The cultural influence of Dutch and Polish consumers’ beliefs on the online purchase intention of personalized products

(1)

The cultural influence of Dutch and Polish

consumers’ beliefs on the online purchase intention

of personalized products

Master Thesis by

Artur Olszewski

under the supervision of

Adriana Krawczyk, PhD

MSc. in Business Studies - Marketing Track University of Amsterdam

Faculty of Economics and Business

29th June 2015 Amsterdam

(2)

2

Statement of originality

This document is written by Student Artur Olszewski 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.

(3)

3

List of tables and figures

Tables

Table 1 Demographic profiles of the Dutch and Polish respondents 37

Table 2 Reliability of the constructs 38

Table 3 Means, Standard Deviations and Correlations for Dutch consumers 39 Table 4 Means, Standard Deviations and Correlations for Polish consumers 39

Table 5 Analysis of the models and interaction effects 41

Table 6 Analysis of the variables within the model H1 41

Table 7 Analysis of the variables within the model H4 43

Table 8 An overview of the study 43

Figures

Figure 1 The scores on dimensions of Dutch and Polish individuals 24

Figure 2 Conceptual model 29

Figure 3 Conceptual diagram of Hayes’ model 1 40

Figure 4 Statistical diagram of Hayes’ model 1 40

(6)

6

Abstract

Managers and academics pay a lot of attention to consumers’ beliefs towards offered products on the Internet. However, little is known about how beliefs, influenced by cultural values, influence the online purchase intention of consumers, especially when personalized products are offered. Personalization means that the design of a product meets someone’s individual requirements and can be identified as belonging to a particular person. This happens by marking a product with his/her name or a self-selected (for example own) picture. In order to test consumers’ beliefs and the influence of cultural values on the online purchase intention, a conceptual framework is developed and tested. This paper focusses on Dutch and Polish consumers. These two countries are selected due to different scores on Hofstede’s dimensions and on the NRI ranking. The NRI ranking reflects the ICT infrastructure of a country. It is hypothesized that beliefs of Dutch and Polish consumers are moderated by Hofstede’s dimensions Individualism and Uncertainty Avoidance. Results from a sample of 231

respondents from both countries indicate that Individualism moderates beliefs of Dutchmen, while Uncertainty Avoidance moderates beliefs of Poles. Gender was incorporated in the model as control variable, however significant influence is not found. The findings are

presented and discussed. Also, managerial implications and future research recommendations are provided.

Acknowledgement: I would like to thank Adriana Krawczyk for supervising me during the process of

writing this thesis. She provided valuable feedback and suggestions that contributed to this final

version. I would like to thank Tomek Dabrowski for corrections, design and motivation meetings. I

would also like to thank my friends and family for their help with collecting the data. Finally, I would

(7)

7

1. Introduction

The World Wide Web (WWW) and the Internet have been two of the most important

developments of information technology in the last years, enabling companies to expand and develop business models using e-commerce (Aitchison & Stone, 2002). Both recent

developments have revolutionized marketing and the business world; now the Internet enables companies to provide various services, such as communication, information access,

entertainment, banking, advertising, education, buying and selling (Samiee, 1998). Use of the Internet gives a lot of opportunities for businesses in the areas of customer service, lower transaction costs, marketing, customer attraction and retention (Al Kailani & Kumar, 2011). Also, it increases the reach of businesses and enables them to operate worldwide (Al Kailani & Kumar, 2011).

Today, it is not a surprise that consumers buy products or services through direct interaction with online stores (Park & Kim, 2003). The way in which consumers communicate with each other and how they look for information has changed (Ranaweera et al., 2008). The process to search, to select and to buy products online is improved (Zhang & Prybutok, 2003). According to Al Kailani & Kumar (2011), the worldwide number of Internet users is rapidly growing; from 2 million in 1990, 45 million in 1995 and 430 million in 2000, it surpassed 1.2 billion in 2006 and it was estimated that in 2011 there were 2 billion Internet users. Internet World Stats (2014) reports that 40,7% of the world’s population uses the Internet, and this number rises daily (Ashraf et al., 2014). According to Ashraf and others (2014) it is forecasted that by 2016, the Internet economy will grow to $4.2 trillion in the G-20

economies. Although many Internet experts are optimistic about the future of online business, online shopping has its unique problems compared to traditional brick-and-mortar stores (Al Kailani & Kumar, 2011). While a lot of studies have been conducted already, which are described later in this paper, online shopping and e-commerce are still a relatively new

(8)

8

academic topic where only a few robust, validated, and replicated results are found (Zhang & Prybutok, 2003; Zheng, 2013).

Despite e-commerce being successful in European Union countries, the level of Internet use and ICT infrastructure differs across the EU-countries (Seybert, 2011). These findings are reported in the Global Information Technology Report (Bilbao et al., 2013), which ranked many Central- and Eastern European countries much lower in terms of their network readiness compared to Western Europe. Network readiness refers to a country’s ability to exploit opportunities offered by information and communications technology, which is based on the Networked Readiness Index (NRI) (Ashraf et al., 2014). The NRI includes features related to access to and usage of ICT infrastructure, digital resources (software or skills), while also providing proxies to calculate the social and economic influences of a current state of the ICT developments in a certain country (Bilbao et al., 2013). For example, according to this report, the NRI score of the Netherlands is fourth in the world, while Poland is 49th in this ranking. It means that the ICT infrastructure in Poland is poorer than the ICT development in West-European countries (Finland, United Kingdom and Germany score respectively first, seventh and thirteenth). Even though Poland has such a low score on the NRI, this Central Europe country is currently classified as the leader of European e-commerce growth (Thiessen, 2012). Here, a question arises: given the existing differences in ICT

infrastructure across countries, how do individuals in the Netherlands and Poland use the Internet?

According to Seybert (2011), the Internet has become important for daily life, education, work and participation in the society. However, he states that differences exist in access to and usage of the Internet within European Union countries, especially between Western and Central Europe. To compare, in 2011, 92% of the Dutch population and 65% of Poles used the Internet within the last 12 months (Seybert, 2011). These numbers did not

(9)

9

change much in 2014: 94% versus 69%, for the Netherlands and Poland, respectively (Eurostat, 2014). It is also useful to compare how many individuals have never used the Internet: from 7% versus 33% in 2011 to 5% and 28% in 2014 for the Netherlands and Poland, respectively (Eurostat, 2014). From these percentages it can be concluded that although almost everyone in the Netherlands has used the Internet, for Poles this is not the case. According to EcommerceNews.eu (2013), Poland is the fastest growing e-commerce market in the European Union. How do the statistics presented above about Internet usage relate to online purchasing in the Netherlands and Poland?

Based on the statistics available at the website of Eurostat (2014) about online purchases made by Dutch and Polish consumers in 2014 it becomes evident that 71% of Dutchmen purchased online within 12 months, while only 34% of Polish people had shopped online in last year. As Seybert (2011) suggested, the Internet has become an important element of daily life of European consumers, but the usage and online purchasing differs between the countries of the European Union. Therefore, it is interesting to study where those differences in online shopping behavior come from.

1.1. Relevance of the study and the literature gap

In order to explain the differences between Dutch and Polish online consumers the beliefs of consumers should be studied (Zheng, 2013). Research on consumer behavior is mainly based on theories that forecast and explain human behavior (Zheng, 2013). These consumer

purchase theories contribute to a better understanding of the mechanism of consumer behavior and provide the theoretical guide (Zheng, 2013). The related theories that measure consumers’ beliefs are discussed in the literature review.

Although a lot of research on consumer behavior is done, little is known about the role of culture on the online purchase intention (Chai & Pavlou, 2004; Zheng, 2013). Based on the

(10)

10

findings from the literature it is suggested that cultural values can influence the usage of Internet and the online shopping behavior (Overby et al., 2004; Chang et al., 2005; Yoon, 2009; Wang & Sun, 2010; Kim et al., 2013; Zheng, 2013; Ashraf et al., 2014; Petersen et al., 2014;Muk & Chung, 2015). The main findings on this topic are described in the literature review. The aim of this study is a better understanding of the role of cultural values on online shopping behavior. In this paper the cultural values are incorporated into the model, which investigates whether or not consumers’ beliefs influence the online purchase intention.

The recommendations from the literature are used in this study. Firstly, it is not known which cultural values influence the online shopping behavior of consumers (Chai & Pavlou, 2004; Forsythe et al., 2006; Kim et al., 2013; Smith et al., 2013; Belkhamza & Wafa, 2014). Even when beliefs and cultural values are investigated to test whether they influence the online shopping behavior, there is a critique that Western perspectives dominate this research, where mostly the American population is used as a sample (Cayla & Arnould, 2008;

Cleveland et al.; Smith et al., 2013). Secondly, it is recommended to use consumers from two or more European countries in order to investigate the differences in beliefs and cultural values that influence the online shopping behavior (Seybert, 2011; Smith et al., 2013; Zheng, 2013). It is suggested in the literature to study countries that are culturally different (Chai & Pavlou, 2004; Al Kailani & Kumar, 2011; Kim et al., 2013; Smith et al,, 2013; Belkhamza & Wafa, 2014). It is also advised to select more than one country in order to maintain the similar setting of a study, which enables an objective comparison between consumers from more countries (Ng, 2013).

As a sample the Dutch and Polish population are chosen. The Netherlands have been chosen, because this country possess one of the most developed digital infrastructures, which is reflected in a high NRI score, the fourth highest in the world according to Global

(11)

11

chosen for this study is Poland. Regarding the NRI, Poland is ranked much lower than the Netherlands, which is reflected in Poland’s ranking (49th

) in the ranking of the Global Information Technology Report (Bilbao et al., 2013). However, Poland is one of the biggest transition economies with the fastest growth rate of e-commerce development in Europe (EcommerceNews.eu, 2013; Gemius, 2014). According to the report on e-commerce in Poland made by Gemius (2014), it is stated that Polish Internet users ‘love’ online shopping. It is reported that 78% of them visit e-commerce related websites, which is the highest score in Central Europe. The analysts from Gemius argue that ‘the boom’ in the e-commerce market in Poland is still ahead (2014). It is also concluded that Poland is an attractive market for foreign investors (Gemius, 2014). Therefore, from the academic and managerial perspectives it becomes interesting to select and compare consumers from the Netherlands and Poland in this study. Another argument to choose the Netherlands and Poland in the study are the different scores of both countries on cultural values. These differences are described in the literature review. Soares and others (2007) argue that differences in scores on cultural values lead to differences in online shopping behavior. Therefore, it is interesting to study how the cultural values of Dutch and Polish consumers relate to online shopping behavior.

The most popular and useful framework that is used to investigate the role of culture is Hofstede’s dimensions (Zheng, 2013). This framework is used in different contexts and settings in a lot of studies concerning the role of culture (Choi & Geistfeld, 2004; Lim et al., 2004; Singh et al., 2005; Moon et al., 2008; Gong, 2009; Yoon, 2009; Cho & Wang; 2010; Al Kailani & Kumar, 2011; Hwang, & Lee, 2012; Kim et al., 2013; Ng, 2013; Smith et al, 2013; Zheng, 2013; Ashraf et al., 2014; Belkhamza & Wafa, 2014; Griffith & Rubera, 2014;

Petersen et al., 2014). It is suggested by several scholars to use the Hofstede’s dimensions to study online shopping behavior (Gong, 2009; Al Kailani & Kumar, 2011; Hwang & Lee, 2012; Smith et al., 2013; Belkhamza & Wafa, 2014; Makhitha & Dlodlo, 2014). This paper

(12)

12

incorporates two of Hofstede’s dimensions in the conceptual model: Individualism and Uncertainty Avoidance. These cultural values have been chosen, because it is suggested that both dimensions can moderate the influence of consumers’ beliefs on the online purchase intention (Chai & Pavlou, 2004; Moon et al., 2008; Belkhamza & Wafa, 2014).

This paper focusses on products that are personalized. Personalization means that the design of a product meets someone’s individual requirements and can be identified as

belonging to a particular person (Moon et al., 2008). This happens by marking a product with his/her name or a self-selected (for example own) picture. This type of product has been selected, because online personalization provides a new direction in the long-lasting but unresolved debate in the literature about the superiority of standardized or customized products (Moon et al., 2008). It is stated that personalization can benefit from the advantages of both standardized and customized strategies because “it offers individually tailored

products at costs that are almost the same as that of standardized production and mass marketing” (Moon et al., 2008, p.32). Therefore, online personalization seems to be feasible, plausible and profitable for larger and smaller firms (Kramer et al., 2007; Moon et al., 2008). Furthermore, the personalization options enable firms to ask higher prices and increase margins on its products compared to selling standardized products (Moon et al., 2008). Another argument to study personalized products is the growing interest of consumers to buy this type of products online (Moon et al., 2008; Lee et al., 2011). However, little is known about the influence of beliefs and culture on the online purchase intentions of personalized products (Chai & Pavlou, 2004; Goldsmith & Freiden, 2004; Moon et al., 2008; Cho & Wang, 2010; Park et al., 2013). The scholars suggest studying this topic in a cross-cultural setting (Goldsmith & Freiden, 2004; Moon et al., 2008; Cho & Wang, 2010; Sabiote et al., 2012; Li, 2013). This recommendation is met in this study, in which the influence of Dutch and Polish

(13)

13

consumers’ beliefs, moderated by their cultural values, on the online purchase intention of personalized products are investigated. This leads to the following research question:

Do beliefs of Dutch and Polish consumers, moderated by Hofstede dimensions

Individualism and Uncertainty Avoidance, influence the online purchase intention of

personalized products?

The answer on this question will contribute to the academic and managerial world. From the academic perspective, this paper will perform a study, which meets the recommendations of several papers (Chai & Pavlou, 2004; Moon et al., 2008; Gong, 2009; Cho & Wang, 2010; Al Kailani & Kumar, 2011; Hwang & Lee, 2012; Kim et al., 2013; Li, 2013; Smith et al., 2013; Belkhamza & Wafa, 2014; Makhitha & Dlodlo, 2014). This paper will contribute to a better understanding of the influence of consumer’s beliefs and cultural values on personalized products offered online. For the managers, the answer on this question is highly relevant, because it suggests whether and how firms should decide to sell personalized products. Also, these findings facilitate the way that firms should act in order to meet the expectations based on consumers’ beliefs and their cultural values.

After the literature review where the theoretical components and concepts are explained, the theoretical framework with the conceptual model and the corresponding hypotheses are proposed. Then, research methodology and results are presented. After that, the findings are discussed. Finally, the theoretical and managerial contributions will be provided, the same as the recommendations for future research.

(14)

14

2. Literature Review

2.1. Consumers’ beliefs

As stated in the introduction, in order to explain the differences between Dutch and Polish online consumers the beliefs of consumers should be investigated (Zheng, 2013). In order to analyze consumer behavior the theories that explain human behavior are used (Zheng, 2013). The topic ‘consumer behavior’ is well covered by the literature; particularly the theory of Reasoned Action, the theory of Planned Behavior, and the Technology Acceptance Model are used to study online shopping behavior (Gong, 2009; Sabiote et al., 2012; Zheng, 2013). From these named theories the theory of Planned Behavior (in this paper also called TPB alternately) represents people’s beliefs most accurate (Zheng, 2013). Therefore, in this paper it is chosen to rely on the TPB and use components of this theory as independent variables in this study. In order to test the online purchasing intentions, the components of the TPB are used and approved by several scholars (Chai & Pavlou, 2004; Choi & Geistfeld 2004; Chang et al., 2005; Kang, 2008; Andrews & Bianchi, 2013; Kang & Kim, 2013; Zheng, 2013; Ashraf et al., 2014, Belkhamza & Wafa, 2014). In order to understand the TPB it is important to explain firstly the theory of Reasoned Action, which forms the fundamental concept for the TPB (Fishbein & Ajzen, 1975).

(15)

15

2.1.1. The Theory of Reasoned Action

The theory of Reasoned Action states that consumer behavior can be determined by the consumer’s ‘behavioral belief’, which is a function of the ‘attitude toward behavior’ and the ‘normative beliefs’ which is a function of ‘subjective norms’ (Hansen et al., 2004). According to Hansen and others (2004), it means that this theory “predicts intention to perform a

behavior by consumer’s attitude toward that behavior rather than by consumer’s attitude toward a product or service” (p.540). However, it is also possible that “intentions to perform a certain behavior may be influenced by the normative social beliefs held by the consumer” (Hansen et al., 2004, p.540). The Theory of Reasoned Action emphasizes the causal links among beliefs and intention, and can be applied in the context of online consumer behavior (Vijayasarathy, 2002; Zheng 2013). However, the theory of Reasoned Action has the limitation to deal with a behavior over which individuals have no control (Ajzen, 1991; Zheng, 2013). This factor over which an individual has no control is added in the Theory of Planned Behavior, which is described in the next subsection.

2.1.2. The Theory of Planned Behavior

According to the Theory of Planned Behavior, people’s actions are determined by their intentions, which are influenced by attitudes, subjective norm and perceived behavioral control (Ajzen, 1991; Zheng 2013). Perceived behavioral control means “the perception of internal and external resource constraints on performing the behavior” (Ramayah et al., 2009, p.1224). Control beliefs reflect the perceived difficulty (or ease) with which the behavior may be affected and the perceived facility acts as an important weighting (Ajzen, 1991). In other words, the TPB can be conceptualized as “the consumer’s subjective belief about how difficult it will be for that consumer to generate the behavior in question” (Hansen et al., p.541). According to Hansen and others (2004), recent studies indicate that consumers

(16)

16

perceive obstacles and difficulties when they decide to shop online. Therefore, it is suggested that scholars should choose and incorporate the TPB rather than the theory of Reasoned Action (Hansen et al., 2004). Another reason is that online shopping “does require skills, opportunities, and resources, and thus not occur merely because consumers decide to act” (Shim et al., 2001, p.413). This suggests the important role of the additional factor subjective norms, that represents control belief over which an individual has limited or no control (Zheng, 2013).

The three components of the TPB will be incorporated in this study. As written above, these components are: the attitude toward behavior, which is reflected in the behavioral belief, subjective norm which is reflected in the normative belief and perceived behavioral belief which is reflected in the control belief (Chai & Pavlou, 2004; Gong, 2009; Zheng, 2013). The influence on the online purchase intention is found for all three components of the TPB in the literature which means that all three factors are reliable predictors of online purchase intention (Chai & Pavlou, 2004; Choi & Geistfeld 2004; Chang et al., 2005; Kang, 2008; Andrews & Bianchi, 2013; Kang & Kim, 2013; Zheng, 2013; Ashraf et al., 2014, Belkhamza & Wafa, 2014).

(17)

17

2.2. The cultural context in the literature

As written in the previous section, the online consumer behavior is well-explained topic in the literature but relatively few studies analyze the influence of culture (Cheung et al., 2005). According to Cheung and others (2005), prior studies reveal that if cultural values and consumer behaviors differ, then the marketing strategies should also be adapted, accordingly to these differences. Especially when the personalized products are offered it is recommended to understand consumer characteristics and their needs (Al Kailani & Kumar, 2011). This is a call to investigate the cultural influence of consumers’ beliefs on the online purchase

intention. This will help in designing websites and products in a responsive way to meet the cross-cultural preferences of online consumers (Al Kailani & Kumar, 2011). It is also argued that the global spread of online purchasing requires a more systematic understanding of the aspects of online consumer behavior, such as the cultural values (Smith et al., 2007). Due to a limited amount of studies on e-commerce adoption across different cultures, the existing support for the factors that influence online shopping behavior remains equivocal (Ashraf et al., 2014). Also, the literature was focused on the US market and it has been criticized for the implications that can be applied only in the Western, well-developed countries (Cayla & Arnould, 2008; Cleveland et a., 2009; Smith et al., 2013). It is argued that the applicability of the Western perspective, theories and the theoretical interrelationships between online

shopping experience, evaluation and the intention to shop online to different cultures is limited (Nguyen & Barrett, 2006). Even though there is an increase in wealth, and in the use of highly advanced technological devices across nations, it is stated that culture still can be seen as a stable predictor of Internet consumption behavior in different countries (Goodrich & de Mooij, 2011; Kim et al., 2013). Furthermore, while beliefs measure people’s attitudes on the individual level, culture reflects people’s attitudes on a group-level (Zheng, 2013). It is thus important to study consumers’ attitudes on online purchasing intention on both levels

(18)

18

(Zheng, 2013). As stated in the introduction, it is recommended to study the online shopping behavior in more cultural contexts (Ashraf et al., 2014) In order to investigate the role of culture the most useful tool is Hofstede’s dimensions (Al Kailani & Kumar, 2011). After the presentation of Hofstede’s dimensions the findings on this topic are described.

2.2.1. Cultural dimensions of Hofstede

It is argued that a national culture is an important factor of many dimensions of human behavior and decision-making (Soares et al., 2007; Zheng, 2013). The Hofstede’s framework of cultural dimensions can be used to study the cultural values (Lim et al., 2004; Smith et al., 2013; Zheng, 2013). There are also recent studies where Hofstede dimensions were used (Choi & Geistfeld, 2004; Lim et al., 2004; Singh et al., 2005; Moon et al., 2008; Gong, 2009; Yoon, 2009; Al Kailani & Kumar, 2011; Hwang, & Lee, 2012; Kim et al., 2013; Smith et al, 2013; Ashraf et al., 2014; Griffith & Rubera, 2014; Petersen et al., 2014).

Hofstede (1980) defined a culture as “the collective programming of the mind, which distinguishes the members of one human group from another’’ (p.125). He identified four cultural dimensions: Individualism–Collectivism (in this paper called Individualism),

Uncertainty Avoidance, Masculinity and Power Distance (Hofstede, 1991). The first cultural dimension, Individualism assesses the relationships between the individual and societal groups, such as family or workplace within a culture (Smith et al., 2013). In individualistic cultures, an individual cares only for himself, whereas in the collectivistic societies,

individuals belong to groups that look after them in exchange for loyalty (Soares et al., 2007). The dimension Uncertainty Avoidance refers to the tolerance of uncertainty and ambiguity within a society (Hofstede, 1991).Masculinity is the extent to which masculine traits, such as competitiveness and wealth accumulation, are valued over feminine values, such as

(19)

19

“the extent to which less powerful members of a society accept and expect that power is distributed unequally” (de Mooij & Hofstede, 2011, p.182). All four dimensions of the Hofstede’s cultural framework are proved, and the empirical results have been replicated by several scholars (Dorfman & Howell, 1988; Triandis, 1995; Matsumoto et al., 1997; Kolman et al., 2003; Merkin, 2006; Gong et al., 2007; Al Khalaini 2014). Now, the main findings about the influence of the Hofstede’s dimensions on the Internet usage and shopping behavior are presented.

2.2.2. Findings on the influence of Hofstede’s dimensions on shopping behavior.

As stated earlier in this paper, among the cultural dimensions proposed in Hofstede's framework, Uncertainty Avoidance and Individualism are the most relevant to Internet shopping (Chai & Pavlou, 2004; Moon et al., 2008; Gong, 2009; Belkhamza & Wafa, 2014).

According to Gong (2009), in order to study the consumer online behavior the relevant cultural dimensions are Individualism and Uncertainty Avoidance. The first one reflects the individual’s propensity of being innovative, while the second dimensions indicates an individual’s perceived risk and trust associated with online purchasing (Gong, 2009). Uncertainty Avoidance results in the extent in which societies can tolerate uncertainty and ambiguity, whereas cultures that score low on uncertainty avoidance exhibit greater tolerance

for risk and people in such societies have inclinations to be more innovative, entrepreneurial and are more willing to try and discover new things (Gong, 2009). Cultures that score high on

Uncertainty Avoidance dimension tend to value factors such as security and clear rules (Lim et al., 2004). Also, they tend to be more resistant to change from established patterns, are risk-averse and less innovative (Gong, 2009). According to Lim and others (2004), countries that score low on the Uncertainty Avoidance dimension and are individualistic tend to show higher Internet shopping rates than countries with a collectivistic culture. The use of both

(20)

20

dimensions is argued due to “their link to the willingness to accept the potential risks of online shopping and to trust unknown online vendors” (Lim et al., 2004, p.546).

The first study that introduces Hofstede’s cultural dimension to the cross-national e-commerce environment was the paper of Chai and Pavlou (2004). Their study was conducted in Greece and the United States in order to better understand of cross-cultural e-commerce adoption. The results show that Uncertainty Avoidance moderates the consumer’s e-commerce adoption (Chai and Pavlou, 2004). Another contribution of their paper was the application of the Theory of Planned Behavior to investigate the behavioral intentions to transact online. Based on the results, the moderating influence of Uncertainty Avoidance was found to exist with respect to subjective norm and for perceived behavior control, but not for attitude (Chai and Pavlou, 2004). It was explained that no moderation effect for attitude might be caused by the age difference between US and Greece samples (Chai & Pavlou, 2004). It is recommended to maintain similar average ages for samples when more samples are used. According to the scholars, the recommendations for the future research should also take into consideration the incorporation of other cultural dimensions, which will provide a better understanding of e-commerce, conducting the future research in other countries (Chai and Pavlou, 2004).

Also the study of Belkhamza and Wafa (2014) that was conducted in Malaysia and Algeria shows that the dimension Uncertainty Avoidance plays a moderating role in the e-commerce adoption process. The aim of their study was to test the relationship between perceived usefulness, attitude, and subjective norm in order to assess the differences of

behavioral intentions to use e-commerce in Malaysia and Algeria (Belkhamza & Wafa, 2014). From the results it seems that the subjective norm was found to be related to e-commerce usage intention. This relationship was weaker in high uncertainty avoidance culture,

(21)

21

dimension (Belkhamza & Wafa, 2014). This finding confirms the results of the study of Chai and Pavlou (2004), which investigated other countries (Greece and the US). Another factor that will be used in this study, attitude, did not have a significant influence on the intention to use e-commerce. From this study it becomes clear that Uncertainty Avoidance has a potential to moderate the intention to use e-commerce. However, the role of Uncertainty Avoidance on the use of e-commerce can differ from the online purchase intention, which is investigated in this study.

Another study that investigates the intention to use e-commerce was conducted by Yoon (2009). Based on a consumer e-commerce acceptance model Yoon (2009) studied the influence of culture as a moderator of the relationship between perceived usefulness,

perceived ease of use, trust as independent variables and the intention to use e-commerce as a dependent variable. From obtained results it is concluded that Uncertainty Avoidance has a moderate effect on the relationship between trust and intention to use (Yoon, 2009). This relationship was moderated negatively towards the intention to use e-commerce.

In another recent study that emphasizes the role of Uncertainty Avoidance on the intention to purchase online is the study of Kim and others (2013). The aim of their study was the investigation of the role of cultural values in the relationship between consumers’

emotional and cognitive reactions and the purchase intention. The study was performed in the US and South Korea. It is suggested that Korean individuals are more influenced by the online retailer’s reputation than US consumers are. In general, the Korean people perceive higher level of risk during online shopping than the US individuals, which is explained by the higher level of Uncertainty Avoidance in South Korea (Kim et al., 2013). The results also indicated that while US consumers are more emotionally stable towards the online vendor’s reputation, the influence of their emotions on purchase intention is greater than that of Korean consumers (Kim et al., 2013). However, the study of Kim and others was focused on a firm’s

(22)

22

reputation as a key external stimulus for purchase intentions. The shortcoming of their study is the majority of respondents were female, which limits the generalization of the findings.

According to Ng (2013), there is a support for the moderating effect of culture and the mediating role of trust in a social network community on the relationship between social interactions and the online purchasing intention. In her study, which is performed in countries in East Asia and Latin America, two dimensions are tested: Uncertainty Avoidance and Individualism. The study has shown that both closeness and trust in a social network

community are reliable predictors of the purchase intention (Ng, 2013). This effect has been moderated by the Individualism dimension (Ng, 2013). The results for the moderating role of the dimension Uncertainty Avoidance were described as “inconclusive” because no

comparable relationship between the two regions from the sample data could be done (Ng, 2013).

According to Al Kailani and Kumar (2011), a limited amount of research has been devoted to investigate the role of Uncertainty Avoidance on online purchase. To test people’s beliefs the Technology Acceptance Model was used. Their study investigated the role of Uncertainty Avoidance and perceived risk on the intention to purchase online in the US, Jordan and India. The results show that individuals from cultures where Uncertainty

Avoidance was high, such as Jordanians, tend to be less likely to make an online purchase (Al Kailani & Kumar, 2011). It has been suggested that cultures with a relatively high level of Uncertainty Avoidance show slow Internet buying adoption rates. Findings also suggest that American consumers that score low on Uncertainty Avoidance tend to show a highest willingness to online purchase among the three cross-national groups (Al Kailani & Kumar, 2011). However, in the paper of Al Kailani and Kumar the moderating role of Uncertainty Avoidance was not investigated. The authors recommend investigating other cultural regions.

(23)

23

They also suggest studying consumers’ online purchase intentions towards customized and personalized products.

Finally, the only study that investigates the effect of cultural dimensions on online shopping behavior when the offered products are personalized rather than standardized is the study of Moon and others (2008). Their study found that the Individualism dimension is the only Hofstede’s dimension to have a significant effect on the purchase intention. It was argued that consumers from countries that score high on Individualism are more likely to buy personalized products on the Internet, which suggests that moderation is positive (Moon et al., 2008). Furthermore, the results also indicate that a product type and price have significant effects (Moon et al., 2008). However, this study has shortcomings; the first one is the chosen method, which is a laboratory experiment where real life context is neglected. The next shortcoming are the places where the study is performed (in New Zealand and Korea), which thus decreases the generalization of the results for the European managers. Therefore, the scholars suggested conducting a study in another cultural region, preferably cross-cultural (Moon et al., 2008). The last shortcoming is the lack of measurements to assess peoples’ beliefs, which are a reliable predictor of consumers’ purchase intentions (Choi & Geistfeld, 2004; Zheng, 2013).

After the presentation of findings about the role of Hofstede’s dimensions on the Internet or e-commerce use and online purchase intention it becomes clear that only

Individualism and Uncertainty Avoidance are relevant for this study. As written above, some moderation effects of both dimensions are already found, but in most cases the influence of dimensions were tested on the intention to use e-commerce, rather than to online purchase. Also, the Technology Acceptance Model was used, rather than the recommended theory of Planned Behavior (Zheng, 2013).

(24)

24

2.2.3. Related cultural values of Dutch and Polish individuals

In order to assess how the cultural values of the Dutch and Polish influence online purchasing decisions it is crucial to know what the scores on the specific Hofstede’s dimensions are (Kolman et al.2003). The differences in scores lead to different online shopping behavior (Soares et al., 2007; Zheng, 2013). The results are presented in figure 1.

Figure 1: The actual scores on the Hofstede’s dimensions Individualism and Uncertainty Avoidance of

Dutch and Polish individuals. Derived from Geert-Hofstede.com (2014).

Based on the results depicted in figure 1, it becomes clear that that there are important differences between the value orientations in The Netherlands, which represents Western Europe and Poland, a Central-European country (Kolman et al., 2003). The results show that the Dutch are individualistic, whereas the Polish are also more individualistic than

collectivistic, but not as individualistic as the Dutch (Kolman et al, 2003). The score of 53 on the dimension Uncertainty Avoidance shows that the Dutch population exhibits a slight preference for avoiding uncertainty (The Hofstede Center, 2014). The score of 53 can be seen

(25)

25

as a moderate score (Kolman et al., 2003). In contrast, the Polish score high on Uncertainty Avoidance, which means that they try to maintain rigid codes of belief and behavior, and could be intolerant to uncommon ideas (The Hofstede Center, 2014). It may result in some resistance of innovation, which can impede the development of the new ideas in Poland (Kolman et al., 2003). Because the Polish society is normative in nature, this society prefers traditional norms and values and view societal changes with suspicion (The Hofstede Center, 2014).

The findings, of the study done by Kolman and others in 2003, confirms the results which are presented in figure 1. It is found that Polish managers score 55 on the dimension Individualism and 85 on the dimension Uncertainty Avoidance (Kolman et al., 2003). The score of 55 on Individualism is explained as the tendency of Polish managers to form a group with strong trust relations, and work together in “beating the enemy” (Kolman et al., 2003). Also, that the individual responsibility is often avoided (Kolman et al., 2003). According to Kolman and others (2003), the score of the Polish on Uncertainty Avoidance is high and can also be seen in the Polish attitude towards authority. Also, the Polish respondents reported often feeling relatively nervous or tense at work (Kolman et al., 2003). To conclude, the results from the study done by Kolman and others, which was published in 2003 show similar results for the Polish individuals as the scores in 2014, derived from Geert-Hofstede.com.

Although, the current literature focusses mostly on investigation of standardized products this study focusses on products that are personalized. The next subsection provides an explanation of personalization of products. Also, the related the findings from the literature are described.

(26)

26

2.3. Product personalization

As stated in the introduction, personalization means that the design of a product meets someone’s individual requirements and can be identified as belonging to a particular person (Moon et al., 2008). This happens by marking a product with his/her name or a self-selected (for example own) picture (Li, 2013). According to Al Kailani and Kumar (2011), the

marketing strategies of firms should be adapted when differences in consumer behavior occur. According to Kramer and others (2007), whereas customization enables consumers to configure a final product by changing product attributes to match their preferences,

personalization “provides consumers with one or more recommended products currently available in the marketplace that match their measured or stored preferences most closely” (p.246). The main idea of treating each individual customer differently is based on the assumption that individuals rely on their own preferences when are making the choices (Kramer et al., 2007). Also, consumers appreciate companies that provide them with product offers that match their personal tastes the most (Kramer et al., 2007). Due to a growing popularity among consumers to personalize products it is recommended to understand consumer characteristics and their changing needs (Kramer et al., 2007; Moon et al., 2008; Cho & Wang, 2010; Lee et al., 2011; Kang & Kim, 2012; Li, 2013). However, little is known about the influence of culture on the relationship between consumers’ beliefs and their online purchase intention (Chai & Pavlou, 2004; Goldsmith & Freiden, 2004; Moon et al., 2008; Cho & Wang, 2010; Park et al., 2013).

The previous research provides empirical reasons that consumers' individual preferences may often be less relevant than the collective preferences of specific targeted groups (Iyengar & Lepper, 1999; Markus & Kitayama, 1991; Kramer et al., 2007). This suggests that consumers who tend to be interdependent or collectivistic may be less

(27)

27

more positively to suggestions that are based on the preferences of other consumers, the target group. Li (2013) conducted a study where tailored and targeted messages were investigated. When a message is created for an individual based on his or her unique characteristics, it is regarded as tailored customization (Li, 2013). This description relates to the concept of personalization (Moon et al, 2008). On the other hand, when a message is designed for a group of people based on their common characteristics, it is termed targeted customization (Li, 2013). This concept relates to customization (Moon et al, 2008). It is found that when individuals were primed with individualistic meanings, they tended to generate a more favorable attitude toward tailored messages than targeted messages. In contrast, when individuals were primed with collectivistic meanings, they formed a more favorable attitude toward targeted messages than toward tailored ones (Li, 2013). However, this study was not based on consumers’ beliefs. Also, little is known about how the obtained results relate to favorable attitudes toward tailored of targeted messages to the online purchase intention of personalized products.

The study of Park and others (2013) is one of a few studies that tested consumers' psychological antecedents for purchasing customized products on the Internet. It is found that a factor perceived risk has a moderating effect on the intention to buy a customized product. From other studies is known that perceived risk is correlated with the dimension Uncertainty Avoidance (Al Kailani & Kumar, 2011; Zheng, 2013). Therefore, even if the role of

dimensions is not tested in the study of Park and others, it might be of importance to present their findings. It was found that consumers who perceive a higher level of risk for purchasing products on the Internet tend to be more dependent on psychological needs when attitudes toward e-customized products are formed than consumers with a lower level of perceived risk (Park et al., 2013). However, it is also suggested that people who scored low on risk

(28)

28

purchase intention toward e-customized products (Park et al., 2013). According to Park and others (2013), this implicates that individuals who experience a high level of risk are not likely to purchase customized products even when they have a positive attitudes toward e-customized products. Also, the authors argue that individuals with a low perceived risk to buy e-customized products online tend to have positive attitudes toward e-customized products. Furthermore, the study revealed that the need for uniqueness and the status aspiration have a significant impact on forming favorable attitudes toward e-customized products (Park et al., 2013). However, there is a difference between customization and personalization of products. While customization refers to tailoring products on a larger scale, personalization includes producing based on individual his or her unique characteristics and preferences (Kramer et al., 2007).

Finally, as mentioned already in the literature review, the only study that investigates the effect of cultural dimensions on online shopping behavior when the offered products are personalized is the study of Moon and others (2008). The findings, recommendations and shortcomings of their study were already discussed above. Briefly, the moderating effect of Individualism on online purchase intention was found, however the generalization of their study is limited to European managers. Also, the tested model was not based on consumers’ beliefs, which is recommended by several scholars (Choi & Geistfeld, 2004; Zheng, 2013).

(29)

29

3. Theoretical Framework

3.1. Conceptual model

In this study it is investigated whether beliefs of Dutch and Polish consumers, moderated by cultural dimensions Individualism and Uncertainty Avoidance, influence the online purchase intention of personalized products. This conceptual model is depicted below.

Figure 2: Conceptual Model of this study

As suggested in the literature review, the Theory of Planned Behavior consists of three components. The first component is the attitude towards behavior, which is reflected in the behavioral belief (Fishbein & Ajzen, 1975; Zheng, 2013). The second component is the consumer’s subjective norm, which is measured with the normative belief (Fishbein & Ajzen, 1975; Hansen et al., 2004; Zheng, 2013). The last component of the TPB is the perceived behavioral belief, which is reflected in the control belief (Hansen et al., 2004; Zheng, 2013). This component is added to account for the subjective thought of consumer about the

difficultness of the task that will be done, which is a realistic assumption as not all types of behavior are equally easy to perform (Hansen et al., 2004). As presented in the conceptual

(30)

30

model, all three belief components are combined into one independent variable. The reason for this is the approved effect of all three components on the online purchase intention (Choi & Geistfeld, 2004). Another reason for this integration is maintaining simplicity.

3.2. Hypotheses

According to Choi and Geistfeld (2004), the Individualism dimension was found to be closely related to the subjective norm and perceived risk for consumers from the US and South Korea when the relationship between culture and the use of e-shopping was investigated. As stated in the literature review, Park and others (2013) also suggested that the need for uniqueness and status aspiration can influence attitudes towards e-customized products. Because

Individualism reflects the need for uniqueness, it is thus implicitly suggested that the relation between beliefs and the online purchase intention of personalized products can be moderated by a score on the Hofstede’s dimension Individualism. Moon and others (2008) found a moderation effect of Individualism on the intention to online purchase a personalized product. This moderation effect was positive. Therefore, it is suggested that consumers’ beliefs

influence the online purchase intention and this relationship is moderated positively by the dimension Individualism. This leads to the following hypotheses:

Hypothesis 1: The score on the Individualism dimension positively moderates the relationship between Dutch consumers’ beliefs and the online purchase intention of personalized products.

Hypothesis 2: The score on the Individualism dimension positively moderates the relationship between Polish consumers’ beliefs and the online purchase intention of personalized

(31)

31

As stated earlier in this paper, the online purchase intention is often associated with the level of perceived risk. According to Al Kailani & Kumar (2011), perceived risk relates to the Uncertainty Avoidance dimension. This dimension was found to act as an antecedent to perceived self-efficacy and perceived risk for each country (Choi & Geistfeld, 2004; Al Kailani & Kumar, 2011). According to Choi and Geistfeld (2004), it is related to perceived risk for Korean consumers. This may suggests that when perceived risk is high, the level of Uncertainty Avoidance is also high. Choi & Geistfeld (2004) found that a high perceived risk and thus a high level of Uncertainty Avoidance impacts the use of the Internet negatively. The website Geert-Hofstede.com (2014) indicates that the score on Uncertainty Avoidance of Korean individuals (85) is comparable to the score of Poles (93). Although attitudes and beliefs towards Internet use may differ from attitudes and beliefs toward online purchase of personalized products, it is expected that a high level of Uncertainty Avoidance negatively moderates this relationship. The results for Americans in the study of Choi and Geistfeld were also significant. Americans score comparable to Dutch individuals on Uncertainty Avoidance (Americans 46, while Dutchmen 53) (The Hofstede Center, 2014). These high scores might indicate that Uncertainty Avoidance negatively moderates the relationship between

consumers’ beliefs and the online purchase intention of personalized products. These findings lead to the following hypotheses:

Hypothesis 3: The score on the Uncertainty Avoidance dimension negatively moderates the relationship between Dutch consumers’ beliefs and the online purchase intention of

personalized products.

Hypothesis 4:The score on the Uncertainty Avoidance dimension negatively moderates the relationship between Polish consumers’ beliefs and the online purchase intention of

(32)

32

4. Research Methodology

4.1. Sample and Procedure

The sample is drawn from Dutch and Polish people who bought online at least one product in the last 12 months. The online questionnaire was administrated at the same time as the hard-copy questionnaires. The data were collected in May and June of 2015. Confidentiality was ensured.

Two versions of the questionnaire, Dutch and Polish, were written. In order to ensure the internal consistency, between both language versions the questionnaire was translated and back-translated. Back translation is recommended by Brislin (1970) and Zheng (2013). The questionnaire consisted of three parts. In the first part, participants were asked about their cultural values. In the second part of the survey, an advertisement about a fictitious firm selling personalized phone covers is presented. The respondents have to read the

advertisement, after which they could continue answering questions about their beliefs and purchase intentions towards the offered product. The last part of the questionnaire included demographical characteristics of the respondents. The English version of the questionnaire is presented in Appendix I.

In total, 252 respondents filled in the questionnaire. Since respondents from other nationalities than Dutch and Polish participated in the survey as well, 6 questionnaires have been excluded from the analysis. Another 15 questionnaires were not completely filled in (less than 50% of the questions was filled in) and these answers have also been excluded from the analysis. This left 231 questionnaires which means that the response rate amounts to 91,67% (231/252 x 100).

In order to test consumers’ beliefs a type of personalized product had to be selected. It is crucial to choose the right product when consumers’ beliefs are measured (Bhatnagar et al., 2000; Choi & Geistfeld, 2004; Zheng, 2013). The role of product type is also investigated in

(33)

33

the study of Moon and others (2008). The findings from the conducted pre-test on Dutch and Polish individuals show that in the questionnaire a ‘useful’, ‘funny’ and ‘relatively cheap’ product has to be tested. Also, the literature suggested that online offered products should be convenient, versatile and unisex (Bhatnagar et al., 2000). Otherwise consumers might

perceive a too high level of risk and the purchase intention would not be measuring the actual consumers’ beliefs and attitudes (Taylor & Todd, 1995; Bhatnagar et al., 2000).

In order to measure the consumers’ beliefs, it is chosen to test the beliefs and attitude towards a personalized phone cover of a fictitious firm ‘YourOwnCover.com’. The logo of the firm and the advertisement in two language versions can be found in Appendix II. Phone covers are often placed in the category ‘apparel’ or ‘accessories’ on clothing websites (Cui et al., 2007). The personalization of phone covers was also studied and recommended by Cui and others (2007). The study of Cui and others was conducted in order to test the regional cultural differences across consumers. To measure cultural differences, the Individualism dimension was used. Therefore, in order to measure the beliefs and online purchase intentions a phone cover of ‘YourOwnCover.com’ was described in the fictitious advertisement.

(34)

34

4.2. Measurements

4.2.1. Independent variables

In order to establish the intention to online purchase of a personalized product, the beliefs towards an offered product have to be established (Taylor & Todd, 1995). The items to measure consumers’ beliefs come from the study of Taylor and Todd (1995). These measurements were also used in the study of Choi and Geistfeld (2004). Beliefs consist of three determinants: behavioral beliefs, normative beliefs and control beliefs, which are related to attitude, subjective norm and perceived behavioral control respectively (Taylor & Todd, 1995; Choi & Geistfeld, 2004; Zheng, 2013). For example, an item that measures normative beliefs was “People who are important to me would think that I should buy the phone cover offered by YourOwnCover.com”. The scale was 1 to 5, where 1 means that an individual ‘strongly disagrees’ and 5 means the individual ‘strongly agrees’ with the statement. In total, 10 items measured the consumers’ beliefs. One item has been recoded. This question

measured the behavioral belief towards the personalized product: “I think it is a bad idea to use the phone cover of YourOwnCover.com”.

4.2.2. Moderating variables

The eight items to measure the Individualism and Uncertainty Avoidance dimensions (four items for each dimension) come from the paper of Yoo and others (2011). These scholars investigated the validation of measurements on Hofstede’s dimensions. The use of the

developed CVSCALE is recommended by Zheng (2013). The items to measure Individualism have been afterwards recoded, because the original measurements test collectivistic attitudes, rather than individualistic, which are relevant for this study. An example item to test

Individualism is: “Group welfare is more important than individual rewards”. The recoded scale is 1 to 5, where 1 indicates that an individual is collectivistic and 5 means they are

(35)

35

individualistic. An example item to test the level of Uncertainty Avoidance is: “It is important to closely follow instructions and procedures”. The scale ranges from 1 to 5, where 5 indicates a high intention to avoid uncertainties and 1 represents low uncertainty avoidance.

4.2.3. Dependent variable

The dependent variable in this study is the intention to online purchase a personalized

product, which is the personalized phone cover of the fictitious firm YourOwnCover.com. All three items that measure the variable ‘purchase intention’ come from the study of Jarvenpaa and others (2000) and were also used in the study of Park and others (2013). An example item is: “How likely is it that you would consider purchasing this personalized phone cover from YourOwnCover.com in the next 3 months?”. The possible answers range from 1 to 5, where 1 means “very unlikely” and 5 means “very likely”.

4.2.4. Control variable

Besides the cultural background and beliefs there are more factors that can influence consumer behavior (Chang et al., 2005). It is recommended to investigate the role of other variables when beliefs and culture are tested (Al Kailani & Kumar, 2011, Zheng, 2013). In this study gender is a control variable.

Brown and others (2003) stated that gender has an impact on online behavior,

suggesting that men are less sensitive to perceived risk during online shopping. According to Rodgers and Harris (2003), gender relates to e-commerce. From the results it seems that women are less satisfied than men with the online shopping experience. This is also

confirmed in other studies (Doolin et al. 2005; Kim & Kim 2004). Alreck and Settle (2002) reported that women’s general attitudes toward online shopping are quite similar to those of men, even though women have much more positive attitudes toward shopping in general

(36)

36

(Liebermann & Stachevsky 2009). However, Bhatnagar and others (2000) found relatively small effects associated with gender. In this study a dummy variable for gender is included (0=female; 1=male).

(37)

37

5. Results

Collected data was analyzed using a computer software package for processing qualitative and quantitative data of social research - SPSS 22. Each variable was checked for missing data. A frequency test was run for all variables and it became clear that the amount of missing values was lower than 1 percent. In the case of missing values mean substitution was

performed. This method is recommended by Field (2014). The scores on all items which form the conceptual model of this study (Beliefs, Individualism, Uncertainty Avoidance and the Online Purchase Intention) are normally distributed.

5.1. Demographic and descriptive analysis

The data that is analyzed in this study describes 103 Dutch respondents, of which 56 are men and 128 Polish respondents, of which 45 are men. The respondents were grouped into three age categories: 15-19, 20-24 and 25 years or older. All respondents bought online at least one product in the last 12 months. More descriptive statistics about the demographical factors of respondents can be found in table 1.

Table 1: Demographic profiles of the Dutch and Polish respondents

Variable Categories Dutch consumers Polish consumers N % N % Gender Males 56 54,4 45 35,2 Females 47 45,6 83 64,8

Age Between 15-19 years 24 23,3 34 26,6 Between 20-24 years 51 49,5 71 55,5 25 years or older 28 27,2 23 18 Education level Elementary school 2 1,9 8 6,3 Lower secondary 40 38,8 12 9,4 Higher secondary 24 23,3 63 49,2 Tertiary education 37 35,9 45 35,2 Online shopping frequency Once a year 4 3,9 15 11,7 Few times per year 56 54,4 77 60,2 Few times per month 35 34 28 21,9 Few times per week 8 7,8 8 6,3

(38)

38

5.2. Reliability analysis

As Saunders et al. (2012) stated, reliability is the extent to which data collection techniques or analysis procedures yield consistent findings. The internal consistency, measured with

Cronbach’s alpha, has been performed to check the usefulness of each item for the analysis. As exhibited in table 2, all items were estimated above the Cronbach’s alpha level of 0.75 which suggest that the measurements are internally highly consistent (Field, 2014).

Table 2: Reliability of the constructs: Cronbach's Alpha

Variable Number of items Dutch consumers Polish consumers Behavioral Beliefs 4 α=0,85 α=0,81 Normative Beliefs 2 α=0,75 α=0,77 Control Beliefs 4 α=0,79 α=0,81 Beliefs Total 10 α=0,78 α=0,83 Individualism 4 α=0,81 α=0,91 Uncertainty Avoidance 4 α=0,89 α=0,89 Purchase Intention 3 α=0,83 α=0,81

As suggested in the theoretical framework, three beliefs components are combined into one variable which is named ‘Total Beliefs’. ‘Total Beliefs’ represents the combination of all 10 items that measured three components of beliefs. The findings show that this variable is internally highly consistent.

5.3. Correlation analysis

Table 3 exhibits the means, standard deviations and correlations between the variables of the Dutch respondents. Similar results for the Polish consumers are available in table 4.

According to the correlation coefficients, for the Dutch respondents the scores on the

Individualism dimensions are associated with beliefs and purchase intention, while for Poles the dimension Uncertainty Avoidance is negatively related to beliefs and purchase intention. For Polish consumers, Individualism is positively related to beliefs. Also beliefs of both nationalities are positively related to purchase intention. The means, standard deviations and

(39)

39

correlations between the related variables are shown in table 3 (for Dutchmen) and table 4 (for Poles).

Table 3: Means, Standard Deviations and Correlations for Dutch consumers

Variables M SD 1 2 3 4

Beliefs 3,67 0,37 0,78

Individualism 4,25 0,44 0,856** 0,81

Uncertainty Avoidance 3,29 0,67 0,048 -0,101 0,89

Purchase Intention 4,33 0,51 0,66** 0,68** 0,014 0,83

Note: **Correlation is significant at the 0,01 level (2-tailed)

Table 4: Means, Standard Deviations and Correlations for Polish consumers

Variables M SD 1 2 3 4

Beliefs 2,55 0,47 0,83

Individualism 3,41 0,71 0,237** 0,91

Uncertainty Avoidance 3,79 0,6 -0,84** -0,318** 0,89

Purchase Intention 2,27 0,61 0,806** 0,15 -0,779** 0,81

5.4. Testing the hypotheses

To test moderating effects, the SPSS plug-in “Process” by Andrew F. Hayes (2012) is used. The aim of the moderation analysis is to determine whether the size of the effect of an independent variable on a dependent variable interacts with a moderator (Hayes, 2013). This method is recently used in the literature (Rosenstreich & Margalit, 2015). In this study, ‘model 1’ is used. According to ‘model 1’, M represents the moderator, X the independent variable, Y the dependent variable, and XM the product of X and M (Hayes, 2013). In this study, X represents beliefs of Dutch and Polish consumers (tested separately), M refers to the cultural dimensions Individualism and Uncertainty Avoidance (also tested separately), and Y is the online purchase intention of personalized product. This model is depicted as a

(40)

40

Figure 3: Conceptual diagram of ‘model 1’ from Hayes (2013)

(41)

41

As stated earlier, four hypotheses are tested. In order to investigate the moderating role of the dimensions Individualism and Uncertainty Avoidance separately with two respondents groups, four moderation analyses are performed. The results of the conducted analyzes for all four models are presented in table 5.

Table 5: Analysis of the models and interaction effects

Model H1 Model H2 Model H3 Model H4 R 0,73 0,81 0,67 0,84 R² of model 0,53 0,66 0,45 0,7 F of model 37,03** 78,91** 26,74** 96,65** R² change due to interaction 0,05** 0,01 0,01 0,02** F change due to interaction 9,58** 1,42 2,11 6,31**

5.4.1. Results for hypothesis 1

The first hypothesis investigates whether the Dutch consumers’ beliefs, moderated by the Individualism dimension, positively influence the online purchase intention of personalized products. Based on the results, which are presented in table 5, the model of hypothesis 1 is significant with F (37,03); p>0,01. Also, the interaction effect is significant with F (9,58); p>0,01. By adding Individualism as a moderator, the model is able to explain 53% of the variance in Purchase Intention. Also, the interaction effect between Individualism and the independent variable Beliefs is significant, F (9,58); p>0,01. Table 6 provides the analysis of the variables that were tested in the model .

Table 6: Analysis of the variables within the model H1

Variables B t p Constant -6,58 -2,77 0,01 Beliefs 2,19 3,49 0,01 Individualism 2,19 3,91 0,02 B x IDV -0,48 -3,1 0,00 Gender 0,49 0,69 0,49

(42)

42

Table 6 suggests that the variable Individualism positively moderates the relationship between the independent variable Beliefs and the dependent variable Purchase Intention, B= 2,19; p<0,01. This means that hypothesis 1 is supported.

The second hypothesis tests the moderating role of Individualism on the relationship between the beliefs of Polish consumers and their online purchase intention. As reported in table 5, the model explains 66% of the variance in Purchase Intention for Polish consumers, and is

significant with F (78,91); p<0,05. However, the interaction effect was not significant, F (1,42); p>0.05. This means that the moderating role of the variable Individualism is not found. No significant interaction effect means that further investigation of the influence of specific variables that were used to test hypothesis 2 is unnecessary. Hypothesis 2 is thus rejected.

Hypothesis 3 studies whether the Uncertainty Avoidance dimension has a negative moderating effect on the relationship between the independent variable Beliefs of Dutch consumers and the dependent variable Purchase Intention. Table 5 shows that although the model is significant, F (26,74); p<0,01, the interaction effect was not, F (2,11); p>0,05. Since the interaction effect was not found, further analysis is not needed. The moderating role of Uncertainty Avoidance is not supported and hypothesis 3 is rejected.

The last hypothesis, testing whether the dimension Uncertainty Avoidance negatively moderates the relationship between the beliefs and the purchase intention of Polish consumers. The findings coming from table 5 report that the model explains 70% of the

(43)

43

variance in the dependent variable Purchase Intention. The model is significant, F (96,65); p<0,01. Also, the interaction effect between Uncertainty Avoidance and the independent variable Beliefs is significant, F (6,31); p>0,01. Since the interaction effect is found, further analysis of the influence of variables which were used in this model is presented in table 7.

Table 7: Analysis of the variables within the model H4

Variables B t p Constant 1,34 1,47 0,01 Beliefs 0,38 1,06 0,29 Uncertainty Avoidance -0,5 -1,97 0,05 B x UA 0,19 1,94 0,05 Gender 0,04 0,67 0,5

It seems that Uncertainty Avoidance negatively moderates (B=-0,5; p=0,05) the relationship between beliefs of Polish consumers and their online purchase intention of personalized products. These findings support hypothesis 4.

5.5. Testing control variable

In all four conducted model analyses the control variable Gender was incorporated. The control variables are called covariates (Hayes, 2012). According to Hayes (2012), scholars include additional variables in moderation models in order to statistically account for shared associations between variables in the causal system caused by other sources. In this study, the proposed control variable has no significant influence on the presented results.

For a clear overview what is investigated in this paper and what the related findings are an overview is provided below.

Table 8: An overview of the study

Hypothesis Moderating variable Nationality of consumers Significance of the interaction effect B of the tested dimension Result

1 Individualism Dutch Yes, F (9,58); p<0,01 B= 2,19; p<0,01

H1 is supported

2 Individualism Polish No, F (1,42); p=0.24 not significant

H2 is rejected

3

Uncertainty

Avoidance Dutch No, F (2,11); p=0,15 not significant

H3 is rejected

4

Uncertainty

Avoidance Polish Yes, F (6,31); p<0,01 B=-0,5; p<0,05

H4 is supported

The cultural influence of Dutch and Polish consumers’ beliefs on the online purchase intention of personalized products