Does Culture Matter? A Comparative Study on the Motivations for Online Identity Reconstruction Between China and Malaysia

Social network platforms provide a good chance for idealizing and managing online identity. Previous research indicated that people often enhance their public self-image intentionally on social network platforms with positive self-presentation strategies (Lee-Won et al., 2014). As mentioned above, people may reconstruct an online identity on the basis of their own ideas, and they may be motivated by various factors, such as vanity, bridging social capital, disinhibition, and privacy concerns (Hu et al., 2015).

Vanity has been defined as “excessive concern for, and/or a positive (and perhaps inflated) view of one’s physical appearance or personal achievements” (Netemeyer et al., 1995, p. 612). It can be distinguished into two types: physical vanity and achievement vanity. Previous research suggested that people with different cultural backgrounds differ in physical vanity and achievement vanity (Durvasula & Lysonski, 2008). Individuals from China and India had more prominent concerns over physical appearance than those from America and New Zealand, while the self-assessment of professional achievement was higher in Western cultures than in Eastern cultures. As economics develops, vanity is becoming increasingly prominent in Asian countries. Previous research suggested that physical and achievement vanity are important predictors of Chinese consumers’ intention to purchase luxury brands (Hung et al., 2011). Chinese consumers tend to buy luxury goods to show their reputation and success (N. Zhou & Belk, 2004). Similarly, achievement vanity is positively associated with Malaysian people’s intention to buy luxury goods (Mamat et al., 2016). Malaysian women who are concerned about vanity prefer products that demonstrate social status (Chui & Samsinar, 2011).

Bridging social capital describes the benefits that come from casual relationships. It derives from relationships that connect individuals from different social networks (Putnam, 2000). Bridging social capital has significant effects on the use of social network platforms in China and Malaysia. For example, in China, the social need (such as making new friends) is one of the important factors that drive individuals to engage in social network platforms, and the gratification of social needs predicts the perceived bridging social capital online (Mo & Leung, 2015). In addition, Chinese users who perceived a higher level of bridging social capital were more likely to feel satisfied (Chang & Zhu, 2012). Studies conducted in Malaysia also indicated that bridging social capital is one of the most important reasons why college students use social network services (Almadhoun et al., 2012; Helou, 2014).

Disinhibition is an effect that makes people loosen up and behave freely online (Suler, 2004). Due to some interrelated factors (such as anonymity, invisibility, and minimized authority), online disinhibition can result in both positive and negative effects. The positive effects of disinhibition are referred to as benign disinhibition (e.g., a greater extent of self-disclosure); the negative effects of disinhibition are referred to as toxic disinhibition (e.g., flaming and harsh criticisms; Suler, 2004). Previous research found that reduced nonverbal cues (such as visual and context cues) diminished people’s inhibition in computer-mediated communications (Lapidot-Lefler & Barak, 2015). The feelings of disinhibition enhanced people’s online self-disclosure (Schouten et al., 2007). People disclose more emotion online when they are anonymous and invisible to others (Lapidot-Lefler & Barak, 2015). However, invisibility, anonymity, and lack of eye contact can lead to toxic disinhibition, such as flaming, negative atmosphere, and even threats (Lapidot-Lefler & Barak, 2012). It was also found that online disinhibition was positively associated with cyberbullying behavior (Udris, 2014).

Privacy concerns refer to a sense of anxiety related to the infringement of privacy (Tan et al., 2012). It reflects a person’s concerns about private information. It is suggested that culture plays an essential role in people’s privacy concerns (Lowry et al., 2011). Using a Chinese sample, scholars have found that users’ privacy concerns were positively associated with their perceptions of risks and uncertainty when using social network platforms (T. Zhou & Li, 2014). A study conducted in Malaysia also found that if individuals thought that the consequences of privacy loss were serious, they would be more concerned with privacy (Mohamed & Ahmad, 2012).

By the end of 2018, the population of China stood at approximately 1.4 billion. The Chinese population consists of multiple ethnicities (56 in total). As the main group, the Han ethnic group accounted for more than 91% of the total population (The National Bureau of Statistics [NBS], 2019). The other 55 ethnic groups made up the remaining population (NBS, 2019). The official language of China is Mandarin, which is spoken by most Chinese people (Odinye, 2015).

For Malaysia, its population reached 32.58 million by the second quarter of 2019 (DOSM, 2019). The major ethnic groups are the Malaysian-Malays (also known as Bumiputera, the indigenous people of Malaysia), the Malaysian-Chinese, and the Malaysian-Indians. The Malaysian-Malays accounted for 69.3% of the total population, whereas the Malaysian-Chinese and the Malaysian-Indians accounted for 22.8% and 6.9%, respectively (DOSM, 2019). People from different ethnic groups prefer their mother tongues or English. For example, the Malaysian-Malays speak Bahasa Melayu (the Malaysian language), whereas the Malaysian-Chinese prefer Mandarin, Cantonese, and Hakka, and the Indians speak Tamil (Tamam, 2011; Wan Jafar, 2014).

Various theories have been proposed by previous researchers to operationalize culture. Hofstede was the first researcher to quantify national culture with six cultural dimensions, namely, individualism versus collectivism, power distance, masculinity versus femininity, uncertainty avoidance, long-term versus short-term orientation, and indulgence versus restraint (Hofstede, 1980; Hofstede et al., 2010). Inspired by the work of Hofstede, the GLOBE project tries to capture the values of a society with nine cultural dimensions (House et al., 2004). An alternative cross-cultural theory, Schwartz’s cultural value orientation, identified seven values at the country level (Schwartz, 2013).

In this study, we adopted Hofstede’s cultural dimension theory instead of others. The more recent cultural frameworks have been found to be highly correlated with Hofstede’s measure (Laskovaia et al., 2017; Magnusson et al., 2008). Although Hofstede’s framework has been criticized (Taras et al., 2012), it is validated by many studies in different fields (Kirkman et al., 2006; Magnusson et al., 2014; F. Zhao, 2011). The six dimensions are introduced in the following content.

Power distance is defined as “the extent to which the less powerful members of organizations and institutions (like the family) accept and expect that power is distributed unequally” (Hofstede, 2011, p. 9). It reflects the degree of inequality in society. In countries with a high level of power distance, followers endorse the unequal distribution of power as much as the leaders do. Followers expect to be guided about what to do. In countries with a low level of power distance, leaders and followers consider each other as equals. Leaders are supposed to consult the followers when making decisions (Hofstede, 2001).

Individualism versus collectivism stands for “the degree to which people in a society are integrated into groups” (Hofstede, 2011, p. 11). In countries with a high level of individualism, people prefer to act as individuals. An individual is supposed to take care of himself or herself and his or her own family. In countries with a low level of individualism (namely, a high level of collectivism), people are connected with close in-groups (such as relatives and friends). In-group members are expected to look after each other and be loyal to the group (Hofstede, 2001).

Masculinity versus femininity is defined as “the degree to which masculine values prevail over feminine values” (Hofstede, 2011, p. 12). In countries with high masculinity, the dominant values are masculine (such as assertiveness, competitiveness, success, and ambition). People place emphasis on achievements and performance. In societies with low masculinity (namely, high femininity), the dominant values are feminine (such as caring for others, warm personal relationships, quality of life, and modesty). People prefer a friendly atmosphere and cooperation (Hofstede, 2001).

Uncertainty avoidance has been defined as “to what extent a culture programs its members to feel either uncomfortable or comfortable in unstructured situations” (Hofstede, 2011, p. 10). In countries with a high level of uncertainty avoidance, individuals feel threatened by ambiguity and try to avoid it; conversely, in countries with a low level of uncertainty avoidance, individuals are comfortable with ambiguous situations and accept uncertainty (Hofstede, 2011).

Long-term versus short-term orientation refers to the extent to which people focus on the future (Hofstede, 2011). In long-term oriented countries, future-focused values are encouraged (e.g., thrift and persistence). People foster virtues to plan for the future. In short-term oriented countries, past- and present-focused values are highlighted (e.g., respect for tradition and personal steadiness and stability). People have a suspicious attitude toward societal changes, focusing more on achieving quick results (Hofstede et al., 2010).

Indulgence versus restraint is defined as the degree of “gratification of basic and natural human desires related to enjoying life and having fun” (Hofstede, 2011, p. 15). In societies with high indulgence, the control of basic desires gratification is relatively weak. People place more emphasis on leisure time and have a tendency toward optimism. In societies with low indulgence (namely, high restraint), the control for the gratification of needs is relatively strong. People feel constrained by strict social norms and perceive that leisure is not important (Hofstede et al., 2010).

Existing cross-cultural studies concerning social network platforms have predominantly investigated cultural differences from the perspective of individualism versus collectivism (e.g., Kim & Papacharissi, 2003; Rui & Stefanone, 2013; C. Zhao & Jiang, 2011). However, individualism versus collectivism is only one of the six dimensions in the cultural dimension framework (Hofstede et al., 2010). As such, there is a need to take into account the remaining cultural dimensions when investigating people’s behavior on social network platforms in the cross-cultural context.

Although the cultural dimension framework was originally developed based on work-related data and initially applied to human resource management, it is now widely referenced in psychology, sociology, business, marketing, and management studies, particularly when examining the cultural differences in people’s emotions, attitudes, behavior, and job performance (Taras et al., 2010). This framework has been used at different levels (i.e., country level, group/organization level, and individual level). For example, at the country level, a prior study found that power distance and individualism were closely related to the level of e-government development (F. Zhao, 2011). In addition, it was found that national culture was significantly associated with the rate of information technology (IT) adoption (Erumban & de Jong, 2006). The IT adoption rate was higher in countries with a lower level of uncertainty avoidance and/or power distance (Erumban & de Jong, 2006).

Some other studies have also used the cultural dimension framework at the group/organization level (Choi et al., 2016; Magnusson et al., 2014). For instance, it has been suggested that members of Korean groups send messages more frequently to each other than members of American groups, and the content of the messages are more friendly (Choi et al., 2016). Magnusson et al. (2014) postulated that individualistic and masculine cultural values have significant impact on the collaboration between the sales division and other departments of organizations.

In addition to the country level and the group/organization level, a majority of researchers have adapted the cultural dimension framework for individual-level studies. For example, Baptista and Oliveira (2015) used the cultural dimension framework in a single country to explore the role of cultural values in people’s adoption of mobile banking services. They found that the relationship between individuals’ behavioral intention and the actual use of mobile banking service was moderated by cultural values (Baptista & Oliveira, 2015). In the cross-cultural context, Chiu and Kosinski (1999) found that nurses from individualistic cultures (i.e., Australia and America) showed more prominent job satisfaction and less work strain than those from collectivistic cultures (i.e., Singapore and Hong Kong).

To provide more precise explanations for people’s behavior at the individual level, Srite and Karahanna (2006) proposed the “espoused national cultural values” on the basis of four cultural dimensions. In their study, individualism versus collectivism is defined as the extent to which people “prefer to act as an individual rather than as a member of a group” (Srite & Karahanna, 2006, p. 682). Power distance is defined as the degree to which the people consider the large differentials of power as normal and accept that people in higher positions have more power. Uncertainty avoidance stands for the degree to which the individual tolerates risk. Masculinity versus femininity is defined as the degree to which the individual espouses gender inequalities (Srite & Karahanna, 2006).

According to Srite and Karahanna’s (2006) study, people who are masculinity oriented place more emphasis on masculine values (e.g., achievements). Their behavior is motivated to fulfill their work goals (Hofstede et al., 2010; Srite & Karahanna, 2006), such as performance, assertiveness, advancement, competitiveness, and earnings. In contrast, people who are femininity oriented place more emphasis on feminine values (e.g., physical attractiveness; Hofstede et al., 2010). Their behavior is usually motivated to fulfill their personal goals (Srite & Karahanna, 2006), such as life satisfaction, warm personal relationships, and a friendly work environment.

Reconstructing identity on social network platforms fulfills people’s need for achievement and physical attractiveness (Hu et al., 2015). It is suggested that physical appearance is related to feminine attributes, whereas the accomplishments and achievements are typical characteristics of masculinity (Fiore et al., 2008; Mankowski & Maton, 2010). According to the work of Hofstede et al. (2010), on average, Chinese people espouse more masculine values than Malaysian people, whereas Malaysian people espouse more feminine values than Chinese people. Therefore, we hypothesize the following:

On the basis of the cultural dimension framework, Srite and Karahanna (2006) postulated that people who are individualism oriented place more emphasis on personal task prevails, and they “prefer to act as an individual rather than as a member of a group” (Srite & Karahanna, 2006, p. 682). In contrast, people who are collectivism oriented pay more attention to maintaining harmony within groups and building relationships with others (Srite & Karahanna, 2006). Collectivistic people tend to be more interdependent and place more emphasis on interpersonal relationships (Hofstede et al., 2010).

Bridging social capital lies in the loose relationships between individuals with different backgrounds (Putnam, 2000). It is easy to establish this kind of relationship with strangers on social network platforms (Ellison et al., 2007). Previous research suggested that there is a positive association between collectivism and the bridging social capital online (Lin et al., 2012). According to the research of Hofstede et al. (2010), China has a higher level of collectivism than Malaysia. In addition, a prior study suggested that Guanxi, which refers to the personal connections with family, acquaintances, and strangers, is very important for Chinese people (X. P. Chen & Chen, 2004). Hence, it is likely that Chinese people tend to focus more on bridging social capital than Malaysian people.

In addition, bridging social capital can be beneficial for people’s personal growth and long-term success. It has been suggested that bridging social capital is positively associated with innovations (Hauser et al., 2007). It also helps people adapt to the new environment (Lin et al., 2012) and benefits their professional development (Fox & Wilson, 2015). Bridging social capital can provide people with new resources that are different from what they can obtain with existing close relationships, such as new information about job opportunities and novel perspectives (Granovetter, 1973; Putnam, 2000). Access to nonredundant information and resources has positive influences on career success (Seibert et al., 2001), which is crucial for people’s future development.

On the social network platforms, for people who reconstruct their online identity because of bridging social capital, they are motivated to make friends with other online users (Hu et al., 2015). Being more collectivistic and more long-term orientated (Hofstede et al., 2010), Chinese people are likely to place more emphasis on bridging social capital than Malaysian people. A reconstructed online identity can facilitate the process of gaining bridging social capital (Hu et al., 2015). As such, we posit the following:

According to Hofstede et al. (2010), Chinese society is more restrained than the Malaysian society in some aspects. For example, Chinese people may feel that their behavior is restrained by strict rules and social norms (Hofstede et al., 2010). The relaxing environment online (e.g., social network platforms) makes individuals feel less pressured, enabling them to behave and express themselves freely. People who feel restrained can take advantage of the benign disinhibition effect and express their suppressed feelings (Suler, 2004). In addition, Chinese people do not accept the inequality in power distribution as much as Malaysian people do (Hofstede et al., 2010). It is likely that Chinese people have more complains about the hierarchical society than the Malaysian people. On the internet, people perceive a lower level of surveillance and feel that their online behaviors are disconnected from their real life (Selwyn, 2008); thus, individuals can capitalize on the toxic disinhibition and give vent to their negative emotions (Suler, 2004).

With a reconstructed identity, people tend to feel less inhibited and free to behave in the way they want (Hu et al., 2015). Being more restrained by social norms and less tolerant of the inequality in power distribution, the online disinhibition effect (both benign and toxic effects) may be more appealing to Chinese people than to Malaysian people. Therefore, we hypothesize the following:

It has been suggested that people who are more uncertainty-avoiding tend to be more anxious, less willing to take risks, and more concerned over the outcomes of their behaviors (Hofstede et al., 2010; Srite & Karahanna, 2006). Sharing private information on social network platforms can lead to uncertain outcomes. The personal information may be accessed by unauthorized third parties or be misused by malicious individuals (Gross & Acquisti, 2005). Given that people who are uncertainty-avoiding are less tolerant of ambiguous situations, Milberg et al. (1995) posited that they are likely to have greater concerns over privacy. Cao and Everard (2008) also suggested that uncertainty avoidance is positively associated with people’s privacy concerns and awareness.

By reconstructing online identity, people can keep their private information safe (Hu et al., 2015), and hence reduce the uncertainty of online interactions. According to Hofstede et al. (2010), Malaysian people have a higher level of uncertainty avoidance than the Chinese. Being less tolerant of uncertain situations, it is likely that Malaysian people may feel more threatened by privacy issues than Chinese people. Thus, we posit the following:

Data were collected from China and Malaysia using an online survey. The online survey was adopted as the method for data collection in this study, because online surveys have many advantages over traditional paper surveys (such as global reach, cost-efficiency, and greater control on design and coding; Lefever et al., 2007; Touvier et al., 2010; Wright, 2005). Scholl et al. (2002) proposed that the lack of representativeness of online survey samples is likely to disappear because an increasing number of people have access to the internet. To make the sample as representative as possible, the online survey was posted as a public message on social network platforms so that it could reach as many users as possible. Given that Chinese people and Malaysian people use different social network platforms, the survey questionnaire only measured people’s perceptions of motivations for online identity reconstruction in general. Participants were asked to answer the survey questions on the basis of their overall experience of social network platform use, not limited to a specific one. The survey was anonymous. Only some very basic personal information was required, such as gender and age range. We asked participants to select their age range instead of providing actual age to protect their privacy and facilitate the completion of the survey.

Regarding the required sample size, previous research suggested that a sample with 384 respondents is sufficient for a target population with 1 million people at the confidence level of 95% and a margin of error of 5% (Sarantakos, 2012). As the population size increases, the required sample size increases at a diminishing rate and remains relatively constant at slightly more than 380 cases (Krejcie & Morgan, 1970; Sarantakos, 2012). Therefore, the sample size of this study in both China and Malaysia should be larger than 384 to produce accurate results.

A total of 815 respondents (418 Chinese and 397 Malaysians) were involved in this study. The Chinese sample consisted of 217 (51.9%) male respondents and 201 (48.1%) female respondents, whereas the Malaysian sample consisted of 187 (47.1%) males and 210 (52.9%) females. While Chinese respondents were ethnically homogeneous, Malaysian respondents were from different ethnic groups. Table 1 presents the summarized demographic information of participants.

The vanity scale contained eight items—four items for physical vanity and the other four items for achievement vanity (Netemeyer et al., 1995). The items for bridging social capital were adapted from Williams (2006). The disinhibition scale contained four items—two items for benign disinhibition and two items for toxic disinhibition (Denollet, 2005; Stunkard & Messick, 1985). Privacy concerns were assessed by three items adapted from Son and Kim (2008). The detailed survey items are presented in the appendix.

Participants’ responses to the survey items were measured with a 7-point Likert-type scale (1 = strongly disagree, 7 = strongly agree). The questionnaire was created in English. To collect data from China, the questionnaire was translated into Chinese by one researcher and back-translated by another two bilingual individuals to ensure the consistency between the two versions of the questionnaire. In the pilot study, which involved both Chinese and Malaysian participants, the survey items showed good reliability and validity. Thus, we kept all the items in our formal questionnaire. During the data collection phase, the Chinese questionnaire was used in China, and the English version was used in Malaysia where English is widely used as a second language (Tamam, 2011; Wan Jafar, 2014).

The reliability of the survey items was evaluated using Cronbach’s alpha (CA), while the construct validity was assessed using confirmatory factor analysis. As shown in Table 2, the CA values of the survey items were greater than .7 in both the Chinese and the Malaysian samples, indicating that the reliability of the survey items was satisfactory (Fornell & Larcker, 1981; Nunnally, 1978). In addition, the factor loadings of the survey items were greater than the accepted threshold 0.70 (Hair et al., 2009), and the average variance extracted (AVE) values of the constructs exceeded the suggested threshold 0.50, thus indicating that the convergent validity of the constructs was sufficient (Fornell & Larcker, 1981).

Discriminant validity was evaluated by comparing the square root of AVE value with the construct correlation. The square root of the AVE value of a certain construct should be greater than the correlations between this construct and the others (Hair et al., 2009). Tables 3 and 4 present the construct correlations and the square roots of AVE values for the Chinese and Malaysian samples. As shown in the tables, for both the Chinese sample and the Malaysian sample, the square roots of AVE values were greater than all the correlations between different constructs, thus indicating adequate discriminant validity.

Researchers who use same-source data (such as self-reported data) may face the problem of common method variance, which may cause bias in data analysis (Richardson et al., 2009). We used Harman’s single-factor test and the common latent factor method to detect potential common method variance in this study (Harman, 1960; Podsakoff et al., 2012).

The results of Harman’s single-factor test indicated that the single factor did not explain the majority of the variance for both the Chinese and Malaysian samples (39% for the Chinese sample and 28% for the Malaysian sample). For the common latent factor method, the variance explained by the common method should be less than 50% (Lowry et al., 2012). The results suggested that for the Chinese sample, 17.6% of the variance was due to the common method bias. For the Malaysian sample, the common method explained 24% of the variance. On the basis of the results of Harman’s single-factor test and the common latent factor analysis, it can be concluded that common method bias is not a big concern in this study.

To capture an overall picture of the cultural differences, we first analyzed the data collected from China and Malaysia as a whole. After this, we divided the Malaysian sample into three groups: the Malaysian-Malays, the Malaysian-Chinese, and Others (i.e., the Malaysian-Indians and the Malaysian-Others). We then analyzed the data collected from Chinese participants, the Malaysian-Chinese, and the Malaysian-Malays to examine whether people from different ethnic groups are motivated differently. Because the main focus of this study is Chinese and Malaysian culture, the Malaysian-Indians and the Malaysian-Others were excluded in the analysis of ethnic groups.

It has been suggested that parametric tests (such as t test and analysis of variance [ANOVA]) could be used with Likert-type scale data (Elliott & Woodward, 2007; Norman, 2010; Pallant, 2007). Parametric tests are robust to violations of test assumptions (Norman, 2010). Previous researchers found that parametric tests are more powerful, except when the distributions of the data are the most nonnormal ones (such as mixed-normal distribution; Rasmussen, 1989). Parametric tests have similar Type I error rate with nonparametric tests (Gregoire & Driver, 1987), but have lower Type II error rate (Rasmussen, 1989). Therefore, we used parametric tests in this study. T test and ANOVA were conducted to analyze the data. Table 5 summarizes the results of t test.

As shown in the table, Malaysian participants scored significantly higher on all the physical vanity items than Chinese participants. This suggests that, in comparison with Chinese participants, Malaysian participants were more likely to reconstruct their identity to increase their physical attractiveness on social network platforms. Therefore, H1 was supported. However, there was no significant difference in achievement vanity, except that Chinese participants showed greater concern in comparing their own success with others’. Thus, H2 was not supported. In addition, Chinese participants scored significantly higher on all the bridging social capital items than Malaysian participants, indicating that Chinese participants were more inclined to be motivated by bridging social capital during identity reconstruction on social network platforms, aiming to meet new friends and build various new social connections. Therefore, H3 was supported. Moreover, Chinese participants scored significantly higher on all the disinhibition items than Malaysian participants, suggesting that Chinese participants were more inclined to be motivated by the online disinhibition effect (both benign and toxic disinhibition) when reconstructing their identity online. Therefore, H4 and H5 were supported. Meanwhile, when compared with Chinese participants, Malaysian participants scored significantly higher on privacy concern items, indicating that Malaysian participants were more inclined to reconstruct their identity for the purpose of protecting their privacy on social network platforms. Therefore, H6 was supported.

In addition to examining the differences between Chinese and Malaysian participants, the repeated measure one-way ANOVA was performed to examine whether participants from the same country weight these motivations differently and which motivation is more important for them. The results revealed significant differences in the weight of the motivations in the Chinese sample, F(3.278, 1366.877) = 44.259, p < .001, η² = 0.096. Specifically, Chinese participants had significantly higher mean scores on bridging social capital than on other motivations (as shown in Figure 1), thus indicating that bridging social capital was more important than other motivations for the Chinese participants when they reconstruct their online identity.

In the Malaysian sample, significant differences were also found in the weight of the motivations for online identity reconstruction, F(4.221, 1671.506) = 191.409, p < .001, η² = 0.326. Specifically, all the differences between different motivations were significant. As shown in Figure 2, privacy concern was the most important motivation for the Malaysian participants.

We also made comparisons between different ethnic groups (the Chinese-Chinese, the Malaysian-Chinese, and the Malaysian-Malays). The results (as shown Table 5) indicated that, in comparison with Malaysian-Chinese participants, Chinese-Chinese participants scored significantly higher on all the items of bridging social capital, toxic disinhibition, one of the two benign disinhibition items, and scored significantly lower on all the items of privacy concerns. However, there were no significant differences in physical vanity and achievement vanity between the Malaysian-Chinese and the Chinese-Chinese. In addition, the Malaysian-Malays scored significantly higher on the items of physical vanity and bridging social capital than the Malaysian-Chinese. The Malaysian-Chinese had similar mean scores with the Malaysian-Malays on the remaining motivations (i.e., achievement vanity, begin and toxic disinhibition, and privacy concerns).

The cultural differences were further investigated by examining the differences between ethnic groups (i.e., the Chinese-Chinese, the Malaysian-Chinese, and the Malaysian-Malays). It was found that participants from different ethnic groups were also motivated differently. Specifically, when compared with the Malaysian-Chinese, the Chinese-Chinese were more inclined to reconstruct their identity for bridging social capital and disinhibition (both benign and toxic disinhibition), but less likely to be driven by privacy concerns. Moreover, the Malaysian-Malays were more inclined to reconstruct their online identity due to physical vanity and bridging social capital than the Malaysian-Chinese.

An interesting finding was that the Malaysian-Chinese participants were less motivated by bridging social capital than the Chinese-Chinese and the Malaysian-Malays. The language might be a barrier that makes the Malaysian-Chinese participants less passionate in seeking bridging social capital. In Malaysia, people from different ethnic groups prefer their mother tongues. The Malaysian-Malays speak Bahasa Melayu (the Malaysian language), whereas the Malaysian-Chinese prefer Mandarin and Cantonese (Wan Jafar, 2014). Although English is widely accepted as a second language, individuals face difficulties in communicating with people from other ethnic groups, because not everyone speaks English fluently (Tan, 2005). The Malaysian-Chinese may be more likely to encounter linguistic difficulties than the Malaysian-Malays and the Chinese-Chinese because the dominant group in Malaysia is the Malaysian-Malays and the official language is Bahasa Melayu (Wan Jafar, 2014), whereas in China, most people speak Mandarin (Odinye, 2015). The linguistic advantages are likely to facilitate the process of gaining bridging social capital for the Chinese-Chinese and the Malaysian-Malays, making them more motivated by bridging social capital in online identity reconstruction, although this needs to be verified in future research.

The motivational differences between the Chinese-Chinese participants and the Malaysian-Chinese participants were nearly consistent with those between the Chinese and the Malaysian (as a whole) samples. However, the differences in motivations for identity reconstruction between the Malaysian-Malay participants and the Malaysian-Chinese participants were not consistent with and not as salient as those between the Chinese and the Malaysian (as a whole) samples. The similarities between the Malaysian-Chinese participants and the whole Malaysian sample could be explained by the efforts that the Malaysian government has made to improve the harmony and unity of all ethnic groups, such as balancing economic disparity and political rights (Tamam, 2009). In the multi-ethnic environment of schools, students are able to understand and communicate with peers from other ethnic groups (Tamam, 2009). In a study concerning national identity, nearly 40% of the Malaysian-Chinese participants weighted their national identity (a Malaysian) over racial identity (a Malaysian-Chinese; Tamam, 2011). Although the Malaysian government endeavors to promote the unity of different ethnic groups, some fundamental differences between the Malaysian-Chinese and the Malaysian-Malays are difficult to integrate, such as language and religion (Noor & Leong, 2013). Although most Malaysian-Malays practice Islam, the Malaysian-Chinese follow different religions (such as Buddhism, Taoism, or other traditional Chinese folk religions; Keshavarz & Baharudin, 2009). Religious beliefs may have some influences on personal values and behavior (Bernardo et al., 2016; Mohd Suki & Mohd Suki, 2015).

This study has several contributions to the body of knowledge. This study provides a deeper insight into the motivations for online identity reconstruction by taking into account national culture. The findings suggested that participants from different countries are motivated differently during online identity reconstruction. To the best of our knowledge, this is the first time that vanity and disinhibition have been investigated in a cross-cultural context. This study also adds knowledge to the cross-cultural research concerning social network platforms. Most existing cross-cultural research concerning social network platforms compared cultural differences from the perspective of individualism versus collectivism (e.g., Kim & Papacharissi, 2003; Rui & Stefanone, 2013; C. Zhao & Jiang, 2011). By taking into account multiple cultural dimensions, this study provides a better understanding of the cultural dimension framework. Furthermore, this study sheds light on the specific differences between Chinese and Malaysian users. This study not only compared the Chinese participants with the Malaysian participants but also compared participants from different ethnic groups (e.g., the Malaysian-Chinese and the Malaysian-Malays).

Practically speaking, the findings of this study can provide directions in identifying the particular needs of different users, thereby social network platform practitioners can improve their services accordingly. The findings of this study would be especially beneficial for the companies that provide social networking services in both China and Malaysia (such as WeChat). Based on the results of this study, practitioners could deploy specific strategies to better serve social network platform users from different countries. They are suggested to place emphasis on different aspects when providing services to Chinese and Malaysian users. For example, the results indicated that the Chinese users focus more on bridging social capital and disinhibition than the Malaysian users, whereas the Malaysian users pay more attention to physical vanity and privacy than the Chinese users during online identity reconstruction. Therefore, it might be a good choice for the practitioners to enhance the features that facilitate the formation of bridging social capital (e.g., new friend recommendation) and free self-disclosure in the Chinese market; at the same time, said practitioners could highlight the features related to the presentation of physical attractiveness (such as photo editing service) and privacy protection (such as privacy setting) in the Malaysian market.

For the social network platform companies which mainly serve the Chinese users (such as QQ and Sina Weibo), more attention should be paid to help users gain bridging social capital because it is the most important motivator of online identity reconstruction for the Chinese users. For the social network platform companies whose service is available in Malaysia (such as Facebook), it is important to make users feel that their privacy is being protected because the Malaysian users are more concerned about their privacy during online identity reconstruction.

Moreover, the findings of this study can be meaningful for practitioners in other domains. In addition to social network platforms, the providers of other services may also find this study valuable. For example, when a company wants to extend its services to the Malaysian market (such as mobile pay services), it should pay more attention to protecting users’ privacy. Because the findings in the current study revealed that Malaysian people are more worried about their privacy on social network platforms, it is likely that they will also be concerned about privacy when using other online services. In addition, the results revealed that Chinese people focus more on bridging social capital. Therefore, the integration of social networking features into other services (such as online shopping applications) may make the services more attractive to Chinese users. Moreover, marketers can deliver customized advertisements to different users based on their cultural background, thereby making the advertisements more effective.

While contributing to the literature about online identity reconstruction and cross-cultural research, the current study also has limitations. Although Hofstede’s cultural dimension framework provides a foundation for cross-cultural research, it is not free from criticism. For example, Ess and Sudweeks (2005) questioned the generalizability of Hofstede’s results because the data were collected from a single organization. There are also concerns over cultural changes. It is uncertain whether Hofstede’s data are still reliable for representing contemporary reality (Inglehart & Baker, 2000). Although most of the findings in this study are in line with Hofstede’s framework, future research is recommended to combine the framework with other theories to elicit more comprehensive insights about online identity reconstruction.

In this study, the participants were recruited using online surveys. However, the number of online users is large, and there are many “lurkers” who only read posts and do not make their presence known to others. It is often difficult to obtain comprehensive knowledge about the target population and track the nonresponse rate (Andrews et al., 2003). Considering the disadvantages of convenience sampling method, the findings of this study may have limited generalizability. Future studies are suggested to use probability sampling methods to validate the results and generalize the findings to a broader population. Given that the authors cannot speak the Bahasa Melayu (the Malaysian language), we cannot ensure the quality of the questionnaire if it is translated into the Malaysian language. Therefore, English questionnaires were used in Malaysia. However, a questionnaire in the native language may help the participants understand the research better. The data in this study were collected at a specific point of time; thus, the limitations of cross-sectional studies should be considered when generalizing the findings. Future researchers could design a longitudinal study to examine whether people’s motivations change over time. At the same time, they can measure the outcomes of online identity reconstruction by comparing participants’ status before and after online identity reconstruction. Such a longitudinal study can reveal whether people obtain what they want (e.g., more bridging social capital) after reconstructing their online identity.