Interest-oriented versus relationship-oriented social network sites in China
First Monday

Interest-oriented versus relationship-oriented social network sites in China by Weiyu Zhang and Rong Wang



Abstract
This paper examines interest–oriented vs. relationship–oriented social network sites in China and their different implications for collective action. By utilizing a structural analysis of the design features and a survey of members of the social networks, this paper shows that the way a social network site is designed strongly suggests the formation and maintenance of different types of social ties. The social networks formed among strangers who share common interests imply different types of collective action, compared to the social networks that aim at the replication and strengthening of off–line relationships.

Contents

Introduction
Collective action and social network sites
Social network sites in China
Methodology
Results
Discussion and conclusions

 


 

Introduction

As social network sites (SNSs) become extremely popular among Internet users, research attention has been drawn to an array of Web sites such as Facebook, YouTube, LinkedIn, MySpace and many others. The focus has often been on contemplating the public/private nature of these sites (e.g., boyd, 2007), identifying the needs and gratifications associated with the usage of SNSs (e.g., Bumgarner, 2007; Joinson, 2008), discussing the risks of using SNSs (e.g., Ibrahim, 2008; Livingstone, 2008), and evaluating the contribution of SNSs to social capital (e.g., Ellison, et al., 2007) and political participation (e.g., Skoric, et al., 2009). SNSs as a genre of Web sites have been well clarified but the diversity within SNSs has yet to be discussed in light of academic thinking. A pioneering attempt was made by Papacharissi (2009), when she compared the underlying structures of three different SNSs and analyzed how their structures influence user interactions. This lack of evidence hinders our understanding of the complexity of SNSs because the phenomenon is simplified to a handful of successful cases such as Facebook. This paper is devoted to studying one dimension of the diversity of SNSs by distinguishing between interest–oriented and relationship–oriented SNSs.

This paper utilizes collective action as a guiding concept to look into the political implications of interest–oriented vs. relationship–oriented SNSs in China, an emerging civil society. The modernization and industrialization have fundamentally changed the way Chinese citizens associate with each other. From the focus on primary groups (e.g., extended family, friends and colleagues) to the booming of interest groups (e.g., the hiking club), the Internet plays a significant role in transforming social relations in China. Given that social networks are the foundation of collective action, the transformed social relations also change the outlook and essence of contemporary collective action. We argue that the structural design of different sites affords and encourages different types of social networks and networking behaviors. Relying on a structural analysis of the Web sites and an online survey of members of social networks, we try to show how an interest–oriented SNS functions differently in allowing and promoting collective action in comparison to a relationship–oriented SNS. The implications of such spaces for both Chinese civil society and collective action in general are discussed.

 

++++++++++

Collective action and social network sites

Collective action, as defined by Bimber and colleagues (2006), refers to a set of communication processes involving the crossing of boundaries between private and public life. Social network theories are used to explain collective action because the crossing of boundaries is a social behavior. Ties with different strengths in social networks have different influences on collective action. Strong ties seem to be particularly effective in nurturing social trust and fostering social capital (Putnam, 2000). However, there are also weaknesses in strong ties. Macy and colleagues (1997) argued that strong ties may discourage the use of peer pressure to sanction free–riding behaviors. Moreover, Gargiulo and Benassi (1998) found that cohesive ties function as a source of rigidity that hinders the coordination of complex organizational tasks. Strong ties may not be as flexible as weak ties with regard to the accommodation of change.

In addition, Granovetter (1981) argued that weak ties have their respective strengths — weak ties that bridge different social networks are particularly efficient in organizing collective action that involves novelty or controversies. Empirical findings show that when the calls for collective action are new or controversial, groups formed on the basis of weak ties, compared to groups based on strong ties, are more successful in recruiting members and mobilizing resources (Steinberg, 1980). Weak ties perform better in such a situation because they are more likely to introduce diverse information (Hansen, 1999) and bring different network segments together (Granovetter, 1981).

Coleman (1990) argued that opportunities for collective action were threatened by the decay of a wide range of traditional civic associations that were once to be the social network sites of face–to–face engagement. Putnam (2000) found that these relationship–oriented groups, many of which were dated from the American industrial revolution and Progressive eras, have suffered nearly universal declines in membership (often declining 50 percent from peak twentieth–century levels). In contrast, the anonymous interest–oriented groups have grown rapidly. These groups typically involve a shared interest, anonymous membership, the exchange of some kind of value such as dues for political representation or information and newsletters, but no personal interaction or accountability among members. This paper distinguishes relationship–oriented networks from interest–oriented networks by looking at the organizational principle of the networks. In relationship–oriented networks, the principle is to establish and maintain strong social relations. For instance, members of a bowling club may share an interest in bowling, but the way in which they find out about each other is through existing social contacts (neighbors, friends of friends, etc.) and the way in which they maintain their relationship is through regular personal interactions (going bowling together every week). In contrast, in interest–oriented networks, strangers get together because of the shared interest(s). They find each other through their shared interest(s) (e.g., subscribing to the same newsletter) without having known each other beforehand. They maintain a lukewarm relationship without intensive personal interactions. Nevertheless, they can work together for collective action such as signing a petition, donating to a cause, and so on.

Most studies on SNSs focus on relationship–oriented sites such as Facebook and MySpace [1]. Although the functions such as search can facilitate the formation of new ties (Ellison, et al., 2007), these relationship–oriented sites are found to be mainly used to develop strong ties among existing social contacts (boyd and Ellison, 2007). Users of relationship–oriented SNSs regularly interact with only a sub–portion of their listed contacts (e.g., friends on twitter) and their levels of involvement (e.g., number of tweets) are correlated with the number of this sub–portion rather than the total number of contacts (Huberman, et al., 2009). Gilbert and Karahalios (2009) found that the intensity of interaction (e.g., number of words exchanged in wall posts on Facebook) serves as the best predictor of the perceived strength of ties. Analyzed together, the two studies suggest that users of relationship–oriented SNSs only interact with their strong ties on a regular basis. One problem with strong ties is that they tend to be homogeneous (Flanagin, et al., 2006). Homogeneity or homophily in social networks may discourage tolerance and encourage the enclaving of small groups, which is argued to be unhealthy for democracy (Sunstein, 2007). If we define collective action as crossing boundaries between public and private life, homogeneous networks seem to reinforce the connections in private life but demonstrates no particular strength in turning private activities into public ones. In other words, homogeneity of the network puts constraints on the scale and the type of collective action that may take place.

SNSs seem to encourage homogeneity if we only examine the relationship–oriented sites (Thelwall, 2008; Vie, 2007). Liu and colleagues (2006) suggested that the profiles on relationship–oriented sites imply deeper patterns of culture and taste. But this fabric of taste, as described by the authors, is only a latent one. This means that private interests such as music, books, films, and food revealed in user profiles are not the primary organizational principle of the social networks afforded by these sites. Offline pre–existing contacts dominate the formation of ties on such sites. Our investigation of the phenomenon of SNSs is limited by the focus on the relationship–oriented sites. The interest–oriented sites that privilege the formation of new ties among strangers who share some common interests and keep such connections as weak yet bridging ties may provide distinct implications regarding collective action.

 

++++++++++

Social network sites in China

This paper aims to examine such a site in China, namely, Douban.com, in comparison to the Chinese version of Facebook, Xiaonei.com [2]. Internet in China has gone through a rapid development since the 1990s. The most recent statistics (China Internet Network Information Center [CNNIC], 2009) showed that the amount of Internet users in China has reached 298 million, with 279 million broadband users. By the end of 2008, the Internet penetration rate of 22.6 percent in China has surpassed the global average level of 21.9 percent for the first time. Nineteen percent of Internet users visit SNSs, including dating sites, sites targeted at students, and those targeted at professionals. Although Chinese users still spend most of their online time on entertainment and socialization, the Internet becomes a significant source of information and a powerful tool for civic engagement. For example, CNNIC (2009) found that among surveyed Internet users, 61.8 percent of them agreed that they first learned of important news from the Internet and 41.9 percent agreed that the Internet is the main channel through which they can express their opinions.

In this article, we compare Douban.com to Xiaonei.com, a Chinese version of Facebook. Douban.com was launched on 6 March 2004, and was targeted at the general public. The interface has remained almost the same since then, which is simple and designed for practical uses. Initially, this site focused on books and invited members to post book reviews. In May and July 2005, Douban added movie and music applications respectively. A recommendation system was created to boost sharing behaviors among members. People can rate books, movies and music albums and use the system to share their comments with other users. Douban had already attracted one million registered users in November 2007.

Xiaonei.com was established in December 2005, at first restricting its users to college students. The initial function of Xiaonei was to incite members to communicate with friends by updating them on their current status. Zheng and Lin (2008) found that homogeneity of members is one of the characteristics of Xiaonei. Xiaonei continued to expand its scope and opened access to the high school and young professional markets at the end of 2007. Xiaonei finally extended its access to everybody by August 2009 and changed its domain name into Renren.com. The most recent report by Xiaonei.com (released in April 2009) indicates that there were more than 40 million registered users, with a third of the accounts being active (around 13 million). Figure 1 shows the current state of both sites. Xiaonei is clearly more popular than Douban. The daily reach of Xiaonei ranges from 0.1 to 0.6 percent, whereas the reach of Douban ranges from 0.3 to 1.1 percent. The differences are stable over time.

 

Figure 1: Daily reach of douban.com and xiaonei.com
Figure 1: Daily reach of douban.com and xiaonei.com.
Note: Source = Alexa.com. Reach is determined by the number of unique Alexa users who visit a site on a given day.

 

 

++++++++++

Methodology

This paper utilizes two methods to analyze the differences between interest–oriented and relationship–oriented SNSs. The first method is a structural analysis of the two Web sites, examining the components or functions that are supported via the design of the online spaces. Examples are used to illustrate the different designs. Although the structural features of the Web sites can suggest certain usages, we are not sure whether users really follow these suggestions and behave as suggested. A second method, an online survey of members of a purposive sample (i.e., four social networks recruited from the Web sites, two each site), is operated in order to provide data that tell us how users actually utilize the Web sites.

Structural analysis

The two authors are registered members of both Web sites. The researchers take advantage of their experience to fully explore the functions and components afforded by the Web sites. The focus of the structural analysis is on the possibilities enabled by the available options rather than a systematic content analysis of any of the options, such as profiles. However, in the course of the analysis, the authors examined more than 200 profiles from Douban and approximately 300 profiles from Xiaonei. In addition, sub–sites, forums, groups, and various applications were studied, as well as site documents (FAQs, privacy statements, help, terms of use, etc.) and relevant news reports.

Survey

As a verification of the structural analysis, we conducted an online survey using a purposive sample of the networks formed within the Web sites. Networks, rather than individual users, were sampled because our research is more interested in the differences between networks formed on the two Web sites. Two networks were selected from each site. One network belongs to an experienced user (defined as using the Web site regularly for more than one year) and the other belongs to a new user (defined as using the Web site regularly for less than one year). Invitations to survey were sent out to the contacts of these two users. The response rates to individual invitations were around 30 percent on both sites. The total sample size was 186, in which 94 were Xiaonei users and 92 were Douban users. The survey responses were collected during March and April, 2009.

Our sample has an average age of 23 and an average education of 16 years, which equals a college degree in China. About 54 percent of the sample were female. Over 70 percent of the respondents were relatively experienced users who had been using the two sites for more than one year. The daily usage was evenly distributed among 30 minutes to one hour (27 percent), one to two hours (23 percent), and two to three hours (27 percent) [3]. Our respondents were actual users of the sites rather than mere usernames. Most respondents know about these sites through friends or colleagues (65 percent). However, if we broke down the data according to the two sites, xiaonei users almost exclusively rely on existing social contacts to know xiaonei (94 percent) whereas only 36 percent of douban users knew about douban through old ties. A significant portion of douban users (27 percent) found douban through search engines.

New vs. old ties. The question was phrased as “among your Xiaonei/Douban friends, how many are friends or acquaintances that you already know in your offline life?” 5 refers to “almost none”, 4 to “between 20 percent and 50 percent”, 3 to “between 50 percent and 80 percent”, 2 to “more than 80 percent but less than 90 percent”, and 1 refers to “all of them”.

Bridging social capital. Ellison and colleagues (2007) developed indices to measure different types of social capitals. We used a shortened version of measures of both bridging and bonding social capital. Bridging social capital included seven items: (1) Interacting with my friends makes me want to try new things; (2) Interacting with my friends makes me interested in what people unlike me are thinking; (3) Talking with my friends makes me curious about other places in the world; (4) Interacting with my friends makes me feel like part of a larger community; (5) Interacting with my friends makes me feel connected to the bigger picture; (6) Interacting with my friends reminds me that everyone in the world is connected; (7) Interacting with my friends gives me new people to talk to. A five–point Likert scale was used, with 1 meaning “totally disagree”, to 5 meaning “totally agree”. Cronbach’s alpha is .90 for the entire scale. Therefore, a variable that averages through the seven items was created to represent bridging capital.

Bonding social capital. The index included five items: (1) There are several people I trust to solve my problems; (2) The people I interact with would put their reputation on the line for me; (3) There is someone of my friends I can turn to for advice about making very important decisions; (4) The friends I interact with would be good job references for me; (5) When I feel lonely, there are several people I can talk to. The same five–point Likert scale was used. Cronbach’s alpha is .82 for the entire scale. A variable that averages through the five items was created to represent bonding capital.

Self–reported usage. The usage was classified into five categories: (1) I have used Xiaonei/Douban to check out someone I met socially; (2) I use Xiaonei/Douban to learn more about my classmates/colleagues; (3) I use Xiaonei/Douban to learn more about other people living near me; (4) I use Xiaonei/Douban to keep in touch with my old friends; (5) I use Xiaonei/Douban to meet new people. Again, the five–point Likert scale was used for each measure.

 

++++++++++

Results

The boundaries between private and public

When defining collective action as crossing the boundaries between private and public life, it becomes necessary to first identify what the boundaries are. Papacharissi (2009) argued that there are three levels of boundaries in social network sites. At a preliminary level, the boundary refers to criteria for membership; at a secondary level, the boundary means protocols for access to private information; and, at a tertiary level, the boundary is the ability to control your own surroundings.

Using the same hierarchy of boundaries, we examined how Xiaonei and Douban define their public and private arenas through a scrutiny of the functions they provide, along with the survey data about actual usage behaviors. At the first level, both Web sites are currently open to everyone. An easy registration will grant users the access to the Web sites. Xiaonei requires more personal information than Douban. Whereas Douban only asks for e–mail, password, and a username, Xiaonei requires users to disclose their real names, gender, and occupation. This very first step of registration reveals that Xiaonei desires or encourages a membership that directly links to an individual’s offline identity. In contrast, Douban’s criteria for membership are looser. Our survey asked whether respondents used their real names as user names. The results confirm that almost all Xiaonei users registered with their real names, whereas only eight percent of Douban users chose their real names as usernames.

At the second level, the distinction between the private and the public is set up via access to member profiles. Xiaonei allows different privacy settings to control who can access an individual’s profile. Profiles can be completely public and visible to anybody. They can be visible to friends only and a real name, a snapshot and a fraction of friends list are visible to visitors. They can even be invisible to anybody, including the user himself. Douban users, in contrast, are offered a limited range of controls over the access to their profiles. Douban profiles are public, without an option to hide the front page. Anybody can click on a username and see this person’s front page. Users are only allowed to choose not to be searchable by username and e–mail address. When users add in movies/books/music albums, they can check the option of “do not let others know.” As a consequence, the entry will not show up in the public profile. The user can also choose to make a blog entry only visible to himself. However, the access to information about friends, group memberships, activities, and message board is not allowed to be blocked. The information has to be part of a user’s profile and visible to all visitors. Therefore, the boundary at the second level is more controlled on Xiaonei than on Douban.

At the third level, both Web sites allow users to customize the functions and components they use. Douban offers their users the freedom to arrange the layout of their front page and decide what is broadcasted. Broadcasting is a function similar to the status updates on Facebook. It posts a short sentence describing any moves a user made, including adding a movie/book/music album, adding a friend, joining a group, writing a blog entry, uploading a picture, recommending and sharing, and participating in an activity. Douban allows users to set who is able to see these updates, ranging from nobody to anybody. Moreover, Douban utilizes Web 2.0 technologies to aggregate individual user data and reflect the aggregation on their Web pages through ratings and rankings [4]. This means that a user’s private actions, such as rating a movie, become one piece of contribution to public knowledge, even if the user decides to hide this rating from others. The anonymous aggregation of private data fundamentally changes our presumption of a natural division between the public and the private. In this sense, Douban users cannot opt out of the aggregative modeling used by Douban designers and thus a part of their surrounding (i.e., the recommendation system) is out of their control. Xiaonei, in contrast, allows users to control the visibility of both their activities and their friends’. Xiaonei users are able to not only set who can see their updates but also decide whose updates they want to see. For example, they may choose not to see the updates about picture posting. Therefore, at the tertiary level, Xiaonei provides users with a stronger control over their own surroundings compared to Douban.

In summary, the boundaries between the private and the public are more up to the users’ control on Xiaonei than on Douban, probably because Xiaonei profile is directly linked to a user’s off–line identity. For Xiaonei users who set the boundaries clear, they are not readily available to be mobilized into collective action unless the action is already salient within their networks. For Douban users, they are automatically, to certain degree, present in the public arena by being forced to keep some of their profile information publicly accessible (e.g., groups and friends). They are also forced to be exposed to friends’ status updates, which raises the chances of encountering requests of collective action. In addition, the aggregation design of Douban makes users’ behaviors, regardless of whether they are visible to other users or not, contributions to a public recommendation system. We may summarize that the mobilization of collective action is more dependent on users’ off–line contacts on Xiaonei, whereas unintentional or spontaneous exposure to requests of collective action holds a better chance on Douban.

Characteristics of social networks

How difficult it is to cross the boundaries is only one side of the story. The other side of the story talks about the motivations and perceptions of potential participants. These motivations and perceptions are individual, but still subject to the influence of social ties. Social network theories come into play by suggesting that the characteristics of social networks influence individual involvement in collective action. Based on previous research (Siegel, 2009), we examine the following characteristics of social networks that are found to be influential when explaining collective action: (1) the network structure; (2) the size of the network; (3) the prevalence of weak ties; and, (4) the individual motivations for getting involved in these networks.

We want to first introduce the structure of the two types of networks supported by the sites, through looking at the ways of relationship formation. Xiaonei is a Facebook–type SNS which runs on the premise that people use their real identities to establish online ties. Therefore, the formation of a relationship is predominantly based on off–line ties such as family members, schoolmates, colleagues, and neighbors. The network structure basically shows that users often belong to multiple clusters of social contacts at the same time. Only a few contacts are able to bridge different networks. Thus, a user has to go through these nodes (i.e., people) to get in touch with other people in other networks. This mode of relationship formation constrains the type of new ties a user can forge due to the limited number of the bridging nodes (i.e., people).

Figure 2 presents the network structure of Douban, which shows a distinct pattern of relationship formation through objects rather than people. The graph is generated based on data obtained through a crawler, SocSciBot [5]. The red dot represents the user and the orange dots represent the user’s contacts. The blue dots represent objects, referring to three things: movies, books, and music albums. The white dots represent groups, activities, and other links [6]. Douban allows users to connect to each other through both friends and objects. In this type of network, the nodes are not just people. Objects function as nodes that may link other objects as well as other people. A user can link to another user through a book, a movie, or a music album. For example, there are only two people who have seen one same movie and they do not know each other. The linkage between them is that Douban lists their names together under the category “who watched the movie”. A user may go to the homepage of that movie and find this other user by looking at the list. This mode of connecting is confirmed to be popular among Douban users by survey data. Our Douban respondents indicated that they found new friends by first reading their comments to an object (66 percent); second, by going through existing friends (56 percent); third, by joining the same group (52 percent); forth, by looking through the list of “who watched the movie/read the book/listened to the music album” (34 percent); and, fifth, by participating in the same activity (24 percent).

 

Figure 2: The network structure of douban.com
Figure 2: The network structure of douban.com.

 

One major difference between the network structures of Xiaonei and Douban is their different approaches to relationship formation. The approach of Xiaonei is user–centric, whereas the approach of Douban is object–centric. According to Li and colleagues (2008), the user–centric approach discovers new ties based on the social connections among users. In contrast, the object–centric approach does so based on the common objects fetched by users in an online community. Studies have found that this object–oriented approach can efficiently identify trends and types of interest as well as expertise in the online communities (Kelkar, et al., 2007). Similarly, Sripanidkulchai and colleagues (2003) took advantage of the interest–based shortcuts to efficiently locate content in decentralized peer–to–peer systems. Interest–based shortcuts work well because a peer who has a particular piece of content that one is interested in is very likely to have other items that one is interested in as well. As an object–centric network, Douban offers an efficient means to locate users that may have similar interests, which suggests a different mechanism of mobilization and organization when coming to collective action.

The sizes of network on the two sites are not very different. The survey data show that Xiaonei users had an average of 172 contacts and Douban users had 102 contacts. From a collective action point of view, the scope of the social capital we need depends on the scale of the problems we face. Therefore, size itself cannot tell us much about whether a network is more ready for collective action than another one. However, weak ties, especially bridging weak ties, are considered as crucial for collection actions, especially those involve novelty and controversies (Granovetter, 1981). Weak ties between newly acquainted users were found to be more prevalent on Douban than on Xiaonei. Over 90 percent of Xiaonei friends were users’ off–line contacts. A majority of Douban users (71 percent) indicated that less than half of their friends were off–line contacts. In addition, Douban offers a ‘friending’ option which is not reciprocal, meaning that a user can add a person to the list of “people whom I care about” (similar to the “following” list on Twitter) without getting this person’s approval of being listed. A list of “people who care about me” (similar to the “follower” list on Twitter) is generated if any users add you as “people whom I care about”. Most Douban users (82 percent when asking about “people whom I care about” and 72 percent when asking about “people who care about me”) reported that people listed under these one–way friending categories were almost all new ties. Moreover, our respondents slightly agreed that the new ties on Douban were weak [7] (M = 3.13).

The strengths of weak ties lie in the possibility that they can bridge different networks. As Granovetter (1981) emphasized, only weak ties that bridge are important for collective action. We thus asked our respondents how the social ties formed on the two Web sites help them to bridge different social networks. Using the indices created by Ellison and colleagues (2007), we measured users’ perception of bridging and bonding social capitals supported by Xiaonei vs. Douban. The data show that Douban users perceived more bridging capitals than Xiaonei users. For example, Douban users agreed that interacting with friends makde them become interested in what people who are unlike them are thinking (M = 3.67) whereas Xiaonei users slightly agreed to the same statement (M = 3.04). The difference is statistically significant (t = -4.44, ρ < .001). A full report of such findings is listed in Table 1. The table essentially shows that on all the measures of perceived bridging social capital, Douban users rated higher than Xiaonei users. In contrast, Douban users rated either lower or almost the same as Xiaonei users on all the bonding social capital measures. For instance, the classic measure of job reference shows that Xiaonei friends served a stronger function of bonding ties (M = 3.00) than Douban friends (M = 2.58, t = 3.01, ρ < .01). The perceptions were consistent with the self–reported usages of the two sites. Table 1 shows that Xiaonei users mainly used the Web site to develop strong ties among existing contacts, including friends, classmates/colleagues, people living near by, and off–line acquaintances. Douban users, however, used Douban predominantly for the purpose of forming new ties.

 

Table 1: Bridging and bonding social capitals perceived by Xiaonei vs. Douban users.
Note: Independent samples t–tests were run for each of the measures. ***ρ <.001, **ρ <.01, *ρ <.05, +ρ <.10.
 Xiaonei usersDouban users
Bridging social capital3.15***3.65
Interacting with my friends makes me want to try new things.3.04***3.82
Interacting with my friends makes me interested in what people unlike me are thinking.3.04***3.67
Talking with my friends makes me curious about other places in the world.3.21***3.71
Interacting with my friends makes me feel like part of a larger community.3.16*3.54
Interacting with my friends makes me feel connected to the bigger picture.3.24***3.77
Interacting with my friends reminds me that everyone in the world is connected.3.343.46
Interacting with my friends gives me new people to talk to.3.03***3.57
 
Bonding social capital3.11*2.87
There are several people I trust to solve my problems.3.133.18
The people I interact with would put their reputation on the line for me.3.13***2.42
I can turn to some of my friends for advice about making very important decisions.3.19*2.87
The friends I interact with would be good job references for me.3.00**2.58
When I feel lonely, there are several people I can talk to.3.123.28
 
Self–reported usage  
I have used Xiaonei/Douban to check out someone I met socially.3.05***2.28
I use Xiaonei/Douban to learn more about my classmates/colleagues.3.35***2.37
I use Xiaonei/Douban to learn more about other people living near me.3.26***2.20
I use Xiaonei/Douban to keep in touch with my old friends.3.64***2.40
I use Xiaonei/Douban to meet new people.2.80+3.10

 

In summary, the network structure of Douban, compared to that of Xiaonei, allows a different way to form new ties — through common interests (i.e., objects) rather than existing social contacts (i.e., people). This mechanism of connection encourages users to meet new people. As a consequence, Douban users tend to have more new ties than Xiaonei users. As expected, these new ties are mostly weak ties. Users also agree that these weak ties help them to establish bridging social capital. It is argued that bridging social capitals can efficiently mobilize people into collective action that involve novel and controversial issues (Granovetter, 1981). Therefore, we conclude that the Douban networks provide a better chance for mobilizing novel and controversial collective action than the Xiaonei networks.

 

++++++++++

Discussion and conclusions

This paper examines the differences between an interest–oriented SNS, Douban.com and a relationship–oriented SNS, Xiaonei.com, in terms of their structural features that enable different social networks and networking behaviors to emerge. Our data show that Douban makes the boundaries between public and private life very vague and even dissolves such boundaries through the aggregation mechanisms. As a result, Douban networks encourage the formation of new ties among strangers through shared interests such as books and movies. Moreover, these new ties often develop into bridging ties that connect distinct social networks. Xiaonei, in contrast, provides effective means to users to control the boundaries between the private and the public. Therefore, strong ties among existing social contacts are fostered on Xiaonei.

A social network that connects strangers via common interests and maintains such bridging weak ties has its significant contribution to collective action. One difficulty associated with mobilizing people for collective action is locating the relevant individuals who are interested in the cause behind the action. An interest–oriented SNS like Douban provides efficient ways to locate the relevant individuals because of three factors — the visibility of users’ interests, the connections between strangers, and the usage of interest–based shortcuts to form ties. For example, an online activity called Earth Hour 2010 attracted over 20,000 Douban users to participate. If a Xiaonei user wants to recruit people to join an activity like this, he has to rely on his social network that is based on existing offline contacts. One problem with this type of recruitment is that a user’s contacts may not be motivated to participate because the connections were not established based on the shared concern regarding environmental issues. A Douban user, for example, can go to the page of the movie An Inconvenient Truth and find the users who have rated this movie five–star and send an invitation to all these users. The diverse ways of articulating and observing users’ interests on Douban open up many channels to locate the relevant others. Because the majority of Douban contacts are strangers, the recruitment message is able to travel beyond a particular user’s limited off–line contacts, reach a broad social circle, and increase the probability of recruiting the right people. In addition, recruiting Douban friends is likely to be successful because ties that are connected by one common interest (e.g., An Inconvenient Truth) are likely to share more interests (e.g., Earth Hour).

Moreover, interest–oriented SNSs allow users to have personal interactions, which facilitate the maintenance of weak yet bridging ties. We consider interest–oriented SNSs similar to the large–scale mailing list groups (Putnam, 2000) in terms of its ability to handle massive number of users who do not know each other. However, interest–oriented SNSs overcome one critical shortcoming of these mailing list groups — impersonal interaction — which may lead to lack of social trust among the group members. Personal and direct communications between Douban users are supported by in–site mail messages, wall posts, status updates, and many other tools. Personal interaction fosters a sense of bond between users and facilitates the diffusion of messages such as a call for collective action.

Regarding the development of Chinese civil society, how to mobilize collective action among Chinese citizens is an important issue. Observers (e.g., Yang, 2009; Zhang, 2006) documented the creative acts of online activism in China and argued for a co–evolution of the Internet and Chinese civil society. Internet–based collective action is not only becoming increasingly popular in China, but also contributing to the development of civil society. Civil society in China is still under–developed due to both state control and the relatively recent modernization progress. Identifying and pursuing a common interest with strangers who do not belong to an individual’s primary groups (e.g., an extended family) are still relatively new to the Chinese. SNSs in China, adding to discussion forums (e.g., tianya.com, mop.com) and blogoshere (e.g., blog.sina.com.cn, blogbus.com), provide another means to meet other fellow citizens and act in collective forms. Both old ties and new ties are now open to the possibilities of online mobilization. An interest–oriented SNS, compared to a relationship–oriented SNS, seems to be more open to collective action that centers on new and controversial social issues because such concerns may not be well shared among a person’s primary groups.

However, a further question is what Zheng and Wu (2005) have asked: “[i]t is not a question of whether Internet–based collective action is possible … the question is whether Internet–based collective action can succeed in challenging the state.” The new ties formed on Douban and the vague boundaries between public and private life cannot guarantee Chinese citizens the power to challenge the state. The reason is simple — all Web applications are developed by programmers. These programmers can use codes to encourage collective action and they can use the same codes to forbid collective action. A good example is the shutdown of Douban groups in early 2009. Following the policy of “cleaning online content” issued by the central government, Douban was forced to examine all existing groups and delete those which contain “inappropriate content”. After this wave of shutdowns, all new groups and activities are now subject to censorship before they can be put into effect.

However we disagree that Chinese civil society will be unable to develop if its citizens cannot explicitly subvert state domination. The state has overplayed civil society for a long time when almost every aspect of social life was closely controlled. Along with the adoption of economic freedom, the Chinese are now less linked to the state and more connected to their fellow citizens. To recognize the necessity to interact with other social members in civic ways is the first step to building a civil society. Chinese SNSs in their current stage may only be able to promote shared understanding and connections that serve as the foundation of trust and cooperation. Trust and cooperation are especially valuable when individuals are among strangers, suggesting the unique contribution of interest–oriented SNSs to the formation of civil society. In addition, we do not think the state can operate in isolation altogether from civil society. The state bureaucracy involves individuals, who in turn belong to social networks. If one collective action is widespread in social networks, we can expect the state to be influenced by this collective action as well.

The limitations of this study lie in the fact that only two SNSs were included and analyzed, which point to problems of the representativeness of the sample, the sources of the differences we have found, and the generalizability of our findings. Firstly, two Web sites are by no means representative of the whole Internet, not even the entirety of SNSs in China. However, the key argument of this paper focuses on the different structural features of two sites and their influence on social networking. It would have been ideal if we had included more sites that fit the classification of interest–oriented and relationship–oriented sites. Nevertheless, we think that the sampled sites illustrate how the design of SNSs can have an impact on social networks formed in the sites, and in turn, on collective actions that may be supported.

Secondly, the differences we observe here may actually be explained by other factors such as the different groups of users (e.g., Xiaonei targeted college students while Douban has been open to all) rather than different structural features. However, our survey shows that there are no major disparities in basic demographics between the sites’ users. We tend to think that our findings reflect an interactive procedure between users and structural features. Structural features of the sites encourage certain patterns of use and users appropriate those features that meet their needs. For instance, one user may choose Xiaonei to maintain his off–line social contacts and use Douban to explore new relationships.

Third, whether our findings can be generalized to other cases is unknown. We need to examine more sites that fit into the categories of interest–oriented and relationship–oriented SNSs and test our arguments with new cases. So far, however, we have not yet identified other sites for analysis. Our future efforts will focus on locating and examining SNSs that operate as interest–oriented networks in order to test further our findings.

In addition, future studies should examine more closely innovative applications that challenge our assumptions about collective action. These studies should be placed in larger social and cultural contexts in order to fully understand their significance. SNSs, as a genre of Web sites, contain a great deal of variation and deserve further scholarly attention. End of article

 

About the authors

Weiyu Zhang is an assistant professor at Communications & New Media Programme, National University of Singapore. Her research interest lies in civic engagement and new media. She has also conducted studies on media multitasking, including both traditional and new media.
Correspondences should be sent to cnmzw [at] nus [dot] edu [dot] sg

Rong Wang is a Master’s student in Communications and New Media Programme, National University of Singapore. Her research focuses on social network analysis, online peer community and collective action.

 

Notes

1. Some may argue that dating sites also attract strangers with similar interests (e.g., looking for a life partner). However, dating sites are still relationship–oriented, in terms that users want to establish and maintain strong social relations ultimately. Dating sites fit only one criterion of interest–oriented SNSs — users find each other through shared interests.

2. In August 2009, Xiaonei officially changed its name to Renren, as well as its domain to www.renren.com. This paper keeps using Xiaonei and Xiaonei.com, which were the name and the domain when the data were collected.

3. This skewness towards heavy users actually strengthens the validity of our study, as we are more interested in the possibilities enabled by different SNSs rather than a survey of all users’ perceptions and behavior.

4. The Web site aggregates individual data points and presents the results to users via many means, such as number of users who viewed the movie, number of users who rated the movie five–star, and tags applied by users to describe the movie. The Web site also recommends relevant objects to users based on the users’ behaviors (similar to the recommendation system on amazon.com).

5. Information about this software can be found at http://socscibot.wlv.ac.uk/.

6. Douban users can create groups and any users who know this group can join. Each group has its own front page on which a discussion forum is available. Members of the groups are allowed to post and reply on the forum. Non–members are allowed to browse the posts if the group is set as pubic. Activities are similar to groups in terms of the technical functions, except that activities have a start and an end date. For example, “environmentalists” is a group and “Earth Hour 2010” is an activity. Groups are categorized under a series of themes such as arts, life, and hobbies. There are 147,352 groups on Douban as of 24 March 2010. Activities are differentiated by tags such as photography, charity, and design. The number of activities is too large to estimate.

7. The classic way to define relationship weakness is to look at the frequencies of contact. If there are few exchanges of greetings between a pair of contacts, the relationship is considered to be weak. We used a different prompt of weakness by asking for users’ perception. We explicitly asked them to perceive whether new ties formed on Douban are weak to them by using a five–point Likert scale.

 

References

Bruce Bimber, Andrew J. Flanagin, and Cynthia Stohl, 2005. “Reconceptualizing collective action in the contemporary media environment,” Communication Theory, volume 15, number 4, pp. 365–388.http://dx.doi.org/10.1111/j.1468-2885.2005.tb00340.x

danah boyd, 2007. “Social network sites: Public, private, or what?” Knowledge Tree, at http://kt.flexiblelearning.net.au/tkt2007/edition-13/social-network-sites-public-private-or-what/, accessed 26 April 2009.

danah boyd and Nicole B. Ellison, 2007. “Social network sites: Definition, history, and scholarship,” Journal of Computer–Mediated Communication, volume 13, number 1, at http://jcmc.indiana.edu/vol13/issue1/boyd.ellison.html, accessed 19 October 2008.

Brett A. Bumgarner, 2007. “You have been poked: Exploring the uses and gratifications of Facebook among emerging adults,” First Monday, volume 12, number 11, at http://www.uic.edu/htbin/cgiwrap/bin/ojs/index.php/fm/article/viewArticle/2026/1897, accessed 26 April 2009.

China Internet Network Information Center (CNNIC), 2009. “Statistical survey report on the Internet development in China” (January), at http://www.cnnic.net.cn/uploadfiles/pdf/2009/3/23/153540.pdf, accessed 21 July 2009.

James S. Coleman, 1990. Foundations of social theory. Cambridge, Mass.: Belknap Press of Harvard University Press.

Nicole B. Ellison, Charles Steinfield, and Cliff Lampe, 2007. “The benefits of Facebook ‘friends’: Social capital and college students’ use of online social network sites,” Journal of Computer–Mediated Communication, volume 12, number 4, at http://jcmc.indiana.edu/vol12/issue4/ellison.html, accessed 22 October 2008.

Andrew J. Flanagin, Cynthia Stoh, and Bruce Bimber, 2006. “Modeling the structure of collective action,” Communication Monographs, volume 73, number 1, pp. 29–54.http://dx.doi.org/10.1080/03637750600557099

Martin Gargiulo and Mario Benassi, 1998. “Trapped in your own net? Network cohesion, structural holes, and the adaptation of social capital,” Organization Science, volume 11, number 2, pp. 183–196.http://dx.doi.org/10.1287/orsc.11.2.183.12514

Eric Gilbert and Karrie Karahalios. 2009. “Predicting tie strength with social media,” Proceedings of the 27th International Conference on Human Factors in Computing Systems (Boston), pp. 211–220.

Mark S. Granovetter, 1981. “The strength of weak ties: A network theory revisited,” Sociological Theory, volume 1, pp. 201–233.http://dx.doi.org/10.2307/202051

Morten T. Hansen, 1999. “The search–transfer problem: The role of weak ties in sharing knowledge across organization subunits,” Administrative Science Quarterly, volume 44, number 1, pp. 82–111.http://dx.doi.org/10.2307/2667032

Bernardo A. Huberman, Daniel M. Romero, and Fang Wu. 2009. “Social networks that matter: Twitter under the microscope,” First Monday, volume 14, number 1, at http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/2317/2063, accessed 15 March 2010.

Yasmin Ibrahim, 2008. “The new risk communities: Social networking sites and risk,” International Journal of Media and Cultural Politics, volume 4, number 2, pp. 245–253.http://dx.doi.org/10.1386/macp.4.2.245_3

Adam N. Joinson, 2008. “‘Looking at’, ‘looking up’ or ‘keeping up with’ people? Motives and uses of Facebook,” Proceeding of the twenty–sixth annual SIGCHI conference on human factors in computing systems (Florence, Italy), pp. 1,027–1,036.

Shreeharsh Kelkar, Ajita John, and Doree Seligmann, 2007. “An activity-based perspective of collaborative tagging,” Proceedings of the International Conference on Weblogs and Social Media (Boulder, Colo.), at http://www.icwsm.org/papers/paper9.html, accessed 9 March 2010.

Xin Li, Lei Guo, and Yihong Eric Zhao, 2008. “Tag–based social interest discovery,” Proceeding of the 17th international conference on World Wide Web (Beijing), pp. 675–684.

Hugo Liu, Pattie Maes, and Glorianna Davenport, 2006. “Unraveling the taste fabric of social networks,” International Journal on Semantic Web and Information Systems, volume 2, number 1, pp. 46–78.http://dx.doi.org/10.4018/jswis.2006010102

Sonia Livingstone, 2008. “Taking risky opportunities in youthful content creation: Teenagers’ use of social networking sites for intimacy, privacy and self–expression,” New Media and Society, volume 10, number 3, pp. 393–411.http://dx.doi.org/10.1177/1461444808089415

Michael W. Macy, James A. Kitts, and Andreas Flache, 1997. “The weakness of strong ties II: Collective action failure in a self–organizing social network,” paper prepared for the Mathematical Sociology Section of the American Sociological Association Annual Meeting (Toronto), at ftp://hive.soc.cornell.edu/mwm14/webpage/WST2.PDF, accessed 9 March 2010.

Zizi Papacharissi, 2009. “The virtual geographies of social networks: A comparative analysis of Facebook, LinkedIn and ASmallWorld,” New Media and Society, volume 11, numbers 1–2, pp. 199–220.

Robert D. Putnam, 2000. Bowling alone: The collapse and revival of American community. New York: Simon & Schuster.

David A. Siegel, 2009. “Social networks and collective action,” American Journal of Political Science, volume 53, number 1, pp. 122–138.http://dx.doi.org/10.1111/j.1540-5907.2008.00361.x

Marko M. Skoric, Deborah Ying, and Ying Ng, 2009. “Bowling online, not alone: Online social capital and political participation in Singapore,” Journal of Computer–Mediated Communication, volume 14, number 2, pp. 414–433.http://dx.doi.org/10.1111/j.1083-6101.2009.01447.x

Kunwadee Sripanidkulchai, Bruce Maggs, and Hui Zhang, 2003. “Efficient content location using Interest–based locality in peer–to–peer systems,” Proceedings of the IEEE Infocomm Annual Conference (San Francisco), at http://www.ieee-infocom.org/2003/papers/53_01.PDF, accessed 9 March 2010.

Lois Steinberg, 1980, “Preexisting social ties and conflict group formation,” paper presented at the 1980 meeting of the American Sociological Association (New York).

Cass R. Sunstein, 2007. Republic.com 2.0. Princeton, N.J.: Princeton University Press.

Mike Thelwall, 2008. “Homophily in MySpace,” Journal of the American Society for Information Science and Technology, volume 60, number 2, pp. 219–231.http://dx.doi.org/10.1002/asi.20978

Stephanie Vie, 2007. “Engaging others in online social networking sites: Rhetorical practices in MySpace and Facebook,” Ph.D. dissertation, University of Arizona.

Guobin Yang, 2009. The power of the Internet in China: Citizen activism online. New York: Columbia University Press.

Weiyu Zhang, 2006. “Constructing and disseminating subaltern public discourses in China,” Javnost-The Public, volume 13, number 2, pp. 41–64, and at http://www.javnost-thepublic.org/media/datoteke/13-2-zhang.pdf, accessed 9 March 2009.

Yujun Zheng and Lin Lin, 2008. “When campus SNS sparks reality: Xiaonei net’s computer–mediated interpersonal communication,” Journal of Guangdong Polutechnic Normal University, volume 3.

Yongnian Zheng and Guoguang Wu, 2005. “Information technology, public space, and collective action in China,” Comparative Political Studies, volume 38, number 5, pp. 507–536.http://dx.doi.org/10.1177/0010414004273505

 


Editorial history

Paper received 4 February 2010; revised 15 April 2010; accepted 22 April 2010.


Copyright © 2010, First Monday.
Copyright © 2010, Weiyu Zhang and Rong Wang.

Interest–oriented versus relationship-oriented social network sites in China
by Weiyu Zhang and Rong Wang.
First Monday, Volume 15, Number 8 - 2 August 2010
http://firstmonday.org/ojs/index.php/fm/article/viewArticle/2836/2582





A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2014.