The bandwagon effect on participation in and use of a social networking site
First Monday

The bandwagon effect on participation in and use of a social networking site by W. Wayne Fu, Jaelen Teo, and Seraphina Seng



Abstract
This study examined the bandwagon effect on participation in and use of one particular SNS — Facebook — using a Singapore–based sample of active Internet users. Building on theories addressing the social mechanisms of interactive or communicative technologies, we tested the postulation that the decision of individuals to participate in Facebook would be linked to their perceptions of how widely Facebook had been diffused either at large or among their off–line contacts. Regression analyses confirm this bandwagon tendency in Facebook participation. Moreover, multiple SNS users tended to use Facebook more than all other SNSs if they perceived that it penetrated their off–line contacts more widely. The results suggest that Internet users engage in SNSs to forge personal online networks in ways complying with real–world group proximity and their perception of the prevalence of adoption rather than making random or amorphous choices.

Contents

Introduction
Facebook
Network size, perception, participation, and usage
Method
Regression analysis
Discussion and conclusion

 


 

Introduction

The widespread diffusion of social networking sites (SNS) has inspired an outpouring of research on the motivation and behavior of individuals in using these sites. Studies have pointed to social affordance as the main driver (boyd, 2008; Bumgarner, 2007; Coyle and Vaughn, 2008; Ellison, et al., 2007; Raacke and Bonds–Raacke, 2008). Although the reaction of individuals to SNSs, as with other network goods, has been predicted by personal attributes and preferences (e.g., Hargittai, 2007), it also depends on the choices of other users.

Human use of communication technologies takes place in the context of social influence (Contractor and Eisenberg, 1990; Fu and Sim, 2011; Fulk, 1993; Fulk, et al., 1990; Lin, 2003; Schmitz and Fulk, 1991). The core of this thesis is that people’s attitudes toward and decision about adopting a technology are affected by those of others, and thus will converge in spheres of social interconnectedness. Social influence is especially potent for using technologies that embed communication networks (Markus, 1987; Rice and Aydin, 1991; Rice, et al., 1990; Rogers, 2003; Rogers and Kincaid, 1981; Schmitz and Fulk, 1991), where participants desire the widest possible dyadic connections. This literature stresses that social interdependence precedes or dictates the shaping and evolution of human networks in ways such that individuals are sensitive to the network choices of their peers. When social forces and economic utility, inherently driven by communication or connection, goad actors into deferring to their predecessors’ network choices (Markus, 1987; Rohlfs, 2001; Soe and Markus, 1993), the bandwagon effect arises among participants. This study thus sought to investigate the bandwagon effect on people’s participation in a particular SNS, where participants choose one or more platforms among contending alternatives on which to base their online personal networks.

SNS users create personal profiles on the sites they have signed up with, and then invest time and effort into articulating and snowballing relationships in cyberspace. In the adoption process, people first decide whether to participate in the SNS universe; once they have opted in, they are confronted by a multitude of SNSs to choose from. Upon joining a site, a person is compelled to take into consideration the SNS preferences and choices of others. The person must then be convinced that he or she will receive a large enough benefit from belonging to this SNS in terms of the scope of reach afforded by the site, given the long–term commitment expected with the SNS affiliation. But despite the propagation of SNSs, scholars still need to understand better the social dynamics of an individual’s decision to use an SNS.

How is such a decision of an Internet user influenced by the SNS’s prevalence among her off–line social circle as well as the general public? For a person who takes up more than one SNS, is the SNS in which she engages the most related to the choices of others as well? Further, is the extent of her involvement in an SNS associated with the size of her social network (i.e., how many “friends” she owns) through the site? Such questions orbiting the “social” workings of SNSs have yet to be answered systematically. Building on the sociological theory of critical mass and the economic theory of network externalities, we thus examined the bandwagon effect on individuals’ use of one particular SNS, namely Facebook.

The network externalities theory characterizes the bandwagon effect on the economic behavior of network customers. Network externalities exist when the value to a person of consuming a product increases with the number of other users (Rohlfs, 1974; Squire, 1973; see Shapiro and Varian, 1999, for a discussion). Therein, the participation of each user means that not only does the user receive some internal (or personal) consumption value from belonging to a network, but also that some external (or social) benefit is created for other members. Because of the value externalities, a larger network will be more valuable to a prospective user insofar as it enables her to connect to more people than a smaller network would. This view has also been raised by communication scholars: The more direct communication links a person has with a set of potential or existing users of a communication system, the greater the value of the system to that person and the more likely the person will adopt it (Hiltz, 1984; Rice, et al., 1990).

As a result, individual network shoppers are attracted to that network among its competitors that has recruited the most subscribers. Given that network users defer to what preceding peers have opted for, a network service that has accumulated a larger user base becomes the de facto choice for newcomers in the market. The “big–get–bigger” progression or the “winner–takes–all” consequence is commonly seen in network contexts (Easley and Kleinberg, 2010; Frank and Cook, 1995; Katz and Shapiro, 1994).

A related research realm that considers the size effect on network diffusion is critical mass theory. Economic models have proved mathematically that network services must reach a definitive threshold in user base size, called “critical mass,” to achieve prevalence in their coverage (Economides and Himmelberg, 1995; Rohlfs, 1974). A similar concept about interactive media adoption, expounded in the communication or diffusion literature, points to the initial number of adopters necessary to fashion a self–sustaining value from the system and to stimulate subsequent collective actions in adoption (Markus, 1987; Rogers, 2003). Sociologists have contextualized the critical mass theory by identifying the conditions for a “public good” to realize universal adoption (Oliver, et al., 1985; Oliver and Marwell, 1988). A key condition they have explained, which was later refined by Markus (1987), is the reciprocal interdependence between earlier and later adopters in terms of the value reaped from joining the same system. On one hand, late adopters will join the system only when enough people have already done so to make the good sufficiently desirable. On the other hand, early adopters will join it only when they expect strong follower recruitment to come afterward; they will then remain in the system only if their expectation is fulfilled (Markus, 1987; Rogers, 2003). If not reciprocated with membership reinforcement, early adopters are likely to discontinue using the system (Rice, 1982). Any member dropout will then reduce benefits to the remaining users, thus spurring further withdrawals.

Markus (1987) illustrated the consequence of critical mass through the time trajectory of user base accumulation. His prediction is that those technologies that attain the critical user mass will spread to the entire population; those that cannot will wither to extinction. Owen and Wildman (1992) and Soe and Markus (1993) also asserted that users generally choose the communication medium most widely available within their communication community, whether or not it is their own preference, because the choice enables them to communicate with the largest number of contacts with the least effort or cost. In sum, the critical mass literature has conveyed the implications of the bandwagon effect — one’s participation in a system will be contingent on the presence of a critical mass of communication partners (Rice, et al., 1990).

In accordance with the reciprocal interdependence notion of critical mass, it is helpful to draw a subtle but crucial distinction between adoption and participation in examining SNS membership. In the literature, adoption for the most part refers to the start of using a new technology, that is, jumping on the bandwagon. In comparison, participation means continuing to use the technology until a given time, as the person has chosen not to get off the bandwagon. In the circumstance of sequential adoption, critical mass adoption may or may not antecede a given person’s adoption, depending on the person’s timing of the adoption (early versus late) and on the threshold of membership size according to the diffusion model (Rogers, 2003). But when system membership has saturated, or at least almost saturated, the person’s participation should hinge on critical mass participation. The reason, as stated above, is that one will use, or continue to use, the technology only when enough current participants exist when system membership is saturated. Indeed, discontinuation in SNS participation is common: A U.S. survey showed that 10 percent of U.S. Internet users had deleted a profile on an SNS, and that 28 percent of current social network users had stopped using an SNS (Lenhart, 2009). Thus, this study aimed to examine how one’s participation in an SNS is related to critical mass participation.

Speaking to the germaneness of this bandwagon conjecture, the SNS landscape has indeed gravitated toward domination by Facebook (Hesse, 2009). Departing from niche or specialized sites that feature a certain theme or interest (e.g., music) or that target specific demographic or geographic segments, Facebook has successfully positioned itself as an all–inclusive social medium for general networking. With its open, down–to–earth design, the site swiftly captured many adopters early on and then continued mobilizing its forces of attraction through viral peer referrals; it has since outrun its competitors in membership by a widening distance. Its strategies and designs have appeared to fuel a sign–up bandwagon among friends and acquaintances, resulting in their “hightailing it” to Facebook.

 

++++++++++

Facebook

Facebook is a social networking Web site that was initially built for university communities but that has won acceptance by the general public as well. As the leading SNS globally, the site reported in February 2010 more than 400 million active users, 50 percent of whom log in on any given day. Individuals register to join Facebook either through e–mail invitations from people they already know off–line or out of their own interest. Once registered, a user must enter personal information on the formatted Web page profile designed by Facebook, similar to other SNSs; the user then enlists other registered users as “friends” to construct her personal social network. Friends are made dyadically whenever a user accepts another’s request to befriend. Both personal profiles and social networks can be disclosed between Facebook friends; also, messages can be posted to each other’s pages. New friend–dyads can spin off from a friend’s friends when a person searches or browses her friends’ friend lists or receives the site’s automatic friend suggestions based on common friend lists.

Online social networks are chiefly an online incarnation or extension of off–line social ties, with a dual function of mapping existing connections and creating new ones (Lampe, et al., 2006; Walther, et al., 2008). Lampe, et al. (2006) found that Facebook users searched for real–world contacts much more often than they browsed for strangers to befriend. A survey found that 91 percent of U.S. teens used SNSs to connect or reconnect with existing friends (Lenhart and Madden, 2007). Also, only 0.4 percent of the Facebook friendships of college students reflected pure online encounters (Mayer and Puller, 2008). Therefore, a person could more rapidly grow her online social network (capital) upon joining if she saw more of her off–line friends on Facebook already. In sum, how popular an SNS is among a person’s off–line–world friends should predict her receptivity to joining it.

 

++++++++++

Network size, perception, participation, and usage

Participation in an SNS

The critical mass and network externality theories expect that the size of membership in a network will determine the benefits individual users receive. Even when networks are differentiated in theme or purpose (e.g., SNSs with distinct orientations or targets) and hence less substitutable, the users (current and potential) of a given network will still be sensitive about network reach, which could fundamentally shift the value of their participation within it, as explained above. Therefore, one’s decision to engage in a network is not simply a matter of personal taste and preference; it is contingent on the choices made by others. Thus, since user choices are interdependent, network shoppers will be compelled to go to one that has more connections to more people.

This bandwagon effect on network platform choice resembles the tendency for social constituents to emulate the focal choice among a critical mass of their peers (Bass, 1969; Simon, 1954). Such social conformity has been studied in a range of behavioral sciences. But despite the dependence of a network’s value on size, individuals may or may not learn, or be able to learn, a network’s exact membership size in reality (Lou, et al., 2000). In fact, it is through the perceptions of network membership gathered from daily–life experiences, peer information sharing, or media announcements that an installed user base is able to shift the intentions of people to participate. Individuals rely on impressions about a network’s current reach to appraise its long–term value during the formation of a network market. The perceived market penetration of a network molds individuals’ expectations as to whether this service will thrive or fade eventually, and so influences their present decision of whether to participate, or continue to do so. The enduring viability of a network among alternatives matters, because people want to avoid wasting time and money on one that is unlikely to flourish anyway. The size–value perception underlies what economists call a self–fulfilling prophecy about market responses to contending networks. If a critical number of consumers foresee that a network will eventually spread in their market and thus adopt it, later adopters will respond to this expectation (Katz and Shapiro, 1986; 1994). Scholars have found that the perception of how commonly a communication medium is used, or its critical mass, is predictive for the decision to adopt (Chun and Hahn, 2007; Ilie, et al., 2005; Lou, et al., 2000; Sledgianowski and Kulviwat, 2009; Soe and Markus, 1993; Van Slyke, et al., 2007; Wang, et al., 2005).

Following this logic, we arrived at our first hypothesis about Facebook participation:

H1: Individuals who believe that Facebook has the most active users will be more likely to participate than those who do not.

Previous research has also found that the makeup of a personal network affects the adoption of communication technology (Boase, 2010). Although the general diffusion of Facebook can affect one’s inclination to participate, the presence of members on the site with whom one wants to be in contact constitutes an immediate enticement to participate. The prospective value of a communication system to a person increases with the extent to which others to whom the person desires to connect have adopted the system (Markus, 1987; Rice, 1982; Rice and Shook, 1988). Researchers have termed the subset of the total network–user base of an online service consisting of direct real–life contacts one’s “personal network” (Birke and Swann, 2006). A rational person should want to select a platform where he or she will encounter a large personal network. Studies have shown that users of various communication services tend to choose the same service networks as their contacts in larger groups (Birke and Swann, 2006; Chun and Hahn, 2007; Corrocher and Zirulia, 2009; Kim and Kwon, 2003). Soe and Markus (1993) tested the critical mass effect on a person’s use of communication media by measuring the perceived diffusion of the media within the person’s own communication community. Ilie, et al. (2005), Lou, et al. (2000), Sledgianowski and Kulviwat (2009), and Van Slyke, et al. (2007) also validated this conjecture.

In addition to the “push” motivation described above, a personal network also wields a “pull” force in recruiting members. The pull force stems from the side of existing members, who have a motive to enroll more contacts into their networks to achieve more extensive connections (Henkel and Block, 2007; Markus, 1987; Rogers, 2003). This incentive is best manifested by the deluge of peer invitations and referrals for SNS sign–ups that are sent out by current SNS participants. Both the push and pull forces work to stretch preexisting social relations in “real” space to “virtual” space. In brief, the size of the installed personal network should be related to inducing participation in Facebook, thus leading to our second hypothesis:

H2: People will be more likely to participate in Facebook if it penetrates their off–line contact circles more widely.

Although Rice, et al. (1990) tested the effect of critical mass adoption on an individual’s perceived outcomes of adopting an electronic messaging system, H1 and H2 in contrast concern the relationship of critical mass participation to the likelihood that one will participate.

Additionally, a person may use more than one SNS because of the need to cover more of the entire SNS population, which is dispersed over separate SNSs. In a recent U.S. survey, Lenhart (2009) discovered that 24 percent of SNS users had multiple profiles to keep better track of their friends, who were spread across multiple sites. Even so, the same bandwagon effect would drive these “multi–homers” to engage more often with the SNS connected to most of their off–line contacts than with any other. Thus, we proposed our third hypothesis:

H3: Multiple SNS users are more likely to use Facebook more than all other SNSs if it penetrates their off–line contact circles more widely.

Usage of an SNS

The degree to which a person uses an SNS should be related to the size of her personal network built over the site. Here, following Lewis, et al. (2008) and Roberts, et al. (2009), we defined a person’s Facebook network size as the number of “alters” with whom “ego” has a specified direct relationship — that is, the number of friends she has on her account. The causality for this association can go either way. First, when holding a larger roster of dyadic contacts on an SNS, a person should find the SNS affiliation more enjoyable and therefore use the site more often. Second, those starting with a keener interest in the SNS will spend more time on the site, building their profiles and presence online, thus naturally expanding their SNS friendships. Moreover, in a more practical sense, a larger number of contacts on the SNS, and thus a wider scope of social interaction, will require more attention to maintain the social network.

Empirical studies examining social influence have explained level of use by individuals through the diffusion of a technology among their contact communities. Specifically, Soe and Markus (1993) showed that a person’s estimate of the penetration of a given communication technology among her communication community (i.e., work group) predicted the level of her own use of the technology. Fulk (1993) and Schmitz and Fulk (1991) found the usage level among a six–person ego–network to predict one’s e–mail usage as well. Thus, we posited this additional hypothesis:

H4: The larger the size of a person’s Facebook friendship network, the more time that person will spend on Facebook.

Other than benefits activated by social interconnectedness, various technology adoption models have accounted for the practical value provided to users by intrinsic technological features, functions, and capabilities. Models such as information or media richness (Daft and Lengel, 1986, 1984; Trevino, et al., 1987), uses and gratifications (e.g., Leung and Wei, 2000), technological utility (Soe and Markus, 1993), organizational information processing (Rice, et al., 1990), and perceived usefulness (Davis, 1989; Davis, et al., 1989), albeit with a focus on specific facets of technological applications, share an analogous ground of reasoning: People are attracted to a technology that promises utilitarian performance, efficacy, and fit at leisure or at a task. These adoption perspectives treat the utilitarian attributes as an antecedent to a user’s receptivity to a technology. Empirical studies have commonly discovered utility/usefulness measures to be highly predictive of an individual’s technology use (e.g., Schmitz and Fulk, 1991; Soe and Markus, 1993; Rice, et al., 1990).

This study further sought to account for perceived usefulness to ensure that it identified the independent effect of the bandwagon variables on Facebook usage. Perceived usefulness has been defined as the subjective evaluation or experience of prospective users as to how using a specific application system will help in achieving their personal goals (Davis, 1989; Davis, et al., 1989), and has been empirically tested in many IT–related contexts (see Park, et al., 2007, for a review). Thus, we proposed the following:

H5: The perceived usefulness of Facebook will be positively associated with the level of Facebook usage.

 

++++++++++

Method

Survey

We administered a Web–based survey about SNS participation and usage in March 2009. Participants were recruited using a combination of cross–sectional and snowball sampling, with undergraduate students majoring in communication at a Singapore university as the basis from which to disseminate the survey. Students on the mailing list were emailed an invitation containing a link to the survey questionnaire online, which they were instructed to forward to extended family members who use the Internet daily, but not beyond, while they were not to participate in the survey themselves. This was designed to approximate the active Internet–user population in Singapore in general rather than targeting a student group. Despite being nonprobability sampling, the snowballing was rolled out from a coherent cross–section and was limited to merely a one–step referral [1].

The sampling was confined to active Internet users, because non–users and inactive users are situated in different environments with regard to adopting an online service. Mixing them all would introduce undesirable sample heterogeneity in terms of individual capabilities or incentives to participate in SNSs. The study’s focus on the bandwagon effect thus made it necessary to control such confounds.

Thus, only daily users of the Internet, aged 18 and above, with Singapore citizenship and permanent residency, were qualified to participate in the survey for the sake of sample homogeneity. Data cleaning produced a final sample of 378 valid respondents, with an ethnic makeup of 85 percent Chinese, 4.3 percent Malay, 2.9 percent Indian, and 7.9 percent other. Table 1 summarizes the demographic statistics, including gender, age, marital status, education, and income. The vast majority (93 percent) of respondents were aged between 16 and 40. We compared the sample’s demographic profile with that of a random Singapore Internet user sample collected in 2007 (Skoric, et al., 2009), and found that the age, education, and income averages among the national sample were in line with the present sample’s descriptive statistics. This demographic similarity implied that our sample was representative. Moreover, the SNS penetration among the sample (91 percent) resonated fairly with a recent national survey by Nielsen, which found that 95 percent of the 15–19 age group and 89 percent of the 20s age group were SNS users (Shafawi, 2010).

 

Table 1: Summary statistics of variables.
VariableEntire sample
(N = 378)
Multiple SNS users
(N = 287)
Facebook participants
(N = 326)
 MSDMSDMSD
1. Age 
10s9%   7% 
20s73%   75% 
30s15%   16% 
40s3%   2% 
50s1%   0% 
2. Gender 
Male52%   56% 
Female48%   44% 
3. Marital status 
Single87%   89% 
Married11%   9% 
Divorced1%   2% 
Widowed1%   0% 
4. Education 
Primary school1%   0% 
O/N/A level25%   26% 
Professional/technical certificate3%   1% 
Diploma14%   14% 
Degree46%   49% 
Graduate degree10%   10% 
Other1%   0% 
5. Income 
< S$2,50059%   60% 
S$2,500–S$4,99927%   29% 
S$5,000–S$7,4994%   3% 
S$7,500–S$9,9994%   3% 
> S$10,0005%   6% 
6. Whether one participates in Facebook
(1 for yes; 0 for no)
0.860.35    
7. Whether one uses Facebook the most among all SNSs  0.920.27  
8. Whether Facebook is the SNS most widely used among one’s off–line friends  0.750.43  
9. Number of off–line friends who use Facebook229.54219.30  318.59225.12
10. Percentage of SNS–using off–line friends who use Facebook56%21%60%18%  
11. Whether one believes Facebook to be the SNS having the most active users worldwide0.680.47    
12. Whether one believes Facebook to be the SNS having the most active users in Singapore0.710.45    
13. Frequency of Facebook logins    4.091.09
Several times a day    42.6% 
About once a day    35.5% 
Once every few days    9.6% 
Once a week    4.9% 
Once every fortnight    4.3% 
14. Hours spent on Facebook in the last week    7.3810.33
15. Number of online Facebook friends    83.7762.06
16. Perceived usefulness of Facebook    3.760.61

 

Measures

The questionnaire addressed all common SNSs equally with respect to the aspects described below. The survey did not in any way single out Facebook from other SNSs, with questions prelisting the SNSs for respondents to report about, although our analysis later focused on the case of Facebook participation. The respondents were thereby not sensitized about the study’s interest in this particular SNS. After the survey, we extracted measures that pertained to predicting participation in and use of Facebook specifically.

Participation. Respondents reported which SNS(s) they were currently using and which one the most often. The questionnaire defined use of an SNS as owning and maintaining a personal profile on it. Respondents who participated in any SNS comprised 91 percent of the sample. Facebook was used by the most respondents (86 percent), followed by Friendster (62 percent), MySpace (9 percent), LinkedIn (8 percent), Bebo (1%), and other (7 percent). “Multi–homing” was fairly common among SNS users in the sample: 24 percent of the SNS users were reported as single–SNS users, 59 percent double–SNS users, and 12 percent triple–SNS users, while the remainder used even more SNSs. Even so, 95 percent of the multiple–SNS participants used Facebook the most compared with the other SNSs.

Usage. We measured SNS usage in two ways. First, for any SNS reported as being used, the respondents were asked to rate how often they had logged into the SNS on a 5–point Likert scale ranging from once every fortnight through several times a day. Second, they reported how many hours they had used that SNS in the past week. Among Facebook users, the averages of the two usage measures were 4.09 points and 7.38 hours, respectively. The corresponding averages for the other SNSs were as follows: 3.30 points and 5.63 hours (Friendster), 2.77 and 2.98 (MySpace), 3.25 and 1.88 (LinkedIn), and 2.04 and 1.31(Bebo). The cross–SNS comparison revealed that individuals tended to use an SNS showing wider penetration more frequently and for longer periods. This size–sensitive usage was consistent with the bandwagon postulation about the value of the different SNSs.

Perceived diffusion of SNSs. The respondents also reported which SNS they believed had the most active users both in Singapore and worldwide. The percentages of all respondents who believed that a particular SNS had the widest popularity were as follows: globally, Facebook 68.3 percent, Friendster 4.7 percent, MySpace 24.4 percent, LinkedIn 1.4 percent, and other 1.1 percent; in Singapore, Facebook 70.6 percent, Friendster 23.8 percent, MySpace 2.9 percent, LinkedIn 0.2 percent, and other 0.7 percent. This breakdown suggests that far more people perceived Facebook to be the most widely used SNS than any other. The perceptions among the sample about the popularity of SNSs in Singapore were accurate in reflecting the ranking of the SNSs in terms of the actual penetration found within the sample, as shown in Table 2. Interestingly, however, how closely the perceived popularity matched the sampled penetration varied across SNSs. The higher the penetration ranking an SNS had, the more accurate was the sample’s perception about the SNS’s popularity in Singapore. This is reflected as a decrease in the B/A ratio in popularity ranking in Table 2. In other words, Facebook, the most popular SNS in reality, was believed to be so by far a greater proportion of its users than a less popular SNS, such as Friendster or even MySpace.

 

Table 2: Penetration rates and perceived popularity of SNSs.
 Ranking(A)
Penetration % of sample
(B)
Perceived as most popular in Singapore by % of sample
B/A
Facebook186%70.6%0.821
Friendster262%23.8%0.384
MySpace39%2.9%0.322
LinkedIn48%0.2%0.025

 

Reach of Facebook among social circles. The survey then asked respondents about the reach of each SNS among their personal social circles, both off–line and online. It inquired about the following specific aspects: (a) whether the SNS was the one most commonly used among off–line friends; (b) the number of off–line friends who used the SNS; (c) the percentage of their entire SNS–using off–line friends who used the SNS; and, (d) the number of online friends owned on the SNS. These reports reflected various aspects of the diffusion of an SNS among the respondents’ off–line contact communities while also measuring both the relative and absolute sizes of their personal networks held through that SNS.

Perceived usefulness. We also constructed a scale for the perceived usefulness of an SNS. From studies examining the reasons or motives for using an SNS (i.e., boyd, 2008; Bumgarner, 2007; Coyle and Vaughn, 2008; Ellison, et al., 2007; Lenhart, 2009; Raacke and Bonds–Raacke, 2008; Sledgianowski and Kulviwat, 2009), we identified a list of 17 questions around the usefulness of SNSs, each of which was measured in the survey on a 5–point Likert scale, ranging from 1 (strongly agree) through 5 (strongly disagree). We then factor–analyzed the 17 items with the principal component procedure and the Kaiser rule. Only those items whose communalities as calculated exceeded 0.7 remained in the subsequent analysis. Five items survived the criterion: (a) “The SNS fulfills my expectations of a social networking site”; (b) “It provides me with a fast and efficient way to communicate with my friends”; (c) “It allows me to keep track of what my friends are doing”; (d) “I feel that there are a lot of activities to do on the SNS”; and, (e) “I believe that the SNS is continually improving itself.” Among Facebook participants, the five–item usefulness scale reported M = 3.76, SD = 0.61, with Cronbach’s α = 0.825 and composite reliability = 0.831. Thus, the internal consistency and reliability indexes supported the quality of the measures.

Table 1 summarizes the basic statistics of the variables pertaining to the entire sample, the multiple SNS user subset, and the Facebook user subset, separately, while Table 3 reports the zero–order correlations between the variables.

 

Table 3: Zero–order correlations between variables used in regressions.
 12345678910
1. Whether one adopts Facebook 
2. Whether one uses Facebook the most among all SNSs 
3. Whether Facebook is the SNS most widely used among one’s off–line friends .26 
4. Number of off–line friends who use Facebook.28.17 
5. Percentage of SNS–using off–line friends who use Facebook.51.24.24.15 
6. Whether one believes Facebook to be the SNS having the most active users worldwide.12.18.57.10.21 
7. Whether one believes Facebook to be the SNS having the most active users in Singapore.07.18.63.00.12.59 
8. Frequency of Facebook logins 
9. Hours spent on Facebook in the last week .33 
10. Number of online Facebook friends .32.13 
11. Perceived usefulness of Facebook .17.14.10

 

 

++++++++++

Regression analysis

We then ran a series of regressions to test the hypotheses concerning the bandwagon effect on participating in and using Facebook. These regressions were meant to inspect different responses involving the aforementioned subsamples.

Participation

We first tested H1, which postulated that people’s perceptions about the propagation of Facebook would predict their participation decision. We regressed a dichotomous variable of whether one used Facebook currently (1 for participant, 0 for nonparticipant) over the perceived popularity of Facebook in the logit model while controlling for the demographic aspects. We entered the perceptual measures for worldwide and Singapore penetration into separate regressions because of their high collinearity (r = .59, p = .00). As Table 4 reports, both the perceived global and Singapore popularity had a statistically significant and positive association with participation. In other words, people who were convinced that Facebook was the most widely deployed SNS were more likely to participate in it than those who were not convinced, thus confirming H1. These two regressions captured a critical fraction of the variations in participation (pseudo R2 = .44 and .54, p < .01).

 

Table 4: Logit regression of Facebook participation over perceived Facebook popularity.
Note: *p < .05; **p < .01; N = 378.
Independent variableColumn 1Column 2
BExp(B)tBExp(B)t
Whether perceived the most popular worldwide1.91**6.764.07 
Whether perceived the most popular in Singapore 3.11**22.475.06
 
Age   10s (base) 
      20s1.173.221.891.283.591.87
      30s2.59*13.332.322.148.521.83
      40s2.249.391.55-0.830.440.54
      50s-0.190.830.12-1.210.300.63
Gender   Male (base) 
      Female0.782.171.710.912.481.74
Marital status   Single 
      Married-1.95*0.142.41-0.460.630.49
      Divorced0.021.020.000.041.030.00
      Widowed-0.080.920.00-0.060.940.00
Education   Primary (base) 
      O/N/A level-1.700.181.48-0.750.470.67
      Technical certificate0.191.200.280.511.670.65
      Diploma1.103.011.701.153.141.69
      Undergraduate degree-0.120.890.120.591.810.57
      Graduate degree-0.030.970.01-0.100.900.12
Income   < S$2,500 (base) 
      S$2,500–S$4,9990.191.210.310.041.040.05
      S$5,000–S$7,499-1.590.201.49-1.440.241.28
      S$7,500–S$9,999-1.200.301.23-1.610.201.49
      > S$10,000-0.020.980.02-0.470.620.34
Constant-0.700.501.10-1.260.281.66
pseudo R20.440.54

 

With the evidence above supporting the effect of general diffusion, we next tested the effect of personal–community diffusion on participation. We regressed participation over the number of off–line contacts who used Facebook and the same demographic controls. Likewise, the model assumed a significant proportion of the participation variations (pseudo R2 = .51, p < .01), as reported in Column 1 of Table 5. The number of contacts was statistically significant in predicting participation (B = 0.01, p < .01). Based on this variable’s coefficient estimate, having one more Facebook–using contact was associated with about a one percent increase in the likelihood of participation, with demographics kept constant.

 

Table 5: Logit regression of Facebook participation over Facebook’s penetration among social contacts.
Note: *p < .05; **p < .01; N = 378.
Independent variableColumn 1Column 2
BExp(B)tBExp(B)t
Number of off–line friends who use Facebook0.01**1.013.72 
Percentage of SNS–using off–line friends who use Facebook 0.11**1.125.01
 
Age   10s (base) 
      20s1.96*7.092.501.836.251.78
      30s3.93**51.063.154.36*78.082.25
      40s4.26**70.792.612.229.211.08
      50s1.143.120.6919.081.93e80.00
Gender   Male (base) 
      Female1.17*3.212.220.862.361.28
Marital status   Single 
      Married-1.330.261.54-1.740.181.38
      Divorced0.481.620.610.852.350.26
      Widowed-0.500.600.37-0.480.620.55
Education   Primary (base) 
      O/N/A level-2.60*0.072.05-2.460.091.25
      Technical certificate0.361.430.47-0.140.870.15
      Diploma0.792.201.080.281.320.33
      Undergraduate degree-0.730.430.71-0.400.670.30
      Graduate degree-0.290.740.82-0.630.530.55
Income   < S$2,500 (base) 
      S$2,500–S$4,9990.001.000.010.702.010.83
      S$5,000–S$7,499-1.740.181.590.451.560.31
      S$7,500–S$9,999-2.53*0.082.36-2.53*0.082.00
      > S$10,000-1.990.141.260.071.080.03
Constant-1.95*0.142.23-5.04**0.013.30
pseudo R20.510.65

 

For a more robust assessment of H2, we used another gauge of personal–community diffusion to explain participation. Column 2 of Table 5 presents a participation regression that specified the relative measure of Facebook penetration — the percentage of a respondent’s off–line contacts who used SNSs — in lieu of the absolute measure, that is, the raw contact count. This alternative measure continued to show a consistent and statistically significant relationship with participation (pseudo R2 = 0.65, p < .01). Its coefficient meant that a one percent increase in the share of SNS–using off–line contacts who were Facebook users corresponded to a 12 percent increase in the likelihood of participation. In sum, the findings based on both absolute and relative diffusion measures supported H2. Next, we examined the bandwagon effect on Facebook usage.

Usage

We approached the bandwagon effect on usage separately for two distinct, though partly overlapping, groups — multiple SNS users and Facebook users, for whom we designed different but nuanced usage measures.

Among multiple SNS users. For multiple SNS users, we aimed to explain whether such a user engaged in Facebook more often than any other SNS. Hence, we ran logit regressions of the binary variable among the subsample of the 287 such respondents. The variable was regressed over the variables of whether Facebook was the SNS most widely adopted among one’s off–line contacts, and what percentage of SNS–using off–line contacts used Facebook. Again, we adjusted for the demographics in the regressions. Table 6 presents the regressions, which included the two key regressors alternatively in Columns 1 and 2. Both were statistically significant. Overall, the results suggest that multiple SNS users tended to use Facebook more often than other SNSs when their real–life social circles received deeper Facebook penetration, hence lending support to H3. Column 2 implies that a one percent increase in the Facebook share of SNS–using off–line contacts increased by 7.6 percent the likelihood that Facebook was the dominant site for multiple SNS users.

 

Table 6: Logit regression of whether a multiple–SNS participant uses Facebook more than other SNSs over Facebook’s penetration among friends.
Note: *p < .05; **p < .01; N = 287.
Independent variableColumn 1Column 2
BExp(B)tBExp(B)t
Whether Facebook is the SNS most widely used among one’s off–line friends1.24**3.463.39 
Percentage of SNS–using off–line friends who use Facebook 0.94*2.602.39
 
Age   10s (base) 
      20s3.40**30.042.992.7716.011.90
      30s3.97*53.142.282.6414.021.29
      40s1.504.470.590.171.190.07
      50s5.08161.440.002.5813.150.00
Gender   Male (base) 
      Female1.69*5.432.400.892.420.86
Marital status   Single 
      Married-1.600.201.43-1.670.191.07
Education   Primary (base) 
      O/N/A level0.802.230.760.922.530.38
      Technical certificate0.491.651.280.621.871.00
      Diploma-1.830.161.76-1.110.330.83
      Undergraduate degree-1.330.270.800.611.850.26
Income   < S$2,500 (base) 
      S$2,500–S$4,999-0.900.411.22 
      S$5,000–S$7,499-1.630.201.22 
      S$7,500–S$9,999-1.270.280.68 
      > S$10,000-1.190.301.60 
Constant1.133.110.72-3.390.031.36
pseudo R20.340.41

 

Among Facebook users. We built another usage regression for Facebook users themselves. We regressed the number of hours spent on Facebook in ordinary least squares (OLS), over the number of online Facebook friends and the perceived usefulness of Facebook simultaneously, in addition to the same demographic controls. The regression entered the number of friends in logarithmic form. The log transformation was technically advantageous, since the Facebook friend count was exceedingly dispersed and the outliers caused undue influence in the regression. Logarithmic rescaling was able to tame this outlier problem (e.g., Lee, 1988). The coefficient of variation for the Facebook friend count, when transformed, dropped from 0.72 to 0.19, whereas that for the number of hours, when transformed, decreased from 1.64 to 0.59.

Table 7 reports that both online friend counts and perceived usefulness had a statistically significant relationship with Facebook usage. According to the coefficient estimate of the friend count in log form (B = 0.23), if a user’s Facebook friends were doubled, that user’s Facebook use time would increase by 17 percent. Also, whenever the measure of perceived usefulness of Facebook went up by one Likert point, use time increased by 39 percent (B = 0.33).

 

Table 7: Regressions of Facebook usage.
Note: *p < .05; **p < .01; N = 326.
a The ordered logit estimation has χ2(17) = 56.30, p = .000. The thresholds for the levels of the dependent variable are 3.00 (t = 2.16), 3.94 (t = 2.84), 4.98 (t = 3.56), and 7.00 (t = 4.87), respectively.
Independent variableOrdinary least squares regressionOrdered logit regressiona
Log (hours on Facebook)Frequency of Facebook logins
BTBt
Log (number of online Facebook friends)0.23**3.030.77**4.50
Perceived usefulness0.33**3.070.68**2.65
 
Age   10s (base) 
      20s-0.341.54-0.490.88
      30s-0.501.56-1.63*2.05
      40s-0.781.46-0.170.13
      50s-1.301.49-0.760.36
Gender   Male (base) 
      Female-0.060.50-0.070.28
Marital status   Single 
      Married0.220.970.460.83
      Divorced0.701.572.631.94
Education   Primary (base) 
      O/N/A level0.681.291.931.35
      Technical certificate0.271.500.521.18
      Diploma-0.261.800.230.68
      Undergraduate degree-0.371.46-0.731.21
Income   < S$2,500 (base) 
      S$2,500–S$4,999-0.040.290.170.52
      S$5,000–S$7,4990.471.421.72*1.98
      S$7,500–S$9,999-0.080.241.431.66
      > S$10,000-0.351.29-0.560.86
Constant-0.490.82 
pseudo R2 or Cox & Snell R20.240.21

 

The frequency of Facebook logins, as a complementary measure of usage, was also related to the same factors. We used an ordered logit regression to fit login frequency, an ordinal measure. Table 7 reports the regression. Again, the Facebook friend count and perceived usefulness remained statistically significant. These regressions thus buttressed H4 and H5.

The results of the control variables were also mostly consistent and stable across the regressions. The regressions showed that gender, marital status, and age were associated with certain differences in the tendency to participate in Facebook. Female or single individuals were more likely to be Facebook participants than were others, whereas teens were less likely to participate than older age groups. These differences, though not always statistically significant, were in line with the usual expectations in regard to the profiles of SNS users.

 

++++++++++

Discussion and conclusion

This study empirically observed the bandwagon effect on individuals’ participation in and use of a popular SNS. We established evidence for the effect by the linkage between use of Facebook and its diffusion at various levels of the community. Whether and how much one used Facebook was unequivocally coupled with its diffusion within the global, local, and communal contexts. Moreover, the bandwagon effect was evinced amid various subsets of Internet users — a general group, multiple SNS users, and Facebook users — and was found significant both in statistical and real terms. All in all, the penetration of Facebook both at large and among a person’s off–line peers was critically responsible for that person’s choice to use it. The series of analyses persistently suggested that individuals tend to jump on an SNS bandwagon that has already been set in motion among the population and around their social circles. The bandwagon process herds people to the same SNS, through which they form online social networks.

In addition to the inferences drawn from the regression analysis, our survey also asked respondents to name explicitly the SNSs that they used most often and that most of their friends belonged to. As the data directly revealed, the choices of 93.5 percent of the respondents and those of their friends coincided.

Although both the perceived global and local (Singapore) diffusion of Facebook predicted a person’s Facebook participation, the respondents were more sensitive to its penetration within than outside Singapore. This is suggested by the larger effect magnitude and greater explanatory power of the local diffusion measure as opposed to that of the worldwide dispersion, shown in Table 4. Perceptions about local and overseas Facebook propagation were expected to be interrelated; in fact, the two measures showed r = .59 (p = .000). But because social or communicative behaviors should be driven more by the society in which people live everyday than by offshore populations, which are portrayed mostly through mass media, users were most sensitive precisely to how Facebook was deployed in their own surroundings.

This finding strikes a chord with the common understanding that online social networking is commonly engaged in to maintain relationships or friendships initiated off–line, rather than to source new ones with strangers from cyberspace. Research has found that SNSs are principally used to support preexisting connections, springing from real–world social commonalities such as attending the same school (Ellison, et al., 2007; Mayer and Puller, 2008). The current results therefore not only revalidated the interface between online SNS engagement and extant social relations, but also pointed to an antecedent to the online–off–line interface — that is, the choices and behaviors of SNS users are responsive to the preferences established among their off–line social groups.

The findings also attest to the perceptual impact of the installed user base on the participation decision. Although MySpace had long retained the largest number of global unique visitors until Facebook recently overtook it (comScore, 2009), the belief that Facebook had the most active users in the world or in Singapore appeared to be ingrained in most of the respondents (68 percent and 71 percent, respectively) and inspired them to set up their Facebook profiles, as shown so far. The same perceptual impact was uncovered from the multiple SNS user group as well. Among this segment, those who used Facebook most and those who did not tended to overlap with those who believed Facebook was the most popular SNS and those who did not, respectively. This was confirmed by the 2X2 cross–tabulation for the two binary variables, which produced χ2(1, n = 287) = 11.51, Cramér’s V = .145, p < 0.001 on the global account, and χ2(1, n = 287) = 10.97, Cramér’s V = .180, p < 0.001 on the Singapore account. This pattern implied that people are drawn to the SNS they believe has the largest subscriber base for maximal connectivity regardless of reality.

We cross–tabulated the binary variable of perceived global popularity with that of whether Facebook was the SNS most widely used among a person’s off–line friends. The results showed that the peer Facebook adoption under a person’s direct observation predicted her or his impression as to whether it was the most popular SNS in Singapore, χ2(1, n = 287) = 132.87, Cramér’s V = .628, p < 0.000. From this cross–tabulation, only 18 percent of the respondents who felt that Facebook was the SNS most used by their peers did not see it as the one most employed in Singapore. Meanwhile, only 12 percent of the respondents who did not consider Facebook to be the SNS most used by their peers agreed that it was the most employed one in Singapore.

We also found that among Facebook users, enlisting more online friends was correlated with higher usage. This was so whether usage was defined as login frequency or use time. In fact, both usage measures were correlated (r = 0.33, p < 0.00).

The relationship found between Facebook usage and the size of one’s online friend network was also consistent with general intentions about social interaction or networking amid SNS users. First, owning a longer Facebook friend inventory may signal an enthusiasm in locating or making friends. Not only does it take time to knit a social web to begin with, but a large friend list per se also requires the user to spend more time and effort to attend to it. Frequent updates and messages from friends keep a user engaged on Facebook.

The perceived usability of Facebook was also positively linked to usage. This result is not surprising, since people will use a service more intensively if they experience it as conducive to serving their needs or achieving their purposes.

Last, we needed to check whether the two usage factors were correlated so that their respective relationships with usage, shown in Table 7, would be confounded. The results showed that the correlation between perceived usefulness and Facebook friend count was rather low and not statistically significant at the five percent level, r = 0.10. With limited collinearity, the statistical significance of each factor signified its independent contribution to explaining Facebook usage. More plainly, it is not impossible, in reality, to use Facebook heavily when a person has few Facebook friends but still feels it is useful, or when a person has many Facebook friends but does not feel it is very useful. Either of these two dimensions could contribute to Facebook usage.

The investigation thus suggests that people choose and use SNSs in compliance with proximity to their off–line groups and perception of use prevalence rather than by random or amorphous selection. This off–line–online group adherence requires no organizational or formal coordination by users; instead, it is grown “organically” among sequential adopters maximizing the benefit received from such a service. When people’s SNS affiliations occur or migrate collectively, this process adds to the domination and concentration of a large SNS among smaller ones. Truly, the playing field of SNSs has tipped toward the leading player, Facebook.

This study, as one of the first to explore the bandwagon effect on SNS participation, is limited in some ways. First, we used a non–probability sample of voluntary participants. Although comparing sample statistics did not suggest our sample to be idiosyncratic with other national samples, further investigations utilizing probability sampling could verify the present results.

Finally, the survey study detected strong correlations between bandwagon factors and individual use using cross–sectional variations. But longitudinal observation of the evolution of a user’s personal networks within respective SNSs and the user’s choices and responses across SNSs would offer more a rigorous examination of the push and pull aspects of the bandwagon effect. It could also add to the robustness in testing the causal relationships. Future studies should thus consider deriving longitudinal data about SNS use both onsite and in real time. End of article

 

About the authors

W. Wayne Fu is an Associate Professor in the Wee Kim Wee School of Communication and Information at Nanyang Technological University, Singapore.
E–mail: TWJFU [at] ntu [dot] edu [dot] sg

Jaelen Teo is a business risk consultant, based in Singapore, specialising in integrity, political and security risk across South–East Asia.
E–mail: jaelen [dot] teo [at] gmail [dot] com

Seraphina Seng is a business development and media strategist for a global provider of advanced technological solutions, based in Singapore.
E–mail: sqianru [at] gmail [dot] com

 

Note

1. Internet users exhibit demographic characteristics and lifestyles that diverge from those that traditional probability sampling methods, such as random digit dialing telephone interviews (Chang and Krosnick, 2009). Internet–based surveys also tend to invite a lower level of random measurement error, survey satisficing, and social desirability than telephone–based surveys (Gosling, et al., 2004).

 

References

D. Birke and G.M.P. Swann, 2006. “Network effects and the choice of mobile phone operator,” Journal of Evolutionary Economics, volume 16, numbers 1–2, pp. 65–84.

J. Boase, 2010. “The consequences of personal networks for Internet use in rural areas,” American Behavioral Scientist, volume 53, number 9, pp. 1,257–1,267.

d.m. boyd, 2008. “Why youth ♥ social network sites: The role of networked publics in teenage social life,” In: D. Buckingham (editor). Youth, identity, and digital media. Cambridge, Mass.: MIT Press, pp. 119–142.

B.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://firstmonday.org/article/viewArticle/2026/1897, accessed 26 September 2008.

L. Chang and J.A. Krosnick, 2009. “National surveys via Rdd telephone interviewing versus the Internet: Comparing sample representativeness and response quality,” Public Opinion Quarterly, volume 73, number 4, pp. 641–678.http://dx.doi.org/10.1093/poq/nfp075

S.Y. Chun and M. Hahn, 2007. “Network externality and future usage of Internet services,” Internet Research, volume 17, number 2, pp. 156–168.http://dx.doi.org/10.1108/10662240710737013

comScore, Inc., 2009. “comScore says Facebook has surpassed MySpace for U.S. users,” at http://www.comscore.com/press/release.asp?press=2396, accessed 17 June 2009.

N.S. Contractor and E.M. Eisenberg, 1990. “Communication networks and new media in organizations,” In J. Fulk and C. Steinfield (editors). Organizations and communication technology. Newbury Park, Calif.: Sage, pp. 143–172.

N. Corrocher and L. Zirulia, 2009. “Me and you and everyone we know: An empirical analysis of local network effects in mobile communications,” Telecommunications Policy, volume 33, numbers 1–2, pp. 68–79.

C.L. Coyle and H. Vaughn, 2008. “Social networking: Communication revolution or evolution?” Bell Labs Technical Journal, volume 13, number 2, pp. 13–17.http://dx.doi.org/10.1002/bltj.20298

R.L. Daft and R.H. Lengel, 1986. “Organizational information requirements, media richness, and structural design,” Management Science, volume 32, number 5, pp. 554–571.http://dx.doi.org/10.1287/mnsc.32.5.554

R.L. Daft and R.H. Lengel, 1984. “Information richness: A new approach to managerial behavior and organizational design,” In L.L. Cummings and B.M. Staw (editors). Research in organizational behavior, volume 6, pp. 191–233.

F.D. Davis, 1989. “Perceived usefulness, perceived ease of use, and user acceptance,” MIS Quarterly, volume 13, number 3, pp. 319–340.http://dx.doi.org/10.2307/249008

F.D. Davis, R.P. Bagozzi, and P.R. Warshaw, 1989. “User acceptance of computer technology: A comparison of two theoretical models,” Management Science, volume 35, number 8, pp. 982–1,003.

D. Easley and J. Kleinberg, 2010. Networks, crowds, and markets: Reasoning about a highly connected world. New York: Cambridge University Press.

N. Economides and C. Himmelberg, 1995. “Critical mass and network evolution in telecommunications,” In: G.W. Brock (editor). Toward a competitive telecommunications industry: Selected papers from the 1994 Telecommunications Policy Research Conference. Mahwah, N.J.: Lawrence Erlbaum Associates, pp. 47–66.

N.B. Ellison, C. Steinfield, and C. 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, pp. 1,143–1,168, and at http://jcmc.indiana.edu/vol12/issue4/ellison.html, accessed 12 April 2012.

R.H. Frank and P.J. Cook, 1995. The winner–take–all society. New York: Free Press.

W.W. Fu and C. Sim, 2011. “Aggregate bandwagon effect on online videos’ viewership: Value uncertainty, popularity cues, and heuristics,” Journal of the American Society for Information Science and Technology, volume 62, number 12, pp. 2,382–2,395.

J. Fulk, 1993. “Social construction of communication technology,” Academy of Management Journal, volume 36, number 5, pp. 921–950.http://dx.doi.org/10.2307/256641

J. Fulk, J. Schmitz, and C. Steinfield, 1990. “A social influence model of technology use,” In: J. Fulk and C. Steinfield (editors). Organizations and communication technology. Newbury Park, Calif.: Sage, pp. 117–140.

S.D. Gosling, S. Vazire, S. Srivastava, and O.P. John, 2004. “Should we trust Web–based studies? A comparative analysis of six preconceptions about Internet questionnaires,” American Psychologist, volume 59, number 2, pp. 93–104.http://dx.doi.org/10.1037/0003-066X.59.2.93

E. Hargittai, 2007. “Whose space? Differences among users and non–users of social network sites,” Journal of Computer–Mediated Communication, volume 13, number 1, at http://jcmc.indiana.edu/vol13/issue1/hargittai.html, accessed 26 September 2008.

J. Henkel and J. Block, 2007. “Peer influence in network markets: An empirical and theoretical analysis,” SSRN Working Paper (1 January), at http://ssrn.com/abstract=956564, accessed 15 November 2008.

M. Hesse, 2009. “Worldwide ebb for Facebook: Now that everybody and his mother is on Facebook, is it time to look for the next big thing in social networking?” Washington Post (19 October), p. C1.

S.R. Hiltz, 1984. Online communities: A case study of the office of the future. Norwood, N.J. : Ablex.

V. Ilie, C. Van Slyke, G. Green, and H. Lou, 2005. “Gender differences in perceptions and use of communication technologies: A diffusion of innovation approach,” Information Resources Management Journal, volume 18, number 3, pp. 13–31.http://dx.doi.org/10.4018/irmj.2005070102

M.L. Katz and C. Shapiro, 1994. “Systems competition and network effects,” Journal of Economic Perspectives, volume 8, number 2, pp. 93–115.http://dx.doi.org/10.1257/jep.8.2.93

M.L. Katz and C. Shapiro, 1986. “Technology adoption in the presence of network externalities,” Journal of Political Economy, volume 94, number 4, pp. 822–841.http://dx.doi.org/10.1086/261409

H.S. Kim and N. Kwon, 2003. “The advantage of network size in acquiring new subscribers: A conditional logit analysis of the Korean mobile telephony market,” Information Economics and Policy, volume 15, number 1, pp. 17–33.http://dx.doi.org/10.1016/S0167-6245(02)00070-7

C. Lampe, N. Ellison, and C. Steinfield, 2006. “A Face(book) in the crowd: Social searching vs. social browsing,” Proceedings of the 2006 20th Anniversary Conference on Computer Supported Cooperative Work. New York: ACM Press, pp. 167–170.

A.H. Lee, 1988. “Assessing partial influence in generalized linear models,” Biometrics, volume 44, number 1, pp. 71–77.http://dx.doi.org/10.2307/2531896

A. Lenhart, 2009. “Adults and social network websites,“ Pew Internet & American Life Project (14 January), at http://www.pewinternet.org/Reports/2009/Adults-and-Social-Network-Websites.aspx, accessed 11 February 2009.

A. Lenhart and M. Madden, 2007. “Teens, privacy, and online social networks,” Pew Internet & American Life Project (18 April), at http://www.pewinternet.org/Reports/2007/Teens-Privacy-and-Online-Social-Networks.aspx, accessed 30 July 2007.

L. Leung and R. Wei, 2000. “More than just talk on the move: Uses and gratifications of the cellular phone,” Journalism and Mass Communication Quarterly, volume 77, number 2, pp. 308–320.http://dx.doi.org/10.1177/107769900007700206

K. Lewis, J. Kaufman, M. Gonzalez, A. Wimmer, and N. Christakis, 2008. “Tastes, ties, and time: A new social network dataset using Facebook.com,” Social Networks, volume 30, number 4, pp. 330–342.http://dx.doi.org/10.1016/j.socnet.2008.07.002

C.A. Lin, 2003. “An interactive communication technology adoption model,” Communication Theory, volume 13, number 4, pp. 345–365.http://dx.doi.org/10.1111/j.1468-2885.2003.tb00296.x

H. Lou, W. Luo, and D. Strong, 2000. “Perceived critical mass effect on groupware acceptance,” European Journal of Information Systems, volume 9, number 2, pp. 91–103.http://dx.doi.org/10.1057/palgrave.ejis.3000358

M.L. Markus, 1987. “Toward a ‘critical mass’ theory of interactive media: Universal access, interdependence and diffusion,” Communication Research, volume 14, number 5, pp. 491–511.http://dx.doi.org/10.1177/009365087014005003

A. Mayer and S.L. Puller, 2008. “The old boy (and girl) network: Social network formation on university campuses,” Journal of Public Economics, volume 92, numbers 1–2, pp. 329–347.

P.E. Oliver and G. Marwell, 1988. “The paradox of group size in collective action: A theory of the critical mass. II,” American Sociological Review, volume 53, number 1, pp. 1–8.http://dx.doi.org/10.2307/2095728

P.E. Oliver, G. Marwell, and R. Teixeira, 1985. “A theory of the critical mass. I. Interdependence, group heterogeneity, and the production of collective action,” American Journal of Sociology, volume 91, number 3, pp. 522–556.http://dx.doi.org/10.1086/228313

B.M. Owen and S.S. Wildman, 1992. Video economics. Cambridge, Mass.: Harvard University Press.

N. Park, K.M. Lee, and P.H. Cheong, 2007. “University instructors’ acceptance of electronic courseware: An application of the technology acceptance model,” Journal of Computer–Mediated Communication, volume 13, number 1, at http://jcmc.indiana.edu/vol13/issue1/park.html, accessed 15 September 2008.

J. Raacke and J. Bonds–Raacke, 2008. “MySpace and Facebook: Applying the uses and gratifications theory to exploring friend–networking sites,” CyberPsychology & Behavior, volume 11, number 2, pp. 169–174.http://dx.doi.org/10.1089/cpb.2007.0056

R.E. Rice, 1982. “Communication networking in computer conferencing systems: A longitudinal study of group roles and system structure,” In: M. Burgoon (editor). Communication Yearbook, volume 6, pp. 925–944.

R.E. Rice and C. Aydin, 1991. “Attitudes toward new organizational technology: Network proximity as a mechanism for social information processing,” Administrative Science Quarterly, volume 36, number 2, pp. 219–244.http://dx.doi.org/10.2307/2393354

R.E. Rice and D.E. Shook, 1988. “Access to, usage of and outcomes from electronic messaging systems,” ACM Transactions on Office Information Systems, volume 6, number 3, pp. 255–276.http://dx.doi.org/10.1145/45945.214325

R.E. Rice, A.E. Grant, J. Schmitz, and J. Torobin, 1990. “Individual and network influences on the adoption and perceived outcomes of electronic messaging,” Social Networks, volume 12, number 1, pp. 27–55.http://dx.doi.org/10.1016/0378-8733(90)90021-Z

S.G.B. Roberts, R.I.M. Dunbar, T.V. Pollet, and T. Kuppens, 2009. “Exploring variation in active network size: Constraints and ego characteristics,” Social Networks, volume 31, number 2, pp. 138–146.http://dx.doi.org/10.1016/j.socnet.2008.12.002

E.M. Rogers, 2003. Diffusion of innovations. Fifth edition. New York: Free Press.

E.M. Rogers and D.L. Kincaid, 1981. Communication networks: Toward a new paradigm for research. New York: Free Press.

J.H. Rohlfs, 2001. Bandwagon effects in high–technology industries. Cambridge, Mass.: MIT Press.

J.H. Rohlfs, 1974. “A theory of interdependent demand for a communications service,” Bell Journal of Economics and Management Science, volume 5, number 1, pp. 16–37.http://dx.doi.org/10.2307/3003090

J. Schmitz and J. Fulk, 1991. “Organizational colleagues, media richness, and electronic mail: A test of the social influence model of technology use,” Communication Research, volume 18, number 4, pp. 487–523.http://dx.doi.org/10.1177/009365091018004003

Mustafa Shafawi, 2010. “Social media part of life in Singapore, especially among the young: Nielsen,” Channel News Asia (2 February), at http://www.channelnewsasia.com/stories/singaporelocalnews/view/1034867/1/.html, accessed 12 April 2012.

C. Shapiro and H.R. Varian, 1999. Information rules: A strategic guide to the network economy. Boston: Harvard Business School Press.

M.M. Skoric, D. Ying, and Y. 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

D. Sledgianowski and S. Kulviwat, 2009. “Using social network sites: The effects of playfulness, critical mass and trust in a hedonic context,” Journal of Computer Information Systems, volume 49, number 4, pp. 74–83.

L.L. Soe and M.L. Markus, 1993. “Technological or social utility? Unraveling explanations of email, vmail, and fax use,” Information Society, volume 9, number 3, pp. 213–236.http://dx.doi.org/10.1080/01972243.1993.9960142

L. Squire, 1973. “Some aspects of optimal pricing for telecommunications,” Bell Journal of Economics and Management Science, volume 4, number 2, pp. 515–525.http://dx.doi.org/10.2307/3003051

L.K. Trevino, R.H. Lengel, and R.L. Daft, 1987. “Media symbolism, media richness, and media choice in organizations: A symbolic interactionist perspective,” Communication Research, volume 14, number 5, pp. 553–574.http://dx.doi.org/10.1177/009365087014005006

C. Van Slyke, V. Ilie, H. Lou, and T. Stafford, 2007. “Perceived critical mass and the adoption of a communication technology,” European Journal of Information Systems, volume 16, number 3, pp. 270–283.http://dx.doi.org/10.1057/palgrave.ejis.3000680

J.B. Walther, B. Van Der Heide, S.–Y. Kim, D. Westerman, and S.T. Tong, 2008. “The role of friends’ appearance and behavior on evaluations of individuals on Facebook: Are we known by the company we keep?” Human Communication Research, volume 34, number 1, pp. 28–49.http://dx.doi.org/10.1111/j.1468-2958.2007.00312.x

C.–C. Wang, Y.H. Hsu, and W. Fang, 2005. “Acceptance of technology with network externalities: An empirical study of Internet instant messaging services,” Journal of Information Technology Theory and Application, volume 6, number 4, pp. 15–32, and at http://aisel.aisnet.org/jitta/vol6/iss4/4/, accessed 12 April 2012.

 


Editorial history

Received 27 February 2012; accepted 10 April 2012.


Copyright © 2012, First Monday.
Copyright © 2012, W. Wayne Fu, Jaelen Teo, and Seraphina Seng. All rights reserved.

The bandwagon effect on participation in and use of a social networking site
by W. Wayne Fu, Jaelen Teo, and Seraphina Seng
First Monday, Volume 17, Number 5 - 7 May 2012
https://www.firstmonday.org/ojs/index.php/fm/article/view/3971/3207
doi:10.5210/fm.v17i5.3971





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

© First Monday, 1995-2019. ISSN 1396-0466.