The influence of cultural value at individual level on people’s willingness to engage in electronic word of mouth and the moderating effect of gender differences

(1)

1

Master Thesis

MSc. in Business Administration

Digital Business Track

The influence of cultural value at individual level on people’s

willingness to engage in electronic word of mouth and the moderating

effect of gender differences

Student name: Zhaohui Wang

Student number: 11585579

Supervisor: Angelos Stamos

(2)

2

Statement of Originality

This document is written by Zhaohui Wang who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document are 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

Acknowledgement

Writing this thesis is the final step of obtaining my master’s degree in Business Administration at the University of Amsterdam. The writing process of this thesis is also an interesting learning experience to apply the knowledge obtained from those master courses. I’m more interested in the electronic word of mouth study now after finishing this thesis.

Firstly I would like to thank my supervisor Mr. Angelos Stamos for his guidance, support and valuable feedbacks during the thesis writing process. Then, I want to give my thanks to my family and friends for their understanding and support.

(4)

4

List of Figures

Figure 1: Conceptual Model ……….25

Figure 2: Process Model 1 ………39

List of Tables

Table 1: Summary of measurement………..30

Table 2: Cronbach’s Alpha………...35

Table 3: Means, Standard deviations and Correlations………36

Table 4: Regression analysis of willingness to engage in e-WOM………37

Table 5: Moderator Analysis……….39

Table 6: Moderator Analysis……….40

(7)

7

Abstract

Electronic word of mouth is an influential way of communication in marketing activities. Thus, it’s important for marketers to know what factors would influence consumers’ willingness to engage in e-WOM. Gender and culture both influence consumer behavior. Hofstede’s national cultural dimension has been used for a long time. However, due to globalization and mixed culture society, it is more reasonable to look at cultural value at individual level. This thesis focuses on measuring cultural values at individual level to find if people’s willingness to engage in e-WOM is influenced by individual cultural value and the interaction effect between gender difference and cultural values. Data is collected by an online survey. Results of analyzing the data show that valuing collectivistic and high uncertainty avoidance culture at individual level has a positive impact on people’s willingness to engage in e-WOM. However, interaction effect is not supported. Implications of theoretical study and managerial issues are given for future research and marketers.

Keywords: e-WOM, individual level cultural value, individualism/collectivism, uncertainty avoidance, gender, masculinity/femininity, power distance

(8)

8

Chapter 1 Introduction

From a study made by Word of Mouth Marketing Association (WOMMA) in 2015, word of mouth has a significant impact on driving sales, which drives 13% of sales across categories. According to an interview made by McKinsey in 2017, Bo Finneman, a McKinsey associate partner in Miami, stresses that word of mouth is a key factor on boosting a customer’s initial consideration of a brand. From theses previous researches, we can see word of mouth is becoming an important factor in today’s management and marketing field. However, now we are in the age of web 2.0, consumers start to share their opinions and experience on the Internet (Hennig-Thurau, Gwinner, Walsh & Gremler, 2004). As Hennig-Thurau, Gwinner, Walsh, and Gremler (2004) say in their article, word of mouth has shifted to electronic word of mouth. Nevertheless, according to Channel Advisor (2010), although over 90% of the survey participants read electronic word of mouth, there are only 40% consumers in Europe and 25% consumers in Australia post and write reviews. There are many different factors influencing people’s willingness to engage in electronic word of mouth (Wenguo Shen et al., 2016, Soonyong Bae et al., 2011, T. Sun et al., 2006).

In McCort and Malhotra’s article (1993), they state that culture has a significant influence on consumer behavior and purchase action. People with different cultural value have different lifestyles, product preference and purchase motivations (Tse et al., 1989). Lam, Lee and Mizerski (2009) state that when people engage in word of mouth, the type and pattern of their activity vary a lot from different cultural values. And there are also other studies about cross-cultural differences influencing consumers’ word of mouth behavior such as perception and diffusion of products (Schumann et al., 2010; Dwyer et al., 2005). Different cultural value brings different attitudes towards engaging in electronic word of mouth. Hofstede proposes his four cultural dimensions by

(9)

9 studying the employees of IBM from 53 countries in 1980. However, it has been about 40 years after Hofstede came up with this theory and things change a lot. People in the same country may have different cultural values due to different growing background, education level and so on. Recent researches have criticized that this theory is difficult to use nowadays consistently. More and more people start to study cultural dimension at individual level (Farah, Hackett & Liang, 2007; Yoo, Donthu & Lenartowicz, 2011; Patterson, Cowley & Prasongsukarn, 2006; Mancheno-Smoak, Endres, Potak & Athanasaw, 2009).

When people use the Internet and shop online, there is a significant difference between male and female (Bimber, B. 2000). As Rodgers and Harris (2003) state in their article, females use the Internet less frequently than males. Female consumers are more likely to depend on face to face communication and recommendation. However, more and more females are using the Internet and the number of female online shoppers are also increasing (Yang & Wu, 2006). It is important to increase female consumers’ willingness to engage in electronic word of mouth.

Culture is known as the common knowledge shared by a large group of people collectively and shaped from one generation to the next (EIGE, 2016). Gender is a state of being male or female. Culture and gender both mean one’s social beliefs and values at specific time and society (Burleson, 2003). Ayman and Korabik (2014) find that gender and culture have an impact on how leaders communicate and interact with their colleagues, supervisors and subordinates. In the research of Shan (2014), culture and gender difference have an interaction effect in negotiation, which is a business communication. Kahttab, AI-Manasra, Zaid & Qutaishat (2012) find there is individualist, collectivist and gender moderated difference on intention to purchase online. According to Julia T. Wood (2009), gender, culture and communication are

(10)

10 interlinked to each other. Electronic word of mouth is also a kind of online communication. Thus, we can’t study only one of them without studying the other two. When it comes to electronic word of mouth, previous researches have studied about simple impact of different factors such as culture and gender on people’s willingness to engage in electronic word of mouth (Shen, Huang &Li, 2016; Bae & Lee, 2011; Sun, Youn, Wu & Kuntaraporn, 2006). But few of them study about interactions between those factors and how the interaction effect influences people’s willingness to engage in electronic word of mouth. There is no investigating about the interaction effect between gender difference and cultural values on electronic word of mouth. In this thesis, I will fill this gap and find the effect of cultural value at individual level on people’s willingness to engage in electronic word of mouth and how gender difference moderates this effect.

The potential contributions of my thesis will mainly be two parts. The first part is theoretical contribution. I want to investigate more deeply about factors influencing people’s willingness to engage in electronic word of mouth. I expect to find an interaction effect between culture and gender instead of previous studies which only concentrate on simple factor influencing people’s willingness to engage in electronic word of mouth and ignore the interaction effect between gender and culture. It would be a contribution and provide insights to existing literature about factors influencing people’s willingness to engage electronic word of mouth. And this thesis will also add more insights to the research about Hofstede’s cultural dimension by studying those dimensions at individual level. The result of this research will show how individual cultural values differ from national cultural values. The second part is managerial implication. Companies nowadays pay great attention to electronic word of mouth management. Marketers can develop more targeted strategies to attract consumers of

(11)

11 different gender with different cultural values to engage in electronic word of mouth and build a more targeted online community.

This thesis will be structured as following parts. I will write a literature review about the key concepts and theories of word of mouth, electronic word of mouth, gender and culture. Then, there will be a chapter about theoretical framework and the process of forming hypothesis. The next part is about research design and result. Then it will be the part of discussion and conclusion. At last, managerial implication, theoretical contribution and limitations will be explained

(12)

12

Chapter 2 Theoretical Framework

This chapter is about theoretical framework of this thesis. It will be three parts. First, it will be a literature review and basic information of related key words. Next, the process of developing hypotheses is given. After this, the conceptual model of this thesis will be drawn.

2.1 Literature Review

This part is a comprehensive review of previous literature. Previous studies about key concepts of this thesis will be analyzed. First of all, basic information and development of word of mouth and electronic word of mouth will be given. Then, it will be a review of literatures about gender and culture. Finally, the relationship between gender and culture will be analyzed.

2.1.1 Word of Mouth (WOM)

Word of mouth (WOM) has become more and more common in people’s daily life. And it’s also being used by companies in the marketing field. WOM is a traditional form of electronic word of mouth. Before the rising of Internet, people talk about their experience of products or service face to face, which is an offline communication. So at that time, WOM is defined by Arndt (1967) as an oral, person-to-person communication between a receiver and a communicator whom the receiver perceives as non-commercial, regarding a brand, product or service. We can regard product reviews from family and friends’ daily face-to-face chat as WOM (Nakayama, Wan & Sutcliffe, 2010).

As this kind of recommendation happens in a person-to-person communication, WOM beats other media on credibility and trustworthiness. WOM is more persuasive so it’s more influential (Shen, Huang & Li, 2016). People believe they can get more authentic

(13)

13 information from people they know during purchase decision-making instead of depending on formal and organizational information such as advertising from the company. This makes WOM being a powerful marketing force undoubtedly (Sundaram, Mitra & Webster, 1998).

WOM plays an important role in influencing consumers’ behavior. Both positive and negative WOM can influence consumers’ attitudes and purchase intention (Sundaram, Mitra & Webster, 1998). However, the influence of online and offline WOM differs a lot. According to the study of Bounie, Bourreau, Gensollen and Waelbroeck (2005), there is a significant negative impact on consumers’ purchase decision of offline WOM which means a face-to-face WOM communication. Online WOM which means electronic word of mouth plays a positive role in consumers’ buying behavior.

Nowadays WOM is updated because of the rising of Internet and the importance of online communication. Researches about electronic word of mouth are increasing. Marketers need to catch up the trend and know more about electronic word-of-mouth, which will be helpful to launch marketing activities.

2.1.2 Electronic word of mouth (e-WOM)

Electronic word of mouth (e-WOM) is the development of WOM due to Internet. Hennig-Thurau, Gwinner, Walsh and Gremler (2004) defined e-WOM as “any positive or negative statement made by potential, actual, or former customers about a product or company, which is made available to a multitude of people and institutions via the Internet”. E-WOM shares many same characteristics with WOM, but they still have lots of differences. There are four main differences (Sun, Youn, Wu & Kuntaraporn, 2006).

(14)

14 • Convenience: People can simply write down their reviews and searching for useful reviews on the website for different products and service at the same time. What’s more, e-WOM spreads information through written words. Written communication is more formal than oral one. And people can read reviews at their own speed with written words (Bickart & Schindler, 2001).

• Speed: It’s the dispersion of e-WOM. People can spread e-WOM through one-to-many communication such as product reviews and ratings on the website and many-to-many communication such as chatroom and community (Sun, Youn, Wu & Kuntaraporn, 2006). This makes transmission of the information directly and improves transmission efficiency. The information exchanges happen in multi-ways and don’t need to occur at the same time.

• Reach: By using e-WOM, people can search information and post opinions online at any time without the limit of place. E-WOM expands the influence of opinion and experience sharing (Hinz Skiera, Barrot & Becker, 2011). It is more accessible.

• Absence face-to-face pressure: Unlike WOM is reviews from face-to-face communication, e-WOM is reviews from the Internet and online communities (Duan, Gu & Whinston, 2008). People don’t need to worry about others opinion and relationship. They can type down their true experience on the website without the pressure of a face-to-face communication. Lee, Hosanagar and Tan (2015) believe there is a herding effect in rating behavior. Friends tends to give similar reviews to a same

(15)

15 product, while the crowd with information from weak ties tends to give different views. People can get different aspects of reviews from e-WOM. But this can also lead to the trustworthiness problem of e-WOM because e-WOM “takes place in an anonymous asynchronous online environment” (Davis & Khazanchi, 2008) and you don’t know the person writing the review.

Furthermore, e-WOM is more measurable than traditional WOM (Cheung & Thadani, 2012). For e-WOM, it is easier to be retrieved online as quantitative data. E-WOM can be divided into two forms: numerical rating and opinionated review. Numerical rating can be a quantitative form such as a five-star rating and opinionated review can be a qualitative form such as comments or written reviews (Mudambi & Schuff, 2010). People can get other’s opinions and attitudes to a product or service intuitively by numerical rating. From opinionated review, people can know more and further information about the product or service. When we consider about people providing e-WOM, both numerical rating and opinionated review should be taken into account. E-WOM is more observable than traditional E-WOM according to previous research (Cheung & Thadani, 2012).

People’s willingness to engage in e-WOM can be triggered by several factors. There are eight main factors influencing people’s willingness to engage in e-WOM: problem-solving support, venting negative feelings, concern for others, positive self enhancement, social benefits, economic incentives, helping the company and advice seeking (Hennig‐Thurau, Gwinner, Walsh & Gremler, 2004). According to Lovett, Peres and Shachar (2013), reasons for people engaging in e-WOM are expressing themselves to others which is a social motive and sharing information which is a functional motive. From Shen, Maceli, Zhao and Baack’s research and Ma’s research,

(16)

16 the relationship between concern for others, advice seeking and culture difference is significant and the relationship between concern for others, advice seeking and gender difference is significant. In this research, the main item measuring people’s willingness to engage in e-WOM will be concern for others.

2.1.3 Gender

There are lots of researchers studying gender differences. And there are signiﬁcant gender differences in consumers’ perception of online consumer reviews (Bae & Lee, 2011). According to this article, females are more likely to listen to other’s opinion and rely on recommendations while males are more certain about what they want to get. From Herring’s study in 1996, we know that when communicating, males are more willing to take control of the conversation and show their knowledge to others by giving recommendations while females put more focus on cooperation and collaboration. What’s more, females feel more comfortable with a face-to-face chat while males prefer an asynchronous forum (Herring, 2003).

There is also previous research about gender difference and behavior online. Females are less likely to invest much energy online and less inspired by web (Garbarino & Strahilevitz, 2004). Females find less social communication when browsing the website and shopping online.

Meyers-Levy (1989) came up with a selectivity model to describe the difference between male and female when they shopping online. Males consider about function when shopping online while women focus more on emotional involvement (Dittmar, Long & Meek, 2004). In the selectivity model, a person’s psychological gender identity is developed by biological sex. That’s to say, males are more independent and like

(17)

17 buying positional goods to show their social status, while females are less independent and like helping others (Meyers-Levy & Maheswaran, 1991).

From those studies we can see that male consumers have more willingness to provide reviews to convey information in e-WOM, while female consumers are more used to participate in traditional WOM.

2.1.4 Culture

Hofstede defined culture as ‘the collective programming of the mind that distinguish the members of one group or category of people from others’ (Hofstede, 2011, p.3). People’s behavior, characteristic and habits differ a lot with different cultural background.

2.1.4.1 Hofstede’s cultural dimension

Geert Hofstede studies about cross-cultural consumer behavior and comes up with a cultural dimensions theory. This theory has been widely used in analyzing national cultural differences in consumer behavior (Chau, Cole, Massey, Montoya-Weiss & O’Keefe, 2002) and WOM (Lam, Lee & Mizerski, 2009). According to previous studies, WOM behavior of consumers is tightly relevant to Hofstede’s national cultural dimension theory (Ma, 2013). There are four dimensions in the original version of this theory and they will be studied in this thesis.

Individualism versus Collectivism (IDV)

The first one is individualism and collectivism dimension. In the individualism culture, people pay great attention to realize their own value and gain profit through their own efforts (Hofstede, 1983). However, in the collectivism dimension, people tend to rely on group. They want to get help from other people in the group but at the same time,

(18)

18 they are also loyal to the group. According to the study of Hofstede, most of the countries in North Europe and North America value individualism culture. However, countries in East Asia such as China and Japan value collectivism culture. (Hofstede, 1983).

Uncertainty Avoidance Index (UAI)

The second one is uncertainty avoidance dimension. This dimension is dealing with the tolerance of uncertainty and ambiguity in a society (Hofstede, 1983). Hofstede states that this dimension is about anxiety and distrust toward unknown things and willingness to know the truth. In a low uncertainty avoidance culture, people seek for adventure and risk. New products are easier adopted by people in this culture. However, in a high uncertainty avoidance culture, people are more likely to traditional beliefs and clear instructions (Ward & Kennedy, 1993).

Masculinity versus Femininity (MAS)

The third one is masculinity and femininity dimension. This dimension shows gender roles in a society. In a masculinity society, the distinction of male and female is evident. Male should be confident and eager for success and money, while female should be humble and deal with daily life affairs (Hofstede, 1983). People seek for success and want to make contribution. In a femininity society, there are not so huge difference between male and female. People are all humble and gentle and male can also be weak. The gap between genders is smaller in this society. People seek for high quality of life and equality (Hofstede, 1998). Countries like Germany, Italy, New Zealand and Japan are masculinity society and countries like Finland, Denmark and Netherland are femininity society (Gudykunst, 2002).

(19)

19

Power Distance Index (PDI)

The forth one is power distance dimension. This dimension shows the different treatment in different culture to inequalities. In a high power distance culture, people with power are always truth and try to keep their power as long as possible (Hofstede, 2003). Opinion leader and authority are valued in this culture. In a low power distance culture, people with power try to make themselves look less powerful and want to close the gap between themselves and people with less power.

In this thesis, I will use individualism and uncertainty avoidance these two dimensions as independent variables. Power distance and masculinity will be used as control variables.

2.1.4.2 Individual level

There are huge amount of researches using Hofstede’s cultural dimension to study and explain consumers’ purchase intention, buying behavior and attitude (Chau, Cole, Massey, Montoya-Weiss & O’Keefe, 2002; Ma, 2013; Lam, Lee & Mizerski, 2009). As Hofstede’ cultural dimension theory is based on national culture, results of those studies are all about the significant impact of national cultural values on consumer behavior, in which all participants with same nationality are treated as the same cultural values.

However, according to Donthu and Yoo (1998), cultural values are not necessarily related to country or nationality. It is difficult to judge people’s cultural values simply based on nationalities. Culture varies among individuals (Yoo, Donthu, 2002). Using Hofstede’s national cultural dimensions at individual level is reasonable because Donthu and Yoo (1998) have confirmed cultural value at individual level shares the same dimension with cultural value at national level.

(20)

20 Hofstede proposes his cultural dimension theory in 1980 after studying the employees of IBM from 53 countries (Hofstede, 1980). However, things change a lot due to digitalization in marketing and globalization around the world. People living in the same country may have different attitudes toward cultural values (Yoo, Donthu & Lenartowicz, 2011). More and more people live in a cross-cultural environment. Being influenced by increased mobility, cross-cultural environment and communication on the Internet, it is possible that a person living in a country valuing collectivist culture may value individualist culture at individual level.

For marketing activities, Keillor, Amico and Horton (2001) state that targeting on consumers rather than country will increase the success of marketing activities. Yoo, Donthu and Lenartowicz also suggest that findings about cultural value at individual level is more relevant and useful for marketing activities and managerial implications. As explained in the previous paragraphs, it’s important for companies and marketers to look Hofstede’s cultural dimension at an individual level nowadays.

2.1.5 Relationship between gender and culture

Culture is defined as ‘a learned system of knowledge, behavior, attitudes, beliefs, values, rules and norms that is shared by a group of people and shaped from one generation to the next’ by Julia T. Wood. Culture impacts many aspects of people’s life. Gender is the state of being male or female at social and cultural level, which is also influenced and shaped by culture. The meaning of gender heavily depends on cultural values. But at the same time, the cultural dimension of masculinity and femininity also shows that the behavior of male and female in a society influence cultural views (Karam, 2016). Gender and culture are both social beliefs and values created by specific society at a specific time. There are some researches about gender and culture interaction effect in emotion, consumer behavior and communication.

(21)

21 Bagozzi, Wong and Yi (1999) find that there is an interaction effect between gender and culture when producing different patterns of association with positive and negative emotions. People in countries valuing independent culture experience oppositional emotions. However, people in countries valuing interdependent culture experience emotions in dialectic ways. And these different patterns are stronger for female than male in these two cultures.

Parboteeah, Hoegl, and Cullen (2008) also find a positive relationship between valuing high power distance culture and traditional gender roles. Their research shows that in a society valuing high power distance culture, females’ social position is more likely to be in the lower ends of societal hierarchy. And people are always willing to and used to this inequality.

Chen, Li and Ceglarska (2008) find in their research that communication in a team with different gender and different cultural background leads to a least satisfied situation comparing to team with the same gender or same cultural background. This means that there is an interaction effect between gender and culture diversity in team communication.

In the study of Kahttab, AI-Manasra, Zaid and Qutaishat (2012), they find that there is a significant interaction effect between gender and collectivist/individualistic on online purchase intention. Female consumers valuing individualistic culture have a lower intention to engage in online activities.

According to the research of Shan (2014), the researcher studies culture and gender difference together at an interactive perspective on negotiation performance. In low context cultures, male negotiators perform better than female, while in high context

(22)

22 cultures, females perform better than male. It shows there is an interaction between gender and culture difference in business communication.

From these previous research, we can see that gender and culture are interlinked. It is not possible for us to study any one of them without having a good understanding of another one. However, there is no investigating about the interaction effect of gender difference and cultural values at individual level on e-WOM. As e-WOM is also an online activity and a kind of communication which happens on the Internet during buying process, I would expect there is an interaction between gender and culture when people engage in e-WOM.

2.2 Hypothesis Development

In this part, I will develop hypotheses of this research based on previous studies and researches. The developing process of hypotheses will be explained in the following part.

As mentioned in the literature review, people’s willingness to engage in e-WOM are motivated by eight factors. And two of these eight factors: concern for others and advice seeking are the only two factors that have a significant relationship between gender and culture difference (Shen, Maceli, Zhao & Baack, 2014; Ma, 2014). Concern for others means the desire to help other consumer make a purchase decision (Sundaram, Mitra & Webster, 1998). People engage in e-WOM to provide helpful and useful experience for friends or strangers online. Advice seeking means people engage in e-WOM to search for information about price, applications and product recommendation (Ma, 2014). In this thesis, concern for others and advice seeking will be the measurements of people’s willingness to engage in e-WOM.

(23)

23 As for the process of interaction hypotheses developing, there is no existing literature about the interaction effect of gender difference and cultural values at individual level on people’ willingness to engage in e-WOM. Thus, the development of interaction hypotheses will base on previous researches related to communication and online activities, which are behaviors similar to e-WOM.

Individualism

Previous study has found that people in the individualism culture value personal goals while people in the collectivistic culture value group goals. Markus and Kitayama (1991) finds that people valuing individualistic culture are more likely to measure a personal success by achieving personal goals while people valuing collectivistic culture enjoy serving and helping others. According to these researches, people valuing collectivistic culture are more willing to provide and share information to help others than people valuing individualistic culture, which is a concern for others regarding to engagement in e-WOM.

This leads to the following hypothesis:

• H1. There is a negative relationship between valuing individualistic culture and willingness to engage in e-WOM.

Gender is expected to be a moderator between valuing individualistic culture and willingness to engage in e-WOM communication according to previous researches which indicate that there is an interaction effect between culture and gender difference on communication (Burleson, 2003; Shan, 2014; Chen, Li & Ceglarska, 2008; Parboteeah, Hoegl & Cullen, 2008). From the literature review above we have already know that male consumers have more willingness to provide reviews to convey information in e-WOM, while female consumers are more used to participate in

(24)

24 traditional WOM and have a face-to-face communication (Garbarino & Strahilevitz, 2004; Meyers-Levy & Maheswaran, 1991). Pavlou and Chai (2002) state that online purchase intention is a behavior to get involved in online transaction which is influenced by e-WOM (Lau & Ng, 2001; Cheung, Luo, Sia & Chen, 2009). In the study of Kahttab, AI-Manasra, Zaid and Qutaishat (2012), they find that there is a significant interaction effect between gender and collectivist/individualistic on online purchase intention. Female consumers valuing individualistic culture have a lower intention to engage in online activities (Kahttab, AI-Manasra, Zaid & Qutaishat, 2012). As there is a positive relationship between willingness to engage in e-WOM and intention to purchase online (Yusuf, Che & Busalim, 2018) and e-WOM communication is also a kind of online activity as online purchase, when it comes to this thesis, I expect female consumers valuing individualistic culture have less willingness to engage in e-WOM than male consumers according to the previous research.

• H2. The negative relationship between valuing individualistic culture and willingness to engage in e-WOM is moderated by different gender, so that this relationship is stronger for female.

Uncertainty avoidance

From the previous literature review we already know that people valuing high uncertainty avoidance culture don’t like changes and are favor of obeying rules and principles. People valuing high uncertainty avoidance culture live under a high level of anxiety. They need to reduce ambiguity and uncertainty to lower their anxiety (Doney, Cannon & Mullen, 1998). Seeking for others’ advice and instruction is a way to solve the problem according to Doney, Cannon and Mullen (1998). And advice seeking is

(25)

25 one of the motivations of people’s willingness to engage in e-WOM. That’s to say, people valuing high uncertainty avoidance is associated with a high level of seeking for information and advice to reduce risk. They have more willingness to engage in e-WOM than people valuing low uncertainty avoidance.

• H3. There is a positive relationship between valuing high uncertainty avoidance culture and willingness to engage in e-WOM.

Gender is expected to be a moderator between valuing high uncertainty avoidance culture and willingness to engage in e-WOM communication according to previous researches which indicate that there is an interaction effect between culture and gender difference on communication (Burleson, 2003; Shan, 2014; Chen, Li & Ceglarska, 2008; Parboteeah, Hoegl & Cullen, 2008). Male consumers are more willing to engage in e-WOM according to previous literature review (Garbarino & Strahilevitz, 2004; Meyers-Levy & Maheswaran, 1991). From the selectivity model of gender, males are independent and prefer online activities such as e-WOM communication while females are interdependent and prefer face-to-face communication such as WOM communication (Garbarino & Strahilevitz, 2004). The selectivity model also shows a traditional gender role attitude. And people valuing high uncertainty avoidance prefer obeying rules and holding traditional attitude (Parboteeah, Hoegl & Cullen, 2008). That’s to say, male consumers valuing high uncertainty avoidance have more willingness to engage in e-WOM than female consumers.

(26)

26 • H4. The positive relationship between valuing uncertainty avoidance culture and willingness to engage in e-WOM is moderated by different gender, so that this relationship is stronger for male.

2.3 Conceptual Model

(27)

27

Chapter 3: Methodology

This chapter describes the research design of this thesis. First, the method used and information of sample in this thesis are shown. Next, variables and measurements to measure these variables are given.

3.1 Method

This study will be quantitative by using an online survey design. It is easy to reach a large amount of people within a short time when we use an online survey. In order to reach people from different country cultural backgrounds and increase the accuracy of results, this research will include a survey with two language versions: English and Chinese. In order to ensure the consistency of the survey’s content, the survey is translated and back-translated.

This research is about the influence of cultural value at individual level on people’s willingness to engage in e-WOM and the moderating effect of gender difference. The survey is divided into 3 parts. Participants will be asked about their willingness to engage in e-WOM in the beginning. The second part of the survey is about participants’ personal cultural value. The last part includes questions about demographical characteristics. The English version of the survey is presented in Appendix 2.

Before I launch the survey, I will pilot test it on my friends and family members to make some adjustments.

3.2 Sample

The sampling frame is unknown and the size of the sample is large, so it is a non-probability convenience sample. The survey will be sent through e-mail and Facebook. The minimum respondents of this research is expected to be 250 in order to be enough for analysis. The response rate of earlier research differs a lot because of different

(28)

28 choice of platform. As this survey will also be reached via Facebook, it’s hard to predict the response rate. The minimum expected response rate is 50%.

3.3 Measures

All the scales and items used in the previous research are in English. As European consumer has a great understanding of English, European respondents can use English survey directly. But in order to ensure a better and accurate understanding for Chinese consumer, the survey will be translated into Chinese by back-translation.

The questionnaire will ask the respondents questions about demographics such as age (rational variable) and education background (ordinal variable).

This research will use pre-validated scales from previous scientific articles which have already been tested on other samples and have less problems.

3.3.1 Dependent variable

The dependent variable of this research is people’s willingness to engage in e-WOM. The willingness to engage in e-WOM is measured by a 7-point Likert scale rating from 7 (strongly agree) to 1 (strongly disagree) with 3 items (Hennig‐Thurau, 2004). The Cronbach’s Alpha of concern for other consumers is 0.802, that of advice seeking is 0.887. Hennig-Thurau measured 8 factors with 27 items in his original paper. This research will only use 2 factors with 3 items of that scale and make some changes to the questions. For example, one question for concern for others is “I share information with others online to avoid others making a risky decision .”

3.3.2 Independent variables

The independent variable of this research is cultural value at individual level. This research will study two primary dimensions of Hofstede’s cultural value dimensions at the individual level: individualism/collectivism and uncertainty avoidance. Dorfman

(29)

29 and Howell developed an individual-level scale to measure Hofstede’s cultural theory for the first time in 1988 and later studies use this scale as a standard. It is a 7-point Likert scale rating from 7 (strongly agree) to 1 (strongly disagree) with 22 items. According to recent studies, the Cronbach’s Alpha of individualism/collectivism dimension is 0.72 and that of uncertainty avoidance is 0.86 (Robertson & Hoffman, 2000). As the original research pays more attention to organization management, some question in the scale will be adapted to fit this research. For example, the question “Managers should encourage group loyalty even if individual goals suffer” will be adapted to “Group loyalty still need to be encouraged even if individual goals suffer”. The entire version of adapted scales is listed in Appendix 1.

3.3.3 Control variables

The other two primary dimensions of Hofstede’s theory: masculinity/femininity and power distance will be measured as control variables in this research. The measurement will also use Dorfman and Howell’s 7-point Likert scale from 7 (strongly agree) to 1 (strongly disagree). The Cronbach’s Alpha of masculinity/femininity is 0.87 and that of power distance is 0.85 (Robertson & Hoffman, 2000). Some questions will also be adapted. For example, the question “Managers should seldom ask for the opinions of employees” will change to “People in a high power position should seldom ask for the opinions of people in a low power position”. The entire version of adapted scales is listed in Appendix 1.

3.3.4 Moderating variable

The moderating variable of this research is gender difference. This can simply be asked at the demographics question part. All respondents are being asked the question ‘what is your gender’. In order to analyze the data, a dummy variable is created: ‘male’ (0) and ‘female’ (1).

(30)

30 3.4 Summary of measurements

The construct, scale items, source of scales and scale type are listed in the following table.

Table 1: Summary of Measurements

Construct/variable Source Scale Scale type

Willingness to engage in e-WOM

Hennig‐Thurau, T., Gwinner, K. P., Walsh, G., & Gremler, D. D. (2004)

1. I share information with others online to avoid others making a risky decision.

2. Sharing information online is a fun way to communicate in the community.

3. I feel good when I tell others online about my buying success.

7 point-likert scale (ordinal)

Independent variables:

Individualistic Dorfman, P. W., & Howell, J. P. (1988) 1. Group welfare is more important than individual rewards

2. Group success is more important than individual success 3. Being accepted by the members of group is very

important

4. Individuals should only pursue their goals after considering the welfare of the group

(31)

31 5. Group loyalty should be encouraged even if individual

goals suffer.

6. Individuals may be expected to give up their goals in order to benefit group success.

Uncertainty avoidance Dorfman, P. W., & Howell, J. P. (1988) 1. It is important to have instructions spelled out in detail so that I always know what I’m expected to do.

2. It is important to closely follow instructions and procedures.

3. Rules and regulations are important because they inform me what is expected of me.

4. Standard operating procedures are helpful for me. 5. Instructions for operations are important for me.

Moderator:

Gender Nominal

Control variables:

Masculinity Dorfman, P. W., & Howell, J. P. (1988) 1. Projects are usually run more effectively when they are chaired by a man

2. It is more important for men to have a professional career than it is for women.

3. Men usually solve problems with logical analysis; women usually solve problems with intuition.

(32)

32 4. Solving difficult problems usually requires an active

forcible approach which is typical of men

5. It is preferable to have a man in a high level position rather than a woman

Power distance Dorfman, P. W., & Howell, J. P. (1988) 1.People in higher positions should make most decisions without consulting people in lower positions.

2. It is frequently necessary for a manager to use authority and power when dealing with subordinates

3.People in higher positions should seldom ask the opinions of people in lower positions.

4.People in higher positions should avoid off-the-job social contacts with people in lower positions. 5.People in lower positions should not disagree with decisions by people in higher positions.

6.People in higher positions should not delegate important tasks to people in lower positions.

Demographic variables:

Age Rational

(33)

33

Chapter 4 Data Analysis

This chapter is about analysis of collected data and results of this thesis. First, some descriptive statistics of the data are given. Next, how data prepared for analysis is shown. Then, reliability of scales, means of variables and correlation are given. After this, it will be the test of hypotheses.

4.1 Descriptive statistics

There are 427 respondents participated in this survey and 328 of them completely finished the survey. The complete rate is 76.8%. The size of the sample is big enough for this research. Data is collected by Qualtrics Survey Software online. SPSS, a computer software package to process quantitative and qualitative data of research, is being used to analyzing the data collected. 78% respondents fill in the survey of Chinese version. The majority of respondents is below the age of 30 and 71.1% respondents are female.

4.2 Preparation of Data

4.2.1 Missing value: frequency test

First of all, in order to examine if there are any errors in the data, all variables are checked for missing data. A frequency test is run for all variables. For all variables, there are no errors and the percentage of missing data is < 10%. In order to deal with missing data, we exclude those cases and only analyze cases without missing data in any variables.

4.2.2 Recoding

First of all, gender is a moderator in this research. The scale of it is adapted from 1 (Male), 2 (Female) to 0 (Male), 1 (Female). Then is the recoding of counter-indicative items. Q3_IDV1, Q3_IDV2, Q3_IDV3, Q3_IDV4, Q3_IDV5 and Q3_IDV6 are

(34)

34 recoded to IDV1, IDV2, IDV3, IDV4, IDV5 and IDV6. The recoding of individualism is because that it is a negatively-keyed item in the data, which is a counter-indicative item. In the scale I use of individualism, 1 to 7 means individualism to collectivism. I recode it into a new variable and make 1 to 7 means collectivism to individualism which can avoid bias in the result.

4.3 Reliability

In order to check the reliability of the measurements, individualism culture value, uncertainty avoidance culture value and willingness to engage in e-WOM run for reliability checks. The Cronbach’s Alpha of these three variables are all above 0.7 (eWOM α=0.754；IDV α=0.850; UAI α=0.863) This means the internal consistency of the scales is at a high level. The corrected item-total correlations indicate that all the items have a good correlation with the total score of the scale (all above 0.30). Items will not substantially affect reliability if they were deleted. The statistics are showed in Table 1.

Table 2: Cronbach’s Alpha

Variables Cronbach’s Alpha

Willingness to engage in e-WOM 0.754

Individualism culture value 0.850

Uncertainty avoidance culture value 0.863

Descriptive statistics, skewness, kurtosis and normality tests are performed for all variables. All variables of willingness to engage in e-WOM, individualism culture value , uncertainty avoidance culture value and gender are not normally distributed. But

(35)

35 according to Tabachnick and Fidell, skewness and kurtosis will not have a substantive influence to the analysis if we have a large sample. There are 328, which is more than 200 people participate in this research. This means the influence is reduced due to the large sample.

4.4 Correlations & Computing scale means

In order to start the test of hypotheses, the mean of these items need to be calculated to describe a variable. Means and standard deviations of these variables are shown in Table 3.

From the table we can see individualism culture value is a strong predictor of willingness to engage in e-WOM with a r=-.362, p<0.01, which confirms H1. Uncertainty avoidance culture value is a strong predictor of willingness to engage in e-WOM with a r=.410, p<0.01, which confirms H2. Gender is also a strong predictor with a r=.203, p<0.01.

(36)

36

Table 3: Means, Standard deviations and Correlations

Variables Mean SD 1 2 3 4 5 6 7 8 1. Gender .71 .454 - 2. Age 32.12 12.586 -.231** - 3. Education 2.10 .620 -.065 -.027 - 4. e-WOM 5.0149 1.17128 .203** -.011 -.036 (.754) 5. Masculinity 3.5337 1.42294 -.339** _.429** _-.065 _.077 _(.858) 6. Power distance 2.8759 1.09972 -.181** _.176** _-.049 _-.014 _.406** _(.813) 7. Individualism 3.0375 1.11100 .040 -.332** .143* -.362** _-.389** _-.067 _(.850) 8. Uncertainty avoidance 5.4356 1.01206 .069 .227** _-.112* _.410** _.270** _.079 _-.524** _(.863)

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).

(37)

37 4.5 Regression analysis

In order to test the relation of independent variables (Individualism cultural value; Uncertainty avoidance cultural value) and dependent variable (Willingness to engage in e-WOM), a hierarchical regression analysis is performed. Age, education, masculinity cultural value and power distance cultural value are control variables.

In the first step of hierarchical multiple regression, four predictors are entered: age, education, masculinity cultural value and power distance cultural value.

Table 4: Regression Analysis of Willingness to Engage in e-WOM R R2 _R2 Change B SE β t Step 1 .148 .022 Age -.004 .006 -.047 -.701 Education -.128 .113 -.069 -1.137 Masculinity .116 .058 .143* 2.014 Power Distance -.103 .071 -.096 -1.457 Step 2 .466 .217*** .195*** Age -.013 .006 -.140* -2.300 Education -.008 .102 -.004 -.077 Masculinity .007 .054 .009 .131 Power Distance -.055 .064 -.052 -.866 Individualism -.249 .069 -.242*** _-3.601 Uncertainty avoidance .344 .070 .312*** 4.946 Statistical significance: *p <.05; **p <.01; ***p <.001

In step 1, the model is not significant with p=.204, >.05. In step 2, individualism cultural value and uncertainty avoidance cultural value are entered as predictors. The total variance explains by the model as a whole is 21.7% F (6,266) = 12.273; p<.001. In the final model, three out of six predictor variables are statistically significant. Individualism cultural value records a Beta value of -.242, p<.001. That’s to say, if a

(38)

38 person values more individualistic culture than collectivistic culture, the willingness to engage in e-WOM decreases for .242. Uncertainty avoidance cultural value records a Beta value of .312, p<.001. This means the willingness to engage in e-WOM will increase if a person values high uncertainty avoidance culture. Age records a Beta value of -.124, p<.05. The willingness to engage in e-WOM decreases for .124 if age increases.

4.6 Moderating Effect

In order to test the interaction effect, a moderation analysis is performed. This research uses the Process macro written by Andrew F. Hayes for SPSS to test this moderating effect.

Gender is set as a moderator between the relationship of individualistic cultural value, uncertainty avoidance cultural value and willingness to engage in e-WOM. Model 1 of Hayes, which is a conceptual and statistical process model for simple moderation, is used to test hypothesis 2 and 4. Figure 1 shows this model. As there are two independent variables in this research, Process model 1 will be performed for each independent variable.

(39)

39 4.6.1 Test of hypothesis 2

From table 5 we can see that the model of hypothesis 2 is significant (p<.001). But when we look at the significance of interaction effect, the interaction effect is not significant. P value of this interaction is 0.6048 which is higher than 0.005. Therefore, there is no evidence of moderation. Hypothesis 2 is totally rejected. Data and information of the moderation analysis are listed in table 5.

Table 5: Moderator Analysis Model Summary

R R-sq F df1 df2 p

0.4265 .1819 22.4578 3.0000 303.0000 .0000

R-square increase due to interaction(s)

R2-chng F df1 df2 p int_1 .0009 .2683 1.0000 303.0000 .6048

Interaction variables

Variable Coeff SE t p LLCI ULCI

Constant 5.9719 .3396 17.5845 .0000 5.3036 6.6403

Gender .3658 .3854 .9491 .3433 -.3926 -1.1241

Individualism -.4253 .1067 -3.9844 .0001 -.6353 -.2152

Int_1 .0639 .1233 .5180 .6048 -.1787 .3065

4.6.2 Test of hypothesis 4

From table 6 we can see that the model of hypothesis 4 is significant (p<.001). But when we look at the significance of interaction effect, the interaction effect is not significant. P value of this interaction is 0.6532 which is higher than 0.005. Therefore, there is no evidence of moderation. Hypothesis 4 is totally rejected. Data and information of the moderation analysis are listed in table 6.

(40)

40

Table 6: Moderator Analysis Model summary

R R-sq F df1 df2 p .4464 .1993 22.2176 3.0000 302.0000 .0000

R-square increase due to interaction(s)

R2-chng F df1 df2 p int_1 .0007 .2023 1.0000 302.0000 .6532

4.7 Summary of hypotheses

Table 7 is the summary of all the hypotheses of this research and the results of analysis are also included.

Table 7: Hypotheses Summary

Hypothesis Result

H1 There is a negative

relationship between valuing individualistic culture and willingness to

Supported

Interaction variables

Variable Coeff SE t p LLCI ULCI

Constant -.2129 .1148 -1.8540 .0647 -.4388 .0131 Gender .3814 .1279 2.9829 .0031 .1298 .6331 Uncertainty avoidance .4242 .1001 4.2379 .0000 .2272 .6212 Int_1 -.0562 .1250 -.4498 .6532 -.3022 .1897

(41)

41 engage in e-WOM.

H2 The negative relationship between valuing individualistic culture and willingness to engage in e-WOM is moderated by different gender, so that this relationship is stronger for female.

Not supported

H3 There is a positive

relationship between valuing high uncertainty avoidance culture and willingness to engage in e-WOM.

Supported

H4 The positive relationship between valuing uncertainty avoidance culture and willingness to engage in e-WOM is moderated by different gender, so that this relationship is stronger for male.

(42)

42

Chapter 5 Conclusion

This chapter will give a conclusion and discuss the findings of this thesis. Then, implications for theory and management will be given. After that, limitations of this thesis and suggestions for future researches will be provided.

5.1 Conclusions and Discussions

The aim of this research is to test whether people’s personal cultural value has an impact on their willingness to engage in e-WOM, and whether gender differences paly a moderating role in this relationship. There is a significant effect of cultural values on people’s e-WOM behavior (Ma, 2013; Money, Gilly & Graham, 1998). However, these researches are studying cultural values at national level. Yoo et al. (2011) has demonstrates that there is a need to measure cultural value at individual level.

From the hierarchical regression analysis, hypothesis 1 and 3 are proved. There is a negative relationship between individualism culture value and willingness to engage in e-WOM. There is a positive relationship between uncertainty avoidance culture and willingness to engage in e-WOM. For theoretical implications, these findings support the previous researches about relationship between individual culture value and willingness to engage in e-WOM. These results provide more literature for future researchers to look deeper at cultural values at individual level.

However, as for the moderating effect, hypothesis 2 and 4 are rejected. Gender doesn’t play a significant moderating role between individualism culture value, uncertainty avoidance culture value and willingness to engage e-WOM. The reasons for these rejected hypotheses could be a lot and they will be discussed now.

For the interaction hypothesis 2: the moderating effect of gender on the relationship between valuing individualistic culture and willingness to engage in e-WOM, it is not

(43)

43 supported in this research. The study results of gender differences in individualism and collectivism across researches have varied (Lalwani & Shavitt, 2011). Gabriel and Gardner (1999) find no gender differences on different tasks and behavior. Baumeister and Sommer (1997) report that male and female do not differ on individualism. These studies show there is no relationship between gender differences and individualism. However, Kahttab, AI-Manasra, Zaid and Qutaishat (2012) find there is a relationship between gender and individualism/collectivism on online activities. Different researches have different result on this relationship.

For the interaction hypothesis 4: the moderating effect of gender on the relationship between valuing high uncertainty avoidance culture and willingness to engage in e-WOM, it is not supported in this research. The reason for failure may be the sample in this research. The sample used in this research is a non-probability convenience sample. It means the generalizability of the data and results will be reduced and weakened. This study doesn’t collect data of nationality. But there is 2 language version of the survey and 70% respondents fill in the Chinese version. This means the majority of respondents could be Chinese. This might lead to lack of variance as people’s individual cultural value could also be influenced by their living background and society. China ranks really high in uncertainty avoidance dimension of Hofstede’s national cultural dimension. 70% respondents living and being influenced in an environment valuing high uncertainty avoidance culture which could lead to a non-normal distribution in the dataset and influence the result.

Then, previous studies about the moderating effect of gender between the relationship of cultural value and willingness to engage in e-WOM all look at national culture level. However, this study measures cultural value at individual level. This could be a reason why the moderating effect is not supported in this research.

(44)

44 After all, it is worthwhile to know people’s willingness to engage in e-WOM is related to valuing individualistic and uncertainty avoidance culture at individual level.

5.2 Managerial implications

From a managerial perspective, the study provides some important implications.

Companies nowadays pay great attention to electronic word of mouth management. Marketers can develop more targeted strategies to attract consumers valuing different culture to engage in electronic word of mouth.

The results of this study show that people valuing collectivistic culture and people valuing high uncertainty avoidance are more fond of engage in e-WOM. Knowing how e-WOM works in group of people with similar cultural values is important for marketers and companies. It can help companies create a better marketing environment with additional consumer referrals, faster information spreading and increased market share.

This research is about people’s cultural values at individual level rather than national level. And there is a significant relationship between individual level culture and willingness to engage in e-WOM. From these results, marketers should know they can’t estimate consumers’ willingness to engage in e-WOM simply based on national culture dimension. They should consider more about individual perception.

Then, the result of this research shows that people valuing collectivistic culture and high uncertainty avoidance culture are positively related to people’s willingness to engage in e-WOM. Marketers and e-WOM platform developers should put more efforts on building an online community targeting people valuing collectivistic culture and high uncertainty avoidance culture and giving them an interaction and friendly online platform. As this research studies about people’s individual level cultural values within

(45)

45 countries. The results of this study are also important for marketers and companies launching business in a culturally mixed country.

However, the measurement of individual cultural value used in this thesis simply copies from the measurement of Hofstede’s national cultural value. It’s important for marketers to find a new way to capture consumers’ individual cultural value to keep with a rapidly changing marketing environment.

5.3 Limitations & future research

Despite the theoretical and practical implications discussed above, there are still some limitations of the current research.

Firstly, the sample used in this research is a non-probability convenience sample. It means the generalizability of the data and results will be reduced and weakened. 78% respondents fill in the survey of Chinese version. The majority of sample is below the age of 30 and 71.1% respondents are female. This means the representativeness is limited.

As the moderating effect of gender is not proved in this research, it would be interesting for future research to find out if there is really not a moderating effect of gender differences between the relationship of personal culture value and willingness to engage in e-WOM, or that this result is just caused by the research setting of this study. Future research can also try to find if there is other factors may moderate the relationship. This study only tests two dimensions of Hofstede’s cultural theory: individualism and uncertainty avoidance. Masculinity and power distance are used as control variables. Hofstede’s cultural dimension has developed to 6 dimensions: power distance, individualism/collectivism, masculinity/femininity, uncertainty avoidance, long/short

(46)

46 term orientation and indulgence/restraint. Future research can test all these dimensions together to see if there will be any new results.

Furthermore, the measurements used in this research are also a limitation. When measuring people’s willingness to engage in e-WOM, this research only looks at the people’s intention of engaging in e-WOM rather than people’s actual behavior. And in real life, the intentions and willingness do not always fully translate into behavior. When measuring cultural values, this research measures it at individual level by using the scale of Hofstede’s national culture dimension. Blodgett, Bakir and Rose (2008) test the validity of Hofstede’s national level cultural value scale when applied to individual level. However, it lacks sufficient validity when applied to individual level. More researches should be done to find if there is a new scale which is suitable to measure cultural value at individual level.

(47)

47

References

Arndt, J. (1967). Role of product-related conversations in the diffusion of a new product. Journal of marketing Research, 291-295.

Bae, S., & Lee, T. (2011). Gender differences in consumers’ perception of online consumer reviews. Electronic Commerce Research, 11(2), 201-214.

Bagozzi, R. P., Wong, N., & Yi, Y. (1999). The role of culture and gender in the relationship between positive and negative affect. Cognition & Emotion, 13(6), 641-672.

Beldad, A., De Jong, M., & Steehouder, M. (2011). I trust not therefore it must be risky: Determinants of the perceived risks of disclosing personal data for e-government transactions. Computers in Human Behavior, 27(6), 2233-2242.

Bickart, B., & Schindler, R. M. (2001). Internet forums as influential sources of consumer information. Journal of interactive marketing, 15(3), 31-40.

Bimber, B. (2000). Measuring the gender gap on the Internet. Social science quarterly, 868-876.

Blodgett, J. G., Bakir, A., & Rose, G. M. (2008). A test of the validity of Hofstede's cultural framework. Journal of consumer marketing, 25(6), 339-349.

Bounie, D., Bourreau, M., Gensollen, M., & Waelbroeck, P. (2005). The effect of online customer reviews on purchasing decisions: The case of video games. Retrieved July, 8, 2009.

Burleson, B. R. (2003). The experience and effects of emotional support: What the study of cultural and gender differences can tell us about close relationships, emotion, and interpersonal communication. Personal Relationships, 10(1), 1-23.

(48)

48 Cafferata, P., & Tybout, A. M. (1989). Gender Differences in Information Processing: A Selectivity Interpretation, Cognitive and Affective Responses to Advertising.

Cheung, C. M., & Thadani, D. R. (2012). The impact of electronic word-of-mouth communication: A literature analysis and integrative model. Decision support systems, 54(1), 461-470.

ChannelAdvisor., (2010). Through the Eyes of the Consumer: 2010 Consumer Shopping Habits Survey, 14

Davis, A., & Khazanchi, D. (2008). An empirical study of online word of mouth as a predictor for multi‐product category e‐commerce sales. Electronic markets, 18(2), 130-141.

Dittmar, H., Long, K., & Meek, R. (2004). Buying on the Internet: Gender differences in on-line and conventional buying motivations. Sex roles, 50(5-6), 423-444.

Duan, W., Gu, B., & Whinston, A. B. (2008). The dynamics of online word-of-mouth and product sales—An empirical investigation of the movie industry. Journal of retailing, 84(2), 233-242.

Doney, P. M., Cannon, J. P., & Mullen, M. R. (1998). Understanding the influence of national culture on the development of trust. Academy of management review, 23(3), 601-620.

Donthu, N., & Yoo, B. (1998). Retail productivity assessment using data envelopment analysis. Journal of retailing, 74(1), 89-105.

Dorfman, P. W., & Howell, J. P. (1988). Dimensions of national culture and effective leadership patterns: Hofstede revisited. Advances in international comparative management, 3(1), 127-150.