materials (Cheng et al, 2007; p. 144). For Brazil the chances lie in soy and iron ore while Russia needs to focus on oil and natural gas (Cheng et al, 2007; p. 144). As every country has her own characteristics and advantages, predicted is that these five countries have the ability to form a strong economic block that could even be a competitor of the G6 which is an unofficial group of the six richest industrialized countries: France, West Germany, Italy, Japan, the United Kingdom and the United States (Cheng et al, 2007). As is briefly mentioned in the introduction of this thesis, several brands of the BRICS have become multinational players and developed as worldwide known strong brands. Examples of these are Havaianas, the flip-flops brand of Brazil, Huawei, the mobile phone company of China and Embraer, the airplane company that even is the supplier of Dutch’s pride; the KLM.
24 As there are a lot of opportunities for these five countries, the BRICS countries still have a lot of obstacles to face and steps to make (Cheng et al, 2007). Cheng et al. (2007, p. 145) stated that the BRICS countries need to improve their long-term conditions to promote growth including macroeconomic stability, political institutional development, trade and investment openness, and education.
So the BRICS countries have showed their potential for the future with successful growth in the past but certainly also a lot of factors to work on and to develop. Therefore, it would be interesting to see how consumers perceive these countries; are they able to produce quality high standard products and develop strong brands or are they still more seen as developing countries? As it is already confirmed by several studies that consumers do prefer developing countries over developing countries, it might be possible that the BRICS countries could benefit from the same country of origin effect as developed countries. As several studies have shown that the brand origin image positively influences brand image it would be interesting to measure if this also applies to brands originating in one of the BRICS countries. When European consumers see the BRICS countries as upcoming developed countries they would have excited and enthusiastic feelings (building block feelings of customer-based equity pyramid) about brands originating in one of the BRICS countries. This leads to hypothesis 1a which is states as follows:
Hypothesis 1a: The brand origin of brands originated in one of the BRICS countries positively influences brand image.
Insch and McBride (2004) showed that brand origin has more influence on perceived brand quality in developed countries than it has in developing countries as these countries tend to
25 have more interest in style and sophisticated products. When European consumers see the BRICS countries as potential developing countries, they would perceive brands originating in the BRICS countries as brands that deliver qualitative high standard products and therefore hypothesis 1b is as follows:
Hypothesis 1b: The brand origin of brands originated in one of the BRICS countries positively influences perceived brand quality.
Studies have shown that the country of manufacture image strongly positively influences brand quality whereby consumers rate quality products of developed countries much higher than quality of products of developing countries. When European consumers would perceive the BRICS countries as potential developed countries the country of manufacture image of these brands would positively influence the perceived brand quality of brands originating in one of the BRICS countries. Therefore, hypothesis 2 is as follows:
Hypothesis 2: The country of manufacture of brands originated in one of the BRICS countries positively influences perceived brand quality.
2.6. Brand typicality
Hamzaoui-Essoussi et al. (2011) used brand typicality as a moderator in their framework. They defined the concept as ‘’the degree to which an object represents a category’’ (Hamzaoui-Essoussi et al., 2011, p. 974). They stated that due to high typicality, spontaneous category associations can occur (Hamzaoui-Essoussi et al., 2011). Studies about memory association strength confirmed this by stating that high typicality empowers consumers to remember brands after exposure to a brand or category cue (Hamzaoui-Essoussi et al., 2011;
26 Loken and Ward, 1987). This leads to consumers seeing the brand as the category and branded products as elements of that same category (Hamzaoui-Essoussi et al., 2011). So with regard to the country of origin effect, when brands have high brand typicality, consumers will immediately relate the brand origins to these brands. When a brand is typical for her brand origin, the brand can profit from the positive image of the brand origin and recall the brand when they are exposure to the country. This is seen in the example of the Havaianas flip-flops mentioned in the introduction; these flip-flops are a typical representation of the white beaches of Brazil. Hamzaoui-Essoussi et al. (2011) found significant evidence for brand typicality moderating the effects of brand origin on the two dimensions of brand equity. If the model of Hamzaoui-Essoussi et al. (2011) also applies to brands originating in one of the BRICS countries, this would mean that brand typicality would moderate the relations between brand origin/country of manufacture and brand equity of brands originating in one of the BRICS countries. This leads to the following hypothesizes:
Hypothesis 3a: Brand typicality moderates the positive influence of the brand origin on brand image of a brand originated in one of the BRICS countries.
Hypothesis 3b: Brand typicality moderates the positive influence of the brand origin on the brand quality of a brand originated in one of the BRICS countries.
27 3. Research gap
Although the study of the country of origin effect has been studied intensively the last decades, there are still a lot of unclear dimensions regarding this phenomenon (Verlegh and Steenkamp, 1999). Hamzaoui-Essoussi et al. (2011) developed a framework to integrate both the brand origin and country of manufacture in relation with brand equity, measured as brand quality and brand image. Although this framework contributes to the existing literature by integrating more dimensions of both the country of origin effect and brand equity, it does not take into account the globalization of the current economy. The upcoming strong economies such as the BRICS countries could influence the existing literature regarding the country of origin that states that consumers would prefer brands of developed countries to developing countries. Therefore, it might be interesting to examine if the framework of Hamzaoui-Essoussi et al. (2011) also applies to brands originating in one of the BRICS countries, as they are no longer trailing countries.
This leads to the following research question:
‘’To what extent does the country of origin effect influence brand equity of brands originated in one of the BRICS countries?’’
The conceptual framework illustrated in figure 3 offers an overview of the five hypothesizes used in this study. As stated before, this framework is based on the model of Hamzaoui-Essoussi et al. (2011). Though in this study, the framework will be examined through brands originating in one of the BRICS countries.
28 Country of Manufacture Perceived brand quality Brand image Figure 3. Conceptual model
Source: author
Hypothesis 1a: The brand origin of brands originated in one of the BRICS countries positively influences brand image.
Hypothesis 1b: The brand origin of brands originated in one of the BRICS countries positively influences perceived brand quality.
Hypothesis 2: The country of manufacture of brands originated in one of the BRICS countries positively influences perceived brand quality.
Hypothesis 3a: Brand typicality moderates the positive influence of the brand origin on brand image of a brand originated in one of the BRICS countries.
Hypothesis 3b: Brand typicality moderates the positive influence of the brand origin on the brand quality of a brand originated in one of the BRICS countries.
Brand origin Brand Typicality H1A H2 H1B H3A H3B
29 4. Data and Methodology
This chapter describes the sample, data collection and the methodology. It provides an insight into the research design and manner of data collection. This chapter ends with a description of the sample and the measures of the independent, dependent and control variables.
4.1. Research
The purpose of this study was to examine if the findings of the study of Hamzaoui-Essoussi et al. (2011) regarding the positive influence of brand origin and country of manufacture on brand equity also can be used for brands originating in upcoming economies also known as the BRICS countries: Brazil, Russia, India, China or South-Africa. Due to time shortage and difficulty of finding appropriate brands, only three of the six BRICS countries were used as brand origin. Expected is that brand typicality will play a moderating role in this relationship as Hamzaoui-Essoussi et al. (2011) already found that this was the case for their model. The attitude of Dutch consumers towards six different brands (Havaianas, Huawei, Xiaomi, Smirnoff, Lenovo and Embraer) originating in three of the BRICS countries (China, Russia and Brazil) will be analyzed.
4.1.1. Data collection
This study is a quantitative explanatory study and a cross-sectional survey design (single point in time) was used from a relative large amount of respondents (211 respondents). The choice for this design was made because cross-sectional designs have been used in other studies examining the country of origin effect (Hamzaoui-Essoussi et al. 2011). As the population of this study is quite large (see 4.1.2 Sample), a convenient-sampling method on voluntary basis, which falls within non-probability sampling methods, was used to gather the data. A digital questionnaire was distributed through Facebook as this is a medium where both young and older Dutch consumers are active on. Besides Facebook, the link of the digital
30 questionnaire was distributed by e-mail to several potential respondents. The questionnaire started with a brief introduction in which the expected timing of filling in the survey was discussed as well the anonymity was confirmed. The questions of the survey were ranked in logical order and attention has been paid to creating diversity in the subjects (so no brand image after perceived brand quality as they both fall under brand equity). The survey was originally in English but was translated into Dutch to prevent problems in understanding English. Therefore, the parallel translation technique was used, which consists of the translation of two independents individuals. These two translations were combined into one Dutch questionnaire.
4.1.2. Sample
The population for this study was Dutch consumers. Especially chosen was for Dutch consumers as the Netherlands is viewed as a prototypical example of a country with high levels of foreign trade, and as a nation that is well integrated into the global economy (Nijssen and Douglas, 2004). So if the country origin effect is in force for brands originating in the BRICS countries, Dutch consumers would be one of the first to show this. As Dutch consumers is a broad population, no special requirements were set for the sample, the participants only needed to have an age between 16 and 65 years. To ensure that the sample is a well representation of the population, demographic questions were asked (age, gender, educational background). A minimum of 200 responses were required to ensure analyzable data, in the end 211 responses were collected.
4.1.3. Selected brands
The brands used in this study were selected on two conditions. The first condition was recognizably. As it was important that the Dutch consumers had an attitude or idea about the brands, the most well-known brands originating in one of the BRICS countries were selected.
31 As Hamzaoui-Essoussi et al. (2011) made a difference between the country of origin and the country in which the brand produces, it was important to select brands with difference between their brand origin and country of manufacture. These selected brands manufactured in merely foreign countries and therefore the choice is made to select the country as country of manufacture where they produce the most. The requirement to have a different country of manufacture and brand origin made the choice of useful brands quite narrow and therefore no specific requirements could be made for product categories. Besides that, these requirements led to the consequence that the selected brand originated in only three of the six BRICS countries, namely China, Russia and Brazil. Due to these requirements and shortage of time it was not possible to selected brands that originated in all of the BRICS countries. The above discussed criteria led to the following brands: Havaianas, Smirnoff, Embraer, Huawei, Xiaomi and Lenovo (see table. 1).
Table 1. Selected brands
Brand Brand origin Country of Manufacture
Havaianas Brazil Brazil & Argentina
Smirnoff Russia United Kingdom
Embraer Brazil Brazil and the United States
Huawei China China and India
Xiaomi China China and India
Lenovo China China and the United States
Source: author
4.2. Measures
The measures of the dependent, independent, moderating and control variables will be discussed below. All the measures of this research were proceeded from previous research in
32 which all the items had a Cronbach’s alpha higher than 0.70. The constructs of the questionnaire can be found in the appendix.
4.2.1. Dependent variables
The dependent variables contained brand image and perceived brand quality.
Brand image was measured by using a scale of Lee (1993), shortened to six items. The Cronbach’s alpha of this scale was .94. The items were adapted to the used brands, for example: ‘’I think I would be happy with flip-flops of the brand Havaianas’’.
Perceived brand quality was measured by using Dodds & Grewal’s (1991) scale, shortened to five items. A seven-point scale was used whereby each scale position was weighted from -3 to +3. The Cronbach’s alpha of this scale was .95. Again, this measure was adapted to the used brands.
4.2.2. Independent variables
The same scale was used for the country of origin image and the country of manufacture image as they both have the purpose to measure the image of countries. Martin & Eroglu’s (1993) scale was used, shortened to six items. A seven-point scale was used whereby each scale position was weighted from -3 to +3. The Cronbach’s alpha of this scale was .93. The scale was adapted to the different countries in which the brands originated from, namely: Brazil, Russia and China. The same goes for the country of manufacture, namely: Argentina, United Kingdom, United States and India.
4.2.3. Moderating variable
Brand typicality was measured by using Loken & Ward’s (1990) scale, shortened to three items. A five-point and ten-point scale was used with a Cronbach’s alpha of .82. Again, this scale was adapted to the different brands, for example: Havaianas is a good example for the country Brazil.
33
4.2.4. Control variable
To be sure that no other factors influence the results of the country of origin effect, two control variables were used: gender and ethnocentrism. Gender is chosen as control variable as Li-Wen and Stergquist (1994) found that gender could influence attitudes towards foreign brands. Ethnocentrism is the extent to which consumers prefer domestic products/companies over foreign products and/or companies. So when the respondents are highly ethnocentrically, they would evaluate brands domestic brands higher than foreign brands (Chryssochoidis, Krystallis and Perreas, 2007).
Ethnocentrism was measured by using Shimp and Sharma’s (1987) scale, shortened to four items. A seven-point scale was used with a Cronbach’s alpha of .94. The scale was adapted to Europe, for example: A real European should always buy European-made products.
34 5. Results
In the following chapter the results of the collected data are presented. First, the preparation process of the data set, descriptive statistics of the sample, reliability checks and a correlation matrix are given after which the results of the five tested hypotheses are presented.
5.1.1. Data preparation
The first step in this study was the preparation of the data set. No errors were found in the data and missing values were deleted from the whole data set (surveys that were not completed) in such a way that only cases were analyzed that did not contain any missing data, also known as excluding cases list wise. Some recoding was done to three counter indicative items (item 3, 5 and 6) of brand image for all six brands. The scale was adapted from 1 (totally disagree) to 5 (totally agree) into 5 (totally disagree to 1 (totally agree). After the reliability checks (see chapter 5.3), variables were computed. A distinction was made between variables which contained the items of all brands taken together and variables which entailed the items per brand. For example, for the dependent variable brand image one variable was made with all the items of the six brands taken together (BI_TOT) and six variables were made for the items about brand image for each brand separately (BI_HAV, BI_SMIRN, BI_HUAWEI). In this way, analysis could have been done for the overall effect as well for the brands separately.
5.1.2. Demographic statistics
In the following section, the demographic statistics are presented. The survey was filled in by 211 respondents. These respondents were all Dutch as the survey was written in the Dutch language and the respondents were explicitly told at the beginning of the survey that the questionnaire was for Dutch only. As one can read in the pie chart in figure 4, 61.61% of the respondents were female while 38.39% were male with an average age of 28 years (figure 5).
35 As stated in figure 6, the average education level of this sample was ‘University of Applied Science’’ with sub sequentially ‘’University’’ as second most mentioned education.
Figure 4. Distribution of gender Figure 5. Distribution of age.
Source: author Source: author
Figure 6. Distribution of education Source: author 38% 62%
Gender
Male Female 13% 12% 39% 36%Level of education
36
5.1.3. Reliability analysis
In order to be sure that the used variables were consistent, reliabilities checks were done on all the variables. A Cronbach’s alpha score can vary between 0 and 1 and to be sure that the variable is reliable the Cronbach’s alpha needs to be at least .70.
Reliability analysis for total variables
To start with the total variables by which variables are meant that test all the answers of the six brands of the same subject. For example, for the variable brand image the items about brand image of all the six brands were taken together. As one can read in table 2, all the total variables have a higher Cronbach’s alpha than the required .700. This means that these variables were reliable and their items measured the same concept. Some items were deleted in order to increase their Cronbach’s alpha. For the country of manufacture image CI_VS6 was deleted in order to increase the Cronbach’s alpha from .841 to .846. The lowest Cronbach’s alpha belonged to the country of origin image with a score of .841 and the highest is from ethnocentrism with a score of .919. So concluded can be that all the variables are reliable and can be used as the study proceeds.
Table 2. Cronbach’s alpha of total variables
Variable Cronbach’s alpha
Ethnocentrism .919
Brand image .883
Perceived brand quality .893
Brand origin image .841
Country of manufacture image .846
Brand typicality .863
Table 2. Cronbach’s alpha of total variables
37 Reliability analysis for variables of individual brands
After the reliabilities checks on the total variables, reliabilities checks were done on variables of the separate brands. For example, brand image of the brand Havaianas did only contain items concerning brand image of that particular brand (instead of all the items as with brand image of the total variables).
As can be read in table 3, all these variables had a Cronbach’s alpha that was higher than the required .700 which means that all items of these variables measured the same thing. Some items were deleted in order the increase the Cronbach’s alpha. For instance, the BO country image for Havaianas Ci_B6 was deleted in order to increase this Cronbach’s alpha from .766 to 7.86. The same was done for country of manufacture image which increased from .846 to .870 when CI_C6 was deleted. Finally, the item CI_A6 was deleted for country of manufacture image of Embraer and Havaianas in order to increase the Cronbach’s alpha from .880 to .893.
Table 3. Cronbach’s alpha for variables of individual brands
Brand image Perceived brand quality
Brand origin Country of manufacture image Havaianas .809 .900 .786 .893 Huawei .771 .916 .771 .870 Smirnoff .74 .848 .83 .826 Xiaomi .809 .931 .771 .870 Embraer .846 .961 .786 .893 Lenovo .844 .924 .771 .870 Source: author 5.1.4. Correlation analysis
Before this study tests the proposed hypotheses, a correlation matrix was done between all total variables. The matrix in table 4 shows a strong correlation between brand origin image and brand image (.306, p < .05). The same goes for brand origin and perceived brand quality
38 (.291, p < .05). Country of manufacture image seemed to have a strong positive correlation with perceived brand quality as well (.215, p < .05). Lastly, brand typicality has correlations with the dependent variables; brand image (.333, p < .05) and perceived brand quality (.335, p < .05) as well the independent variables: brand origin (.272, p < .05) and country of manufacture (.264, p < .05).
Table 4. Correlation Matrix
Variable M SD 1 2 3 4 5 6 7 1. Gender 1.62 .49 2. Ethnocentrism 2.38 1.19 .109 (.92) 3. Brand image 3.82 .73 -.120 -.228** (.88) 4. Perceived brand quality 3.5 .42 .104 -.106 .460** (.89) 5. Brand origin 4.27 64 .084 -.036 .306** .291** (.84) 6. Country of manufacture image 4.64 .56 -.089 -.060 .328** .215** .569** (.85) 7. Brand typicality 5.77 .91 .113 -.075 .333** .335** .272** .264** (.86)
**. Correlation is significant at the 0.01 level (2-tailed).
Source: author
5.2. Hypothesis testing – Hierarchical regression
After data preparation, reliability checks, variables computing and the calculation of a correlation matrix, the hypothesizes could be tested. To start with the hypothesis 1a.
5.2.1. Hypothesis testing – Hypothesis 1a
Hypothesis 1a: The brand origin of brands originated in one of the BRICS countries positively influences brand image.
In order to test this hypothesis a hierarchical multiple regression was done in order to investigate the ability the influence of the brand origin on brand image, after controlling for gender and ethnocentrism. In the first step of hierarchical multiple regression, the two control variables were entered, namely: ethnocentrism and gender. This model was statistically
39 significant F (2,208) = 6.773; p < .005 and so gender and ethnocentrism explained 6.1% of the difference of brand image (see table. 5). After entry of brand origin image at step 2 the total variance explained by the model as a whole was 15.6% F (3,207) = 12.760; p < .005. When looking only at the variance done by the brand origin, this independent variable turned out to be the cause of 9.5% of variance of brand image (R2 Change = .095; F (1,207) = 23.281; p < .001). In the final model, the brand origin and ethnocentrism turned out to be statistically significant with the brand origin recording the highest Beta value (β = .310; p<.001). As this hierarchical regression showed that the brand origin image does positively influence brand image, hypothesis 1a can be supported. A side note needs to be made that although the result is statistically significant, the variance is not very large which means that the brand origin only is responsible for a little amount of the changed brand image.
40 Table 5. Hierarchical regression hypothesis 1a overall effect
1 R R2 R2 Change B SE β t Step 1 .247 .061** Gender -.144 .100 -.097 -1.433 Ethnocentrism -.133 .041 -.217** -3.215 Step 2 .395 .156** .095*** Gender -.185 .096 .124 -1.928 Ethnocentrism -.124 .039 -.203** -3.158 Brand origin .350 .072 .310*** 4.825
Note. Statistical significance *p<.05; **p<.01; ***p<.001
Source: author As there is statistical significant support for hypothesis 1a, now hierarchical regressions were done for each brand individually to see which brand caused this significant result. As one can read in table 5 no statistically significant evidence was found for gender as control variable therefore this variable was not captured in the following analysis for the individual brands.
1 ‘’R = Correlation coëfficiënt: correlation between observed and predicted values of
dependent variable.
R2 = Coëfficiënt of determination: the proportion of variance in the dependent variable
which can be explained by the independent variables (at step 1 or 2). An overall
measure of strength of association and does not reflect the extent to which any particular independent variable is associated with the dependent variable.
R2 Change = Indicates the proportion and change of variance explained when entering an independent variable.
B = Regression coëfficiënt: values for the regression equation for predicting the dependent variable from the independent.
SE = Standard error: standard errors associated with the coefficients.
β = Standardized coefficient (Beta): represents the change in the outcome resulting from a unit change in the independent variable.
t = T-test: shows if the b-value of each predictor is different from 0.’’ (UCLA: Statistical Consulting Group, 2016).
