Effect of National Culture on the Adoption of
Radical Innovation
Jelle J. Kok
MSc Advanced International Business Management and Marketing J.J.Kok@student.rug.nl
Newcastle University Business School Supervisor: Dr A. Javornik
170807371
University of Groningen Supervisor: Mr H.U. Haq
S2744155
2
ABSTRACT
This research attempts to discover whether cultural differences between countries can explain the differences in the consumer adoption of radical innovation. Although radical innovation adoption strongly differs from incremental innovation and is regarded essential for the economic growth of firms and nations, it has received relatively little attention in literature. In this research, culture has been studied on a national level, using the dimensions of Hofstede (power distance, individualism, masculinity and uncertainty avoidance) and Hall (context). Radical innovation adoption has been measured using data from two Eurobarometer reports, in which people were asked about self-driving vehicles. For this quantitative approach across 26 European countries, correlation tests and t-tests have been executed. In line with the expectations, the cultural dimensions of power distance and uncertainty avoidance are negatively linked to radical innovation adoption. Masculinity is negatively linked to radical innovation adoption, which is opposite of what was expected. It is thought to be that the different nature of radical innovation, among other things by the high degree of consumer learning required for adoption, plays an important role in these relationships. Furthermore, control variables as traffic deaths and education level had an impact on the results. An additional finding is that men are substantially more likely to adopt a radical innovation than women. Overall, this study produces various academic and managerial contributions, for example by acknowledging the specific nature of radical innovation and its adoption. The research is limited by the fact that the data used was not specifically created for this purpose, not every potential factor could be taken into account and only highly developed European countries were studied.
Keywords: Innovation Adoption, Culture, Consumer Behavior, Radical Innovation, Hofstede
3 TABLE OF CONTENTS ABSTRACT 2 LIST OF TABLES 5 LIST OF FIGURES 5 1 INTRODUCTION 6 2 LITERATURE REVIEW 9 2.1 Culture 9
2.2 Cultural dimensions and innovation adoption 10
2.3 Radical innovation adoption 12
2.3.1 Defining radical innovation 13
2.3.2 Relative advantage 14
2.3.3 Compatibility 15
2.3.4 Complexity 16
2.3.5 Perceived risk 18
2.4 The relationship between culture and radical innovation adoption 19
2.4.1 Power distance 19
2.4.2 Individualism - collectivism 20
2.4.3 Masculinity - femininity 21
2.4.4 Uncertainty avoidance 23
2.4.5 High versus low-context cultures 24
3 METHODOLOGY 25
3.1 Strategy 25
4
3.3 Variables and measurement 28
3.3.1 Dependent variable 28 3.3.2 Independent variables 30 3.3.3 Control variables 31 3.4 Data analysis 33 3.5 Reflections 34 3.6 Ethical considerations 35 4 FINDINGS 36 4.1 Assumption testing 36 4.2 Main results 40 4.3 Controlling results 42 4.4 Additional findings 43 5 DISCUSSION 45
5.1 Analysis of main results 45
5.2 Analysis of additional findings 49
6 CONCLUSION 50
6.1 Academic contributions 50
6.2 Managerial implications 51
6.3 Limitations and further research 52
7 REFERENCES 55
5
LIST OF TABLES
Table 2.1. Existing research regarding cultural dimensions and innovation adoption 11
Table 2.2. Types of innovation (based on Chandy and Tellis, 1998) 13
Table 2.3. Differences in adoption of radical innovation and incremental innovation 17
Table 3.1. Descriptive statistics Eurobarometer 2014 and 2017 27
Table 4.1. Shapiro-Wilk test for normality 38
Table 4.2. Self-driving vehicle adoption, Hofstede values and Hall values 39
Table 4.3. Results of correlation tests on Hofstede’s and Hall’s dimensions 41
Table 4.4. Results of independent samples t-test on Hall’s dimension of context 42
Table 4.5. Results of correlation test on control variables 42
Table 4.6. Results of significant relationships tested for control variables 43
Table 4.7. Results of independent samples t-test on gender 44
Table 8.1. Data control variables 74
Table 8.2. Adoption of self-driving vehicles by gender 75
LIST OF FIGURES
Figure 2.1. Conceptual model of cultural dimensions and radical innovation adoption 24
6
1 INTRODUCTION
Due to globalization, rising incomes (of the middle class) and increasing technology, consumers from different countries appear to become more and more similar. Furthermore, factors as global transport and capitalism, which lead to the expansion of multinational corporations, are influencing the global consumer behavior (Cleveland and Laroche, 2007; Ger, 1999). Multiple authors argue that this leads to more homogeneous consumer needs (Bullmore, 2000; Czinkota and Ronkainen, 1998). Levitt (1983) explained that idiosyncratic consumer preferences of specific regions will cease to exist because of globalization. This dynamic is driven by technology, it will lead to the same high quality, low-cost products all over the world. This argument can be seen as something rational, technology and costs will lead to converging consumers behavior. However, consumer behavior is based on more than only rational arguments (Suerdem, 1994; Wood and Moreau, 2006). The supposed homogenization of consumer behavior is merely upheld by anecdotal evidence, consumer cultures are still far from homogeneous (De Mooij and Hofstede, 2002). Current and historical conditions, interacting with the forces of globalization, form specific consumption patterns on each level, national, regional or local (Ger, 1999). Each consumer has a unique context influencing their consumption pattern, based on their own lives in relationship to the rest of the world. Based on this, Kotler (1994) identified four main domains that influence consumers behavior: cultural, social, personal and psychological factors. Of those, cultural factors have “broadest and deepest impact” (p. 174). Culture strongly influence the way people perceive the world and thus make decisions (Briley, Morris and Simonson, 2000).
7 Steenkamp, Ter Hofstede and Wedel, 1999; Yaveroglu and Donthu, 2002; Yeniyurt and Townsend, 2003). They found that the innovation adoption by consumers differs across countries with different cultures. This research will distinguish itself from existing articles by its focus on the adoption of radical innovation. According to multiple researchers, radical innovation has thus far received too little attention, especially in comparison with the amount of research about innovation in general (Chao, Reid and Mavondo, 2012; Cowden and Kalliny, 2013; Hervás-Oliver et al, 2018; McDermott and O’Connor, 2002). It strongly differs from incremental or continuous innovation in its nature, having more impact on organizations and consumers (Veryzer Jr, 1998). For consumers, the adoption of radical innovations requires a significant change in, or even entirely new behavior (Calantone, Chan and Cui, 2006; Gatignon and Robertson, 1991). It is interesting to discover what drives consumers in adopting radical innovation, why would they purchase a whole different, possibly unknown, product or service?
Thus, taking in mind the importance of radical innovation and the impact for consumers, this resulted in the following research question:
How can cultural differences between countries explain the differences in the consumer adoption of radical innovation?
8 The purpose of this research is to find out more about the background of differences in adoption of radical innovation across countries. This is relevant in several ways. First of all, it contributes to the existing research regarding radical innovation adoption. Radical innovation is something that incorporates substantial new knowledge or technology and that meets the needs of customers better than existing products or services (Chandy and Tellis, 1998). Not only does the adoption require entirely new behavior, it also creates new knowledge structures (Gregan-Paxton and John, 1997). It will be interesting to see whether radical innovation is influenced by culture, since culture has a strong impact on behavior and learning patterns (Markus and Kitayama, 1991; Tempelaar et al, 2013).
9
2 LITERATURE REVIEW
2.1 Culture
Culture has a strong influence on every aspect in human behavior (Samuel Craig and Douglas, 2006). Because it is so incorporated in human existence, it is often hard to grasp how and to what extent it impacts life. Cultural norms and values are forceful powers that shape perceptions, emotions and behaviors (Markus and Kitayama, 1991). McCracken (1986) embraces an overarching view of culture, exhibiting in two ways. First of all, culture is “ the lens through which the individual views phenomena” (p. 72). As such, it establishes how people perceive and apprehend phenomena. Second, culture is “the blueprint of human activity, determining the coordinates of social action and productive activity, and specifying the behaviors and objects that issue from both” (p. 72). In this way, it determines how the world will be shaped by human effort.
A high variety of definitions of culture has been established over time. A commonly used definition is established by Engel, Blackwell and Miniard (1995), referring to culture as “a set of values, ideas, artefacts, and other meaningful symbols that help individuals communicate, interpret, and evaluate as members of society” (p. 63). In their book, they state that culture has a strong effect on consumer behavior, it influences what products people purchase and the communication and decision-making about products.
10 and Sood (1989) discovered that consumer behavior can partially be attributed to the home country of the respondent. This ‘national culture’ is regarded to be effective, it has successfully explained aggregate national consumer behavior. It clarifies the differences between consumers of one country and another country (Roth, 1995).
2.2 Cultural dimensions and innovation adoption
11 Table 2.1. Existing research regarding cultural dimensions and innovation adoption
Study Dimensions of culture Measure of
innovation adoption
Findings
Dwyer, Mesak and Hsu (2005)
Hofstede (power distance, individualism,
masculinity, uncertainty avoidance and long-term orientation)
Diffusion rate of technological product innovations
Masculinity and power distance are positively linked with innovation diffusion rate, individualism, uncertainty avoidance and long-term orientation negatively
Singh (2006) Hofstede (power distance, individualism,
masculinity, uncertainty avoidance)
Adoption of new products
High power distance, high uncertainty avoidance and feminine countries adopt via normative influences, collectivistic
countries via interpersonal communications Steenkamp, Ter Hofstede and Wedel (1999) Schwartz (conservation, self-enhancement) and Hofstede (individualism, masculinity, uncertainty avoidance) Consumer innovativeness
Consumers in individualistic, masculine countries are more innovative, consumers focused on conservation, and in uncertainty avoiding countries are less innovative
Tellis, Stremersch and Yin (2002) Hofstede (masculinity, uncertainty avoidance), need for achievement and industriousness
New product take-off
Probability of takeoff increases with higher need for achievement and industriousness and lower uncertainty avoidance, no results found for masculinity
Yaveroglu and Donthu (2002)
Hofstede (power distance, individualism, uncertainty avoidance) and Hall (context) Diffusion of new products, by coefficients of innovation and imitation
Individualism, low uncertainty avoidance, low power distance and high-context are linked with innovation, collectivism, high uncertainty avoidance and low-context countries with imitation
Yeniyurt and Townsend (2003)
Hofstede (power distance, individualism, masculinity, uncertainty avoidance) Penetration rate of new (technology) products
High power distance and high uncertainty avoidance negatively influence penetration rate, individualism has a positive influence, no relation for masculinity found
12
2.3 Radical innovation adoption
This section will discuss what radical innovation is and how its adoption differentiates from incremental innovation adoption. The adoption of innovations is a complex process, consisting of multiple stages (Gatignon and Robertson, 1991; Mittelstaedt et al, 1976). This has been discussed by many authors, and ranges from the first awareness of the consumer with the product or innovation to the actual adoption (Rogers, 2003). The adoption of radical innovation differs strongly from incremental or continuous innovation in various aspects (Aggarwal, Cha and Wilemon, 1998; Reinders, Frambach and Schoormans, 2010; Veryzer Jr, 1998). The characteristics of radical innovation adoption and the differences with incremental innovation adoption will be discussed using Rogers’ (2003) characteristics of innovation adoption. Three of those characteristics, relative advantage, compatibility and complexity, are regarded as the most relevant attributes regarding innovation adoption (Kapoor, Dwivedi and Williams, 2014). Based on Ostlund (1974) another characteristic will be added, perceived risk.
2.3.1 Defining radical innovation
13 than the current rate (Gatignon et al, 2002). A similar concept is ‘discontinuous innovation’, described by Veryzer Jr (1998, p. 137) as “radically new products that involve dramatic leaps in terms of customer familiarity and use”. He argues that those products commonly include the application or development of notable new technologies. In this research, the definition of Chandy and Tellis (1998) will be used, since it most used in other literature and also incorporates the customer perspective .
Table 2.2. Types of innovation (based on Chandy and Tellis, 1998)
Customer need fulfillment per dollar
Low High
Newness of Low Incremental innovation Market breakthrough
technology High Technological breakthrough Radical innovation
2.3.2 Relative advantage
14 Consumers that adopt a radical innovation will attract more attention than adopters of other types of innovation, it will make the consumer stand out of the crowd (Aggarwal, Cha and Wilemon, 1998). Holak and Lehmann (1990) argue that if the design of a new product or innovation is regarded as complex, it will be regarded by customers as more advantageous.
2.3.3 Compatibility
Compatibility is defined as “the degree to which an innovation is perceived as consistent with the existing values, past experiences, and needs of potential adopters” (Rogers, 2003, p. 15). Rogers (2003) identifies two kinds of (in)compatibility of new products: incompatibility with previous or current products and incompatibility with the needs and expectations of consumers. First of all, there is a higher incompatibility for radical innovations than for incremental innovations regarding previous or current products. Incremental innovations tend to fit into existing product categories, whereas radical innovations create new product categories (Aggarwal, Cha and Wilemon, 1998; Hoeffler, 2003). Often, a radically new product is the first and only in its product category, generally for a certain period. Second, in contrast to incremental innovations, radical innovations tend to fulfill latent needs of consumers (Hurmerinta and Sandberg, 2015; Jaworski, Kohli and Sahay, 2000). This means that consumers are not aware of their desire for the new product or service, this will reach their consciousness when it is introduced (Holt, 1976).
15 their openness to change their ‘mental models’, since it will significantly change their consumption or usage patterns.
In contrast to incremental innovation, there is a lack of familiarity for customers when adopting a radical innovation (Veryzer Jr, 1998). This will, in most cases lead to resistance by consumers. Partially, this is caused by the fact that the new product or innovation did not fit in the existing consumption patterns. If people are more familiar with a product, the mental adoption will be quicker. Due to the lack of knowledge about the new product or innovation, consumers might be focused on irrelevant features. Consumers sometimes judge a product by characteristics that are not related to the essential novel aspects, while a crucial aspect like safety is not always taken into account.
2.3.4 Complexity
Rogers defines complexity as “the degree to which an innovation is perceived as relatively difficult to understand and use” (2003, p. 15). Because radical innovations are noticeably dissimilar to existing products in their nature and complexity, there might be a high amount of consumer resistance, even though the benefits are high (Veryzer Jr, 1998). An explanation for this is that most radical innovations are considerably incongruent. This means that they are inconsistent with the current cognitive schema of consumers of a product, which leads to a more negative view (Meyers-Levy and Tybout, 1989).
16 tend to already have some elemental knowledge or experience to understand what the product is (Hoeffler, 2003). Consumers will only have to identify the product category to understand the innovation. To understand radical innovations however, a substantial amount of learning is required from consumers, so new knowledge structures should be constructed (Gregan-Paxton and John, 1997).
17 will reduce their uncertainty. It creates a ‘surrogate experience’ with the radical innovation, which allows consumers to be more accurate in their estimations of the benefits of the product.
2.3.5 Perceived risk
Another critical factor influencing the likelihood of consumers to purchase a new product is perceived risk (Shimp and Bearden, 1982). In the consumer behavior literature, this is defined as “the consumer perceptions of the probability and magnitude of potential negative consequences resulting from a purchase” (Sarin, Sego and Chanvarasuth, 2003, p. 72). Because consumers have no history of usage with the radical innovation, consumers will perceive more risk for this. There is limited information available about the product, which will lead to higher amounts of uncertainty and risk. Furthermore, the uncertainty about the usefulness of radical innovations is higher than for incremental innovations (Hoeffler, 2003). This is caused by the fact that the technological advantages of radical innovations are not immediately visible for consumers, which results in a higher degree of consumer resistance. There is also a financial risk, since radical innovations are typically introduced at relatively high market prices (Aggarwal, Cha and Wilemon, 1998).
Table 2.3. Differences in adoption of radical innovation and incremental innovation Adoption of radical innovation Adoption of incremental
innovation
Product category Creates new product category Fits in existing product category
Consumer learning High amount of learning is
required, primarily by analogies and mental simulations
Low to moderate amount of learning required, primarily by category-based learning
Consumer behavior Requires significant changes or
entirely new behavior
18
Perceived risk High perceived risk, uncertainty
about usefulness
Low perceived risk, usefulness is clear
Relative advantage Higher relative advantage, not
always perceived
Lower relative advantage
2.4 The relationship between culture and radical innovation adoption
In this section, the cultural dimensions used will be examined and the argumentation towards the hypotheses will be drawn. The conceptual model (figure 2.1) at the end of the section will give an overview of the expected relationships.
2.4.1 Power distance
Power distance is described as “the extent to which the less powerful members of institutions and organizations within a country accept that power is distributed unequally” (Hofstede, 2001, p. 9). In countries with a higher score on power distance, there tend to be more formal rules and authority.
19 Because of the importance of authority and centralization, the way people acquire and process information differentiates low power distance countries from high power distance countries. In low power distance countries, people are more verbally oriented, they tend to be more active in acquiring information, mainly through the media and friends. On the other hand, in high power distance countries information is passively transferred through interpersonal, implicit communication. This results in the fact that people from low power distances countries are substantially more likely to view themselves as well-informed consumers than people from high power distance countries (De Mooij, 2010). If people have more knowledge about a new product or innovation, they are more willing to adopt this (Wood and Moreau, 2006). Especially for radical innovations this is relevant, since these require a substantial amount of consumer learning (Hoeffler, 2003).
All in all, this resulted in the following hypothesis:
H1: The cultural dimension of power distance is negatively linked with the adoption of radical innovation.
2.4.2 Individualism - collectivism
20 collective thinking, people will conform with the norms and values of the group. An example of this is Japan, in this collectivistic country housewives indicate that their decision-making process is influenced by at least 8 other housewives (De Mooij, 2010). Singh (2006) argues that for openness to innovations, a consumer should sometimes display new behavior that not adheres to the current values and norms. For radical innovation adoption this is evidently the case, since it requires a significant change in or even entirely new behavior (Calantone, Chan and Cui, 2006; Gatignon and Robertson, 1991).
Consumers that adopt a radical innovation will attract a substantial amount of attention, it will make them stand out of the crowd (Aggarwal, Cha and Wilemon, 1998). In individualistic countries, people want to differentiate themselves from others, whereas in collectivistic countries people tend to behave similar to others (De Mooij, 2003). Uniqueness is a key component in individualistic countries, whereas collectivistic countries focus on uniformity (Ho and Chiu, 1994). It is expected that in individualistic countries there will be more adoption of radical innovation, because in those countries people will take initiative themselves and base their decisions on personal judgments. All things considered, this leads to the hypothesis:
H2: The cultural dimension of individualism is positively linked with the adoption of radical innovation.
2.4.3 Masculinity - femininity
21 modesty, caring for the weak, and the quality of life” (Hofstede, 1984, p. 84). Masculine countries are described as focused on performance, ambition and material values, whereas equal rights, looking for consensus and caring for the weak are values of feminine countries (Hofstede, 2001).
First of all, money and things are essential in masculine countries (Hofstede, 1980). Since radical innovations are generally introduced at relatively high market prices, it is expected that those factors weigh heavily (Aggarwal, Cha and Wilemon, 1998). By adopting new products, consumers could display their wealth and success (Tellis, Stremersch and Yin, 2003). Steenkamp, Ter Hofstede and Wedel (1999) confirm that, mainly because of materialism, consumers in masculine countries are more likely to purchase new products.
Secondly, in masculine countries people that are successful and independent are admired (Hofstede, 1980). By the adoption of radical innovations, consumers will receive a substantial amount of attention and display new or changed behavior (Aggarwal, Cha and Wilemon, 1998; Calantone, Chan and Cui, 2006; Gatignon and Robertson, 1991). Thus, independent decision-making will be necessary for this is and is likely to be higher in masculine countries than in feminine countries (Tellis, Stremersch and Yin, 2003). Furthermore, decision-making will take more time in feminine countries, since all stakeholders involved will be consulted for the decision. In masculine countries decisiveness is regarded as something crucial, it demonstrates masculinity (De Mooij, 2010).
22 H3: The cultural dimension of masculinity is positively linked with the adoption of radical innovation.
2.4.4 Uncertainty avoidance
Uncertainty avoidance is “the extent to which the members of a culture feel threatened by uncertain or unknown situations” (Hofstede, 2001, p. 161). Also defined as the tolerance of a society towards ambiguity, countries with a high score tend to be less open to different thoughts and ideas, focusing on one absolute truth, and vice versa.
Thus, it could be expected that in countries with a high uncertainty avoidance, because of the risk-aversion, people will be less open for innovations. These results are consistent with what Hofstede (2001) argued: in uncertainty avoiding countries, there is a strong resistance against the ‘new’, both through formal and informal rules. Steenkamp, Ter Hofstede and Wedel (1999) support this, consumers in countries with a low uncertainty avoidance are more innovative. It is possible that this effect is stronger for radical innovation adoption than for other types of innovation, since the perceived risk is substantially higher (Aggarwal, Cha and Wilemon, 1998). When adopting a radical innovation consumers take more risks than adopting incremental innovations, for example because of the higher financial investment needed. Because there is often limited information available about the new product, it is unclear what the specific (technological) advantages are. Thus, the uncertainty for consumers about radical innovations is high (Hoeffler, 2003).
23 H4: The cultural dimension of uncertainty avoidance is negatively linked with the adoption of radical innovation.
2.4.5 High versus low-context cultures
Hall (1976) argues that cultures can be categorized when looking at the style of communication. He states that “high-culture transactions feature pre-programmed information that is in the receiver and in the setting, with only minimal information in the transmitted message, whereas low-context culture transactions are the reverse.” (p. 101). This means that in low-context cultures (such as United States and Switzerland) most relevant information is shared in words, thus explicitly. Low-context cultures are seen as factual, whereas high-Low-context cultures are regarded emotional (Camit, Noble and Algie, 2009). In high-context cultures (such as China, Japan and Greece) the context is highly important for the interpretation of information (Czinkota and Ronkainen, 1998).
24 something subtle and might require assertiveness. For this reason, it is thought that innovations are more likely to be adopted in low-context cultures.
This idea is supported by Kim, Pan and Park (1998), who discovered that low-context cultures are better at dealing with new situations than high-context cultures. They state that because people in low-context cultures know how to deal with a more complicated system of functioning, such as complex legal systems, they tend to be more creative when confronted with new things.
All things considered, this lead to the following hypothesis:
H5: Consumers from high-context countries will be less likely to adopt radical innovation than consumers from low-context countries.
25
3 METHODOLOGY
3.1 Strategy
This research will be executed from a positivist point of view. Positivists see reality as something stable, which can be seen and described objectively (Levin, 1988). In research, positivity tends to focus on facts, by formulating hypotheses and testing them statistically. Crowther and Lancaster (2012) argue that positivist research usually follows a deductive approach, which is the case for this research. It does not aim to produce an entirely new theory. Instead, this study attempts to contribute to the existing literature by generating novel insights on the topic of radical innovation and its adoption. An empirical, quantitative approach has been chosen for this study. When doing a deductive research on a country level, a quantitative research design is most suitable. In this research, secondary data will be used to compare countries, since it was financially and time-wise impossible to collect data from multiple countries ourselves. However, there are very suitable data sets already in existence, as explained in the next section.
3.2 Data collection and sample
26 The adoption of radical innovation will be measured using data of two Special Eurobarometer reports of 2014 and 2017 from the European Commission. In the Special Eurobarometer report about autonomous systems (2014), inhabitants of EU member states were surveyed about various topics of robots and automation. The survey was carried out by TNS, 27,801 people from various demographic and groups were consulted in face-to-face interviews in their preferred language. A similar report from the European Commission was published in 2017 about the attitudes towards impact of digitisation and automation of daily life. For this report, 27,901 people from EU countries were questioned in comparable settings as the 2014 report.
The data of these reports are fully available on the European Commission website, making it easy to use. Furthermore, the sample is very large and supposed to be diverse, which can be seen in table 3.1 on the next page. In terms of gender there is a fairly equal distribution between male and female respondents. From most EU countries approximately 1,000 respondents have been interviewed, both in 2014 and 2017. In Germany and the United Kingdom this number is even higher, whereas only in the relatively small countries Luxembourg and Malta a lower amount of people has been interviewed, about 500.
27 not included in the analysis of Hall’s dimension of context. These countries are Estonia, Hungary, Ireland, Latvia, Lithuania, Malta and Romania.
Table 3.1. Descriptive statistics Eurobarometer 2014 and 2017
2014 2017
Total Man Woman Avg. age Total Man Woman Avg. age
28 The data of the control variables GDP per capita and traffic deaths has been retrieved from the World Bank (2017; 2015), the Education Index (2017) is retrieved from United Nations Development Programme. For those variables, the most recent data will be used. The results will be controlled for the average of respondents in a country as well. This data will be retrieved from the Eurobarometer reports of 2014 and 2017. The average of both years will be taken as control variable. The control variables used in the research will be further explained in the following section.
3.3 Variables and measurement
This section will in detail discuss the methodological characteristics of the dependent, independent and control variables and how they are measured.
3.3.1 Dependent variable
29 driving systems across manufacturers (Borenstein, Herkert and Miller, 2017; Howard and Dai, 2014). Further, legal and ethical issues are relevant, questions like ‘whose fault is an accident?’ and ‘what if the vehicle is being hacked?’ raise.
Multiple authors described self-driving vehicles as a disruptive or radical innovation (Bernhart and Winterhoff, 2016; Boeglin, 2015; Fagnant and Kockelman, 2015; Leicht, Chtourou and Youssef, 2018; Talebian and Mishra, 2018). Regarding the framework of Chandy and Tellis (1998), self-driving vehicles meet both requirements to be seen as a radical innovation. It has a high newness of technology as explained before, and it will have a high customer fulfillment per dollar, looking at the possible applications.
30 The results of both Eurobarometer reports will be used to calculate the score of the EU countries on the adoption of radical innovation. For both reports, the questions regarding self-driving vehicles were answered on a scale from 1 to 10. The higher the score on this scale, the more positive the participant is towards self-driving vehicles. In this research someone is regarded to adopt self-driving vehicles if they have a score of 7 to 10 on the question. Thus, a percentage has been calculated for both 2014 and 2017 for every country has been calculated. A higher percentage means that a country is more willing to adopt a self-driving vehicle.
The average of those two years will be used as the proxy for the adoption of radical innovation. The reason those two reports have been taken together is that, although the questioning seems similar, they are different in its nature. Thus, it will give a more all-encompassing insight of how people view self-driving vehicles across EU countries.
3.3.2 Independent variables
Hofstede’s cultural dimensions will be used to measure cultural differences across countries. For each of the dimensions countries have a score between 1 and 100, based on Hofstede’s original research at IBM, later studies of him and the World Values Survey (Minkov and Hofstede, 2012). For the exact scores of every country, see table 4.2.
31 from 1 to 16, where 1 means a totally low-context country and 16 a totally high-context country. Because Hall initially intended context to be divided in categories rather than a continuous scale and thus to make a test between two groups possible, the ranking is divided in groups as well (Morden, 1999). For context this means that a country with a score of 1 to 6 is considered low-context, 7 to 10 is considered ‘medium’-context and 11 to 16 is considered high-context. Some scores were based on a regional level, such as Flanders and Walloon (Belgium), those scores are converted.
3.3.3 Control variables
In order to test whether the cultural dimensions have an influence on the adoption of radical innovation, multiple correlation analyses will be executed. These tests will be controlled for by the GDP per capita adjusted for the purchasing power parity (World Bank, 2017), the amount of traffic deaths per 100,000 people (World Bank, 2015), the average age of respondents in a country and the Education Index (UNDP, 2017). The exact data of the countries on the control variables can be found in table 8.1 in the appendix.
32 vehicles and the technology itself are high, only a small portion of the people will be willing to purchase this for its introductory price (Sergeenkov, 2019; Talebian and Mishra, 2018).
The amount of traffic deaths will be used because of the specific nature of the proxy for the dependent variable, self-driving vehicles. Anderson et al (2014) argue that self-driving vehicles will reduce the amount of traffic accidents significantly or even cease this. Thus, it could be expected that in countries with high traffic death figures, people will be more likely to adopt self-driving vehicles.
Multiple authors discovered that demographics strongly influence the adoption of innovation (Im, Bayus and Mason, 2009; Kumar, 2014; Tellis, Yin and Bell, 2009). Age appears to be a demographic factor that has an essential impact on innovation adoption (Dee Dickerson and Gentry, 1983). On average, people that are willing to adopt innovations are younger. For this reason, the average age of respondents in a country will be used as control variable.
33 children), both expressed as an index obtained by scaling with the corresponding maxima (UNDP, 2018).
3.4 Data analysis
The relationship between all cultural dimensions and the adoption of radical innovation will be tested by using a bivariate Pearson correlation test. If the assumption of normality will not be met for a specific variable, this variable will be measured using a Spearman Rank correlation test as well. This non-parametric test is suitable in case one of the variables of a correlation is not approximately normally distributed (Weaver et al, 2017). In the next section, it will be explained why a correlation will be used and why a regression is not suitable for this kind of research. For Hall’s dimension of context an independent samples t-test will be conducted as well, because initially it was intended to be a two category-scale. It is possible to divide countries in two groups in terms of Hall values. When there are significant results in the correlations, they will be controlled for using a partial correlation. This test will determine whether a possible relationship between the cultural dimensions and the adoption of radical innovation is influenced by any of the control variables. For the tests a significance level of 5% has been set. For all measurements the 25th edition of SPSS will be used, a software package for statistical analysis from IBM.
3.5 Reflections
34 sample size. For the Hofstede variables, the sample size consists of 26 observations, for the Hall variables there are 19 observations. For a multiple linear regression, which would be suitable in this kind of research, since the dependent variables is expected to be explained by multiple independent variables, a higher amount of observations is needed. Green (1991) states that 50 + 8 IV (with IV being the amount of independent variables) is a suitable formula to calculate the amount of observations needed. Thus, for a simple linear regression 58 observations would be required. In this research, with 5 independent variables, 90 observations would be needed to execute a successful multiple regression analysis. Similar results were found using an online tool from Soper (2019) to estimate the required sample size for a successful multiple linear regression. This consists of three factors, the expected effect size (f2), the desired power coefficient and the
35 Another methodological constraint is the multicollinearity between the independent variables. This means that results can be unreliable if one independent variable is highly correlated with another independent variable (Stat Trek, 2019). Smith (2006) discusses that the Hofstede dimensions power distance and individualism are not separable using empirical measures. Hofstede (2014) argues that this is caused by national wealth. Individualistic countries are wealthier, as well as low power distance countries. When only relatively rich countries are compared, this should not be an issue. However, when testing this with a Pearson correlation test, there appears to be a significant relationship between power distance and individualism (r(23) = -.574, p < 0.005). Especially in researches with relatively small samples such as this article, multicollinearity could strongly impact the quality of results (Stat Trek, 2019).
3.6 Ethical considerations
36
4 FINDINGS
4.1 Assumption testing
According to Laerd (2019a), three main assumptions have to be met before a valid Pearson correlation can be conducted. First of all, the data should be measured at a continuous level. This is the case for all variables. When looking at the independent variables, Hofstede’s dimensions are measured on a 0 to 100 scale, Hall’s context on a 1 to 16 scale. The dependent variable is measured as a percentage. The control variables, GDP per capita, traffic deaths, average age of respondents and Education Index score are measured on a continuous level as well.
Figure 4.1. Scatter plot of self-driving vehicle adoption and uncertainty avoidance
A second assumption is that there should be no significant outliers. In correlation tests, especially those with a relatively small sample size, an outlier will have a strong effect on the results. When
37 only looking at the self-driving vehicle adoption, one country stands out. Poland has by a considerable difference the highest score of all EU countries. Laerd (2019a) suggests to study scatter plots with both the dependent and independent variables to check for outliers. In multiple scatter plots, Poland appears to be an outlier. The most evident example of this is the relation between uncertainty avoidance and self-driving vehicle adoption, as can be seen in figure 4.1 on the previous page. Because Poland would be an outlier in several of the tested relationships, it will be not be part of this research.
38 (Weaver et al, 2017). This will be reviewed further in the discussion section. All other data is considered to be normally distributed, as can been seen in table 4.1 below.
Table 4.1. Shapiro-Wilk test for normality Variable
W-statistic
Degrees of
freedom p-value (significance)
Dependent variable
Self-driving vehicle adoption 0.949 26 0.220
Independent continuous variables
Power distance 0.980 26 0.873
Individualism 0.919 26 0.043*
Masculinity 0.976 26 0.785
Uncertainty avoidance 0.941 26 0.142
Low-, high-context 0.882 19 0.023*
Independent categorical variables
High-context 0.956 5 0.781
Low-context 0.809 9 0.026*
* significant at p-value less than 0.05
39 countries is not distributed normally (p = .026). However, because according to Laerd (2019b) the t-test is regarded very robust in terms of violating the assumption of normality, this variable will continue to be included in the results section.
Table 4.2. Self-driving vehicle adoption, Hofstede values and Hall values
*PD is power distance, IND is individualism, MAS is masculinity and UA is uncertainty avoidance
Countries Adoption 2014
Adoption 2017
Average
2014, 2017 PD* IND* MAS* UA*
Low vs high-context Scale 1-16 Low vs high-context Categorical Austria 25.50% 19.42% 22.46% 11 55 79 70 1 Low Belgium 19.84% 20.33% 20.09% 65 75 54 94 6 Low Bulgaria 28.19% 20.87% 24.53% 70 30 40 85 8 Medium Croatia 17.59% 10.65% 14.12% 73 33 40 80 8 Medium
Czech Republic 23.97% 30.23% 27.10% 57 58 57 74 8 Medium
Denmark 34.72% 31.70% 33.21% 18 74 16 23 4 Low Estonia 21.84% 21.98% 21.91% 40 60 30 60 x x Finland 22.40% 23.35% 22.87% 33 63 26 59 4 Low France 16.94% 19.92% 18.43% 68 71 43 86 14 High Germany 17.89% 19.25% 18.57% 35 67 66 65 1 Low Greece 12.70% 15.58% 14.14% 60 35 57 100 14 High Hungary 26.33% 22.80% 24.57% 46 80 88 82 x x Ireland 24.13% 21.11% 22.62% 28 70 68 35 x x Italy 21.92% 29.60% 25.76% 50 76 70 75 14 High Latvia 20.60% 23.55% 22.08% 44 70 9 63 x x Lithuania 31.16% 21.61% 26.39% 42 60 19 65 x x Luxembourg 20.58% 20.56% 20.57% 40 60 50 70 6 Low Malta 17.44% 23.69% 20.57% 56 59 47 96 x x Netherlands 34.95% 34.52% 34.74% 38 80 14 53 6 Low Portugal 19.63% 20.41% 20.02% 63 27 31 99 14 High Romania 22.60% 24.00% 23.30% 90 30 42 90 x x Slovakia 18.24% 15.14% 16.69% 100 52 100 51 8 Medium Slovenia 22.87% 25.32% 24.10% 71 27 19 88 8 Medium Spain 17.44% 15.82% 16.63% 57 51 42 86 14 High Sweden 33.30% 30.51% 31.91% 31 71 5 29 4 Low
40
4.2 Main results
A Pearson correlation test has been executed to test whether the cultural variables affect the adoption of radical innovation, for which self-driving vehicle adoption is a proxy. The two variables that are not normally distributed will be tested using both a Pearson correlation test and a Spearman correlation test. The tests resulted in three significant results. The cultural dimension of power distance is negatively linked with the adoption of radical innovation (r(26) = -.447, p = .022). Because the r-value is between -0.4 and -0.6, this is a moderate (negative) relationship (Dancey and Reidy, 2007). As can be seen in overview in table 4.3, the nature of this relationship is as expected. The positive correlation between individualism and the adoption of radical innovation is close to significance, however it is not significant (r(26) = .377, p = .058). Because individualism is not normally distributed, a Spearman correlation test has been executed as well. This report similar results, with no significant results discovered (r(26) = .330, p = .099). Between masculinity and radical innovation adoption a positive relationship was expected, yet opposite results have been discovered. Masculinity has a moderate, negative relationship with radical innovation adoption (r(26) = -.453, p = .020). A thorough explanation and analysis of the results can be found in the discussion in chapter 5. The expectation was that uncertainty avoidance is negatively linked with the adoption of radical innovation and this hypothesis has been confirmed. A moderate negative correlation has been determined (r(26) = -.506, p = .008). The other two Hofstede dimensions were expected to link positively with the adoption of radical innovation.
41 significant relationship (r(19) = -.355, p = .136). Because the dimension of context is not approximately normally distributed, a Spearman correlation has been executed as well, not leading to significant results (r(19) = -.388, p = .100).
Table 4.3. Results of correlation tests on Hofstede’s and Hall’s dimensions Cultural dimension Hypothesized influence on radical innovation adoption Results Pearson correlation (r) Spearman correlation (r) Hofstede H1 Power distance Negative Confirmed -0.447* H2 Individualism Positive No significant results 0.377 0.330
H3 Masculinity Positive Negative influence -0.453*
H4 Uncertainty avoidance Negative Confirmed -0.506** Hall H5 Low-, high-context Negative No significant results -0.355 -0.388
* significant at p-value less than 0.05 ** significant at p-value less than 0.01
42 Table 4.4. Results of independent samples t-test on Hall’s dimension of context
M (mean) SD (standard deviation) t-value Degrees of freedom p-value (significance) H5 High-context 0.190 0.044 1.84 12 0.090 Low-context 0.250 0.064 4.3 Controlling results
First of all, to examine whether the control variables are directly linked to the adoption of radical innovation, a Pearson correlation has been executed. The tests between GDP per capita, the amount of traffic deaths, the average age of respondents, the Education Index and the adoption of radical innovation did not lead to significant results, as can be seen in table 4.5.
Table 4.5. Results of correlation test on control variables
GDP per capita Traffic deaths Average age Education Index
Adoption of radical innovation
0.155 -0.273 0.179 0.379
43 little influence on the relationship between the two variables (Laerd, 2019c). When controlling for traffic deaths (r(23) = -.377, p = .063) and Education Index (r(23) = -.306, p = .137) however, the link between power distance and the adoption is not significant anymore. This demonstrates that the amount of traffic deaths and the Education Index score have a substantial influence on this relationship. When controlling the (negative) relationship between masculinity and the adoption of radical innovation, the results remain significant. This indicates that the control variables have little influence on this relationship, for the detailed results, see table 4.6. The relationship between uncertainty avoidance and the adoption of radical innovation is influenced by the Education Index (r(23) = .371, p = .068), the other control variables do not have a substantial influence here.
Table 4.6. Results of significant relationships tested for control variables Cultural variable Adoption of radical innovation Controlled for GDP per capita Controlled for traffic deaths Controlled for age of respondents Controlled for Education Index Power distance -0.447* -0.432* -0.377 -0.431* -0.306 Masculinity -0.453* -0.470* -0.475* -0.423* -0.420* Uncertainty avoidance -0.506** -0.490* -0.447* -0.547*** -0.371
* significant at p-value less than 0.05 ** significant at p-value less than 0.01 *** significant at p-value less than 0.005
4.4 Additional findings
44 section. Table 8.2 in the appendix displays the radical innovation adoption by gender of every country. An independent samples t-test has been conducted to assess whether there is a difference between the adoption of radical innovation between men and women. This demonstrate that men (M = .286, SD = .069) have a significantly higher adoption of radical innovation than women (M
= .169, SD = .042), t(50) = 7.33, p = <0.005. The full results of the test can be found in table 4.7.
Table 4.7. Results of independent samples t-test on gender
45
5 DISCUSSION
5.1 Analysis of main results
First of all, it was found that power distance is negatively linked to the adoption of radical innovation. This result is in line with what was expected, since people in countries with a high power distance are less likely to take initiative and are used to maintain the status quo (Hofstede, 1980; Steenkamp, 2001). Further, people in high power distance countries are less likely to regard themselves as “well-informed” (De Mooij, 2010).
However, when controlling this result for the amount of traffic deaths in a country, it is not significant anymore. Reader et al (2015) argues that a high score on safety culture is linked with low power distance countries. This is caused by the fact that in those countries there is an open discussion about safety, and safety actions will be taken proactively. Furthermore, the proxy used for radical innovation, self-driving vehicles, is in multiple ways related to safety. This development could exile the human error in traffic and so decrease the amount of traffic accidents (Anderson et al, 2014). Thus, although there is a significant relationship between power distance and the adoption of radical innovation, it is likely that this is caused or at least strongly influenced by the factor of safety.
46 high power distance countries education is more teacher-centered (Hofstede, 2011). Researchers tend to agree that student-centered education is more favorable, since it stimulates critical thinking and active learning (Weimer, 2002; Wohlfarth et al, 2008). Especially active learning is something that is highly needed for the successful adoption of radical innovations. Hence, it is likely that the adoption is higher in low power distance countries, because the education is more student-centered and thus of higher quality.
47 Between masculinity and the adoption of radical innovation a positive relationship was expected, however opposite results were found. The expectation was that because masculine countries focus on performance and material values, radical innovation would be more likely to be adopted here than in feminine countries. It is possible that this is caused by the high price-consciousness in masculine countries, which origins from the focus on money (Leo, Bennett and Härtel, 2005). For consumers in masculine countries it is important to purchase products for less, while for the adoption of radical innovations a high financial investment is required (Aggarwal, Cha en Wilemon, 1998). Hence, it is likely that consumers from more price-conscious countries are less inclined to spend a high amount of money on radical innovations.
Another possible explanation for this is the specific nature of self-driving vehicles. Van Everdingen and Waarts (2003) discussed that product characteristics could moderate the relationship between masculinity and innovation. Self-driving vehicles attempt to reduce traffic fatalities and could increase the mobility of people with disabilities and underaged people (Anderson et al, 2014). Through this, one could say that this innovation takes care of the weak. This is something that is very specific for a feminine country, being modest and caring (Hofstede, 2001). Conclusively, for this dimension opposite results than expected were found, possibly because of price-consciousness and the type of product chosen for radical innovation.
48 regarding culture and innovation adoption (Steenkamp, Ter Hofstede and Wedel, 1999; Tellis, Stremersch and Yin, 2002; Yeniyurt and Townsend, 2003). Especially for the adoption of radical innovation, which requires a high amount of cognitive effort to change the customers opinion, openness to the new is required . This is the case in low uncertainty avoiding countries and could explain the strong relation with the adoption of radical innovation.
However, the link between uncertainty avoidance and radical innovation adoption is influenced by the Education Index. The explanation could be similar to the power distance dimension, since in low uncertainty avoidance countries learning tend to be student-centered as well (Tempelaar et al, 2013). For the adoption of radical innovation a high degree of consumer learning is required (Hoeffler, 2003). Thus, it is likely that adoption is lower in uncertainty avoidance countries, because the education is less student-centered and for that reason of lower quality.
49 this could have had an impact on the expect negative relationship between high-context cultures and radical innovation adoption.
5.2 Analysis of additional findings
50
6 CONCLUSION
6.1 Academic contributions
This research adds several new insights to the existing literature regarding the relationship between culture and consumer behavior. It gives more understanding of the different nature of the adoption of radical innovation, which has received relatively little attention. It differs from incremental or continuous innovation in various ways, demanding a high amount of consumer learning and being introduced at higher market prices. Furthermore, the adoption of radical innovation induces significantly changed or new behavior and will make the adopter ‘stand out of the crowd’. These specific attributes of radical innovation presumably had an impact on the relationship with culture.
51 Although this was not part of the original research question, this study has confirmed that certain demographics strongly influence the adoption of innovation, and that it is likely that product category plays a role in this. With regard to self-driving vehicles, men have proven to be significantly more likely to adopt radical innovation than women.
6.2 Managerial implications
Not only on an academic level, but also for practitioners this study provides essential implications. First of all, the study intends to acknowledge the importance of radical innovation in companies and for consumers. It is crucial for firms, as it could ensure a strong competitive position on both short and long-term. However, a high number of radical innovation project fails, which is partially contributed to how it is commercialized. This research emphasizes the difference in consumer dynamics between radical innovation adoption and incremental innovation adoption. Especially on essential adoption attributes, such as consumer behavior, consumer learning and perceived risk it is highly dissimilar to incremental innovation adoption. By highlighting this, this research could contribute to managerial decision-making regarding the marketing of radical innovation.
52
6.3 Limitations and further research
As with almost every study, the findings beyond the specific sample (European countries) and specific product group (self-driving vehicles) might be limited. The countries compared appear to be fairly different in terms of culture and demographic factors at first sight, however they are not diverse in every way. First of all, the countries studied all have a relatively high income, all having a GDP per capita of above $20,000 (World Bank, 2017). Furthermore, every country included in the study is regarded a developed country. This is indicated by the fact that they are regarded as having a ‘very high human development’ on the Human Development Index (2018). This makes it debatable whether the findings can be generalized, especially for non-Western, less developed countries.
A second limitation is the proxy used for radical innovation adoption. Self-driving vehicles are characterized by many authors as a radical innovation, however there are some constraints. Because full self-driving vehicles have not been introduced on the market yet, their characteristics remain unclear in many ways. Although their costs and benefits may be described thoroughly, when an innovation is launched to the market the consumer behavior is not always as expected. When people are confronted with a whole new product, they have to obtain some insight in it before an opinion about it can be formed, according to Veryzer Jr (1998). Thus, the specific nature of self-driving vehicles could influence the generalizability of the results.
53 data set was not created with the purpose of researching the adoption of radical innovation. Furthermore, the question people were asked focused on whether they would be comfortable being driven in a self-driving vehicle, not if they would consider purchasing one. This question is possibly not enough focused on consumer behavior. However, a research of this scale would have been impossible without secondary data, as it interviewed more than 25,000 people from a high number of countries.
A fourth limitation is that although four control variables have been tested, the relationship could be caused by other factors. Multiple empirical studies have discovered that various consumer demographics such as mobility, professional status, ethnicity and family size influence influence consumer innovativeness (Kumar, 2014; Tellis, Yin and Bell, 2009; Im, Bayus and Mason, 2003). Not every demographic factor could be included in this research, primarily because the of lack of data about the respondents of the Eurobarometer reports. As a result of the fact that a correlation analysis will be used, it was not be possible to include moderators in the model. Various factors that will be used as control variables in this research, such as education and income, could have been interesting moderators regarding the expected relationship between culture and radical innovation adoption.
54 in their role as workers and not as consumers, this could influence the usefulness in terms of consumer research.
55
7 REFERENCES
Aggarwal, P., Cha, T. and Wilemon, D. (1998) ‘Barriers to the adoption of really-new products and the role of surrogate buyers’, Journal of Consumer Marketing, 15(4), pp. 358-371.
Alexander, D. L., Lynch Jr, J. G. and Wang, Q. (2008) ‘As time goes by: do cold feet follow warm intentions for really new versus incrementally new products?’, Journal of Marketing
Research, 45(3), pp. 307-319.
Ali, A., Krapfel Jr, R. and LaBahn, D. (1995) ‘Product innovativeness and entry strategy: impact on cycle time and break‐even time’, Journal of Product Innovation Management, 12(1), pp. 54-69.
Algina, J. and Olejnik, S. (2003) ‘Sample size tables for correlation analysis with applications in partial correlation and multiple regression analysis’, Multivariate Behavioral Research, 38(3), pp. 309-323.
56 Bernhart, W. and Winterhoff, M. (2016) ‘Autonomous Driving: Disruptive Innovation that Promises to Change the Automotive Industry as We Know It’, in Langheim, J. (ed.) Energy
Consumption and Autonomous Driving. New York, NY: Springer, pp. 3-10.
Boeglin, J. (2015) ‘The costs of self-driving cars: reconciling freedom and privacy with tort liability in autonomous vehicle regulation’, Yale JL & Tech, 17, pp. 171-203.
Borenstein, J., Herkert, J. and Miller, K. (2017) ‘Self-driving cars: Ethical responsibilities of design engineers’, IEEE Technology and Society Magazine, 36(2), pp. 67-75.
Briley, D. A., Morris, M. W. and Simonson, I. (2000) ‘Reasons as carriers of culture: Dynamic versus dispositional models of cultural influence on decision making’, Journal of Consumer
Research, 27(2), pp. 157-178.
Bullmore, J. (2000) ‘Alice in Disneyland, a creative view of international advertising’, in Jones, J. P. (ed.) International advertising: Realities and myths. Thousand Oaks, CA: Sage, pp. 41-56.
57 Calantone, R. J., Chan, K. and Cui, A. S. (2006) ‘Decomposing product innovativeness and its effects on new product success’, Journal of Product Innovation Management, 23(5), pp. 408-421.
Camit, M., Noble, G. I. and Algie, J. (2009) ‘A comparison of factual/low context versus emotional/high context message appeals: persuading Chinese smokers to ring the Chinese quitline’, Proceedings of the Australian and New Zealand Marketing Academy
Conference, pp. 1-8.
Chandy, R. K. and Tellis, G. J. (1998) ‘Organizing for radical product innovation: The
overlooked role of willingness to cannibalize’, Journal of Marketing Research, 35(4), pp. 474-487.
Cleveland, M. and Laroche, M. (2007) ‘Acculturation to the global consumer culture: Scale development and research paradigm’, Journal of business research, 60(3), 249-259.
Chao, C. W., Reid, M. and Mavondo, F. T. (2012) ‘Consumer innovativeness influence on really new product adoption’, Australasian Marketing Journal (AMJ), 20(3), pp. 211-217.
58 Cowden, B. J., and Kalliny, M. (2013) ‘The role of culture in developing disruptive innovation in domestic firms’, in Christiansen, B., Turkina, E. and Williams, N. (ed.) Cultural and
Technological Influences on Global Business. Hershey, PA: IGI Global, pp. 258-271.
Crowther, D. and Lancaster, G. (2008) Research Methods: A Concise Introduction to Research
in Management and Business Consultancy. 2nd edn. London: Routledge
Czinkota, M. R. and Ronkainen, I. A. (1998) International marketing. 5th edn. Fort Worth: Dryden Press.
Daghfous, N., Petrof, J. V. and Pons, F. (1999) ‘Values and adoption of innovations: a cross-cultural study’, Journal of Consumer Marketing, 16(4), pp. 314-331.
Dancey C. P. and Reidy J. (2007) Statistics without Maths for Psychology. Harlow: Pearson Education.
Mooij, M. D. (2003) ‘Convergence and divergence in consumer behaviour: implications for global advertising’, International Journal of advertising, 22(2), pp. 183-202.
De Mooij, M. (2010) Consumer behavior and culture: Consequences for global marketing and
59 De Mooij, M. and Hofstede, G. (2002) ‘Convergence and divergence in consumer behavior: implications for international retailing’, Journal of retailing, 78(1), pp. 61-69.
Dee Dickerson, M. and Gentry, J. W. (1983) ‘Characteristics of adopters and non-adopters of home computers’, Journal of Consumer Research, 10(2), pp. 225-235.
Dewar, R. D. and Dutton, J. E. (1986) ‘The adoption of radical and incremental innovations: An empirical analysis’, Management science, 32(11), pp. 1422-1433.
Engel, J.F., Blackwell, R.D. and Miniard, P.W. (1995) Consumer Behaviour. 8th edn. Fort Worth, TX: The Dryden Press.
European Commision (2015) ‘Special Eurobarometer 427, Autonomous Systems’. Available at: http://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_427_en.pdf (Accessed: 23 March 2019)
European Commision (2017) ‘Special Eurobarometer 460, Attitudes towards the impact of digitisation and automation on daily life’. Available at:
60 Fagnant, D. J. and Kockelman, K. (2015) ‘Preparing a nation for autonomous vehicles:
opportunities, barriers and policy recommendations’, Transportation Research Part A: Policy
and Practice, 77, pp. 167-181.
Gatignon, H., and Robertson, T. S. (1991) ‘Innovative Decision Processes’, in Robertson, T. S. and Kassarjian, H. H. (ed.) Handbook of Consumer Behavior. Englewood Cliffs, NJ: Prentice-Hall, pp. 316-348.
Gatignon, H., Tushman, M. L., Smith, W. and Anderson, P. (2002) ‘A structural approach to assessing innovation: Construct development of innovation locus, type, and characteristics’,
Management science, 48(9), pp. 1103-1122.
Ger, G. (1999) ‘Localizing in the global village: Local firms competing in global markets’,
California Management Review, 41(4), pp. 64-83.
Ghasemi, A. and Zahediasl, S. (2012) ‘Normality tests for statistical analysis: a guide for non-statisticians’, International journal of endocrinology and metabolism, 10(2), pp. 486.
Green, S. B. (1991) ‘How many subjects does it take to do a regression analysis’, Multivariate
behavioral research, 26(3), pp. 499-510.
61 Hall, E. T. (1959) The silent language. Garden City, NY: Doubleday.
Hall, E. T. (1976) Beyond Culture. Garden City, NY: Anchor Press.
Hervás-Oliver, J. L., Albors-Garrigos, J., Estelles-Miguel, S. and Boronat-Moll, C. (2018) ‘Radical innovation in Marshallian industrial districts’, Regional Studies, 52(10), pp. 1388-1397.
Ho, D. Y.-F. and Chiu, C.-Y. (1994) ‘Component ideas of individualism, collectivism, and social organization: An application in the study of Chinese culture’, in Kim, U., Triandis, H. C.,
Kâğitçibaşi, Ç. Choi, S.-C. and Yoon, G. (ed.) Individualism and collectivism: Theory, method,
and applications. Thousand Oaks, CA: Sage, pp. 137-156.
Hoeffler, S. (2003) ‘Measuring preferences for really new products’, Journal of Marketing
Research, 40(4), pp. 406-420.
Hofstede, G. (1980) ‘Motivation, leadership, and organization: do American theories apply abroad?’, Organizational dynamics, 9(1), pp. 42-63.
Hofstede, G. (1984) ‘Cultural dimensions in management and planning’, Asia Pacific journal of
management, 1(2), pp. 81-99.
62 Hofstede, G. (2011) ‘Dimensionalizing cultures: The Hofstede model in context’, Online
readings in psychology and culture, 2(1), pp. 1-26.
Hofstede, G. (2014) Individualism/Collectivism in 10 minutes. Available at:
https://geerthofstede.com/wp-content/uploads/2016/08/Individualism-Collectivism-in-10-minutes-2015-09-05.pptx (Accessed: 16 May 2019).
Hofstede Insights (2019) Country Comparison. Available at: https://www.hofstede-insights.com/country-comparison/ (Accessed: 18 April 2019).
Holak, S. L. and Lehmann, D. R. (1990) ‘Purchase intentions and the dimensions of innovation: An exploratory model’, Journal of Product Innovation Management, 7(1), pp. 59-73.
Holt, K. (1976) ‘Need assessment in product innovation’, International Studies of Management
& Organization, 6(4), pp. 26-44.
Hornikx, J. and le Pair, R. (2017) ‘The influence of high-/low-context culture on perceived Ad complexity and liking’, Journal of Global Marketing, 30(4), pp. 228-237.
63 Human Development Index (2018) ‘Human Development Indices and Indicators: Statistical Update 2018’. Available at:
http://hdr.undp.org/sites/default/files/2018_human_development_statistical_update.pdf (Accessed: 15 May 2019).
Hurmerinta, L. and Sandberg, B. (2015) ‘Sadness bright as glass: the acceptance of emotionally sensitive radical innovation’, Journal of Marketing Management, 31(9-10), pp. 918-939.
Im, S., Bayus, B. L., and Mason, C. H. (2003) ‘An empirical study of innate consumer innovativeness, personal characteristics, and new-product adoption behavior’, Journal of the
academy of marketing science, 31(1), pp. 61-73.
Jaworski, B., Kohli, A. K. and Sahay, A. (2000) ‘Market-driven versus driving markets’, Journal
of the academy of marketing science, 28(1), pp. 45-54.
Kaasa, A. and Vadi, M. (2010) ‘How does culture contribute to innovation? Evidence from European countries’, Economics of innovation and new technology, 19(7), pp. 583-604.
Kapoor, K. K., Dwivedi, Y. K., and Williams, M. D. (2014) ‘Rogers’ innovation adoption attributes: A systematic review and synthesis of existing research’, Information Systems
