This study addresses online political engagement which becomes an important part of current digital culture. Previous research suggests that network diversity is associated with higher levels of political engagement due to enhanced political interest. However, network diversity may also constrain political engagement because politically heterogeneous networks are comprised of others with disparate attitudes and beliefs, which may inhibit communication about politics. The results from an online survey (N=255) suggest that network diversity affords both benefits and burdens to users in terms of online political engagement. Specifically, political diversity is positively correlated with political interest, and political interest predicts political engagement on social networking sites (SNSs). However, political diversity in online networks is also associated with increased self–censoring behaviors. Further, relational diversity is negatively correlated with self–censoring, and this relationship is most pronounced for people with stronger self-disclosure tendencies.
Political engagement on SNSs
Discussion and conclusion
Social networking sites (SNSs), such as Facebook and Twitter, facilitate a wide range of interpersonal communication behaviors. Not surprisingly, they are also platforms for political engagement. A recent Pew report shows that 39 percent of American adults have used SNSs for political activities (Rainie, et al., 2012). Typical political activities on SNSs include commenting on civic and political issues, responding to political posts, calling for participation in political activities such as voting, following candidates and joining political groups formed on SNSs.
One prominent feature of SNSs is increasingly diverse social networks maintained online (Joinson, et al., 2011). Online friends on SNSs may be from various age groups and different ethnic, religious, political and socioeconomic backgrounds. Postings on SNSs resemble broadcasting to various social spheres that have different norms, values, beliefs and expectations. Some scholars have addressed the implications of network diversity in online political engagement and democratic participation (Mutz, 2006). One argument states that diverse networks provide access to diverse sources of information and alternative perspectives, which in turn stimulates political interest and increases political efficacy (Ikeda and Boase, 2010). As political interest and efficacy are positively correlated with actual political participation (Best and Krueger, 2005), diverse networks facilitate political participation.
A contrasting argument states that network diversity is a proxy for the presence of different and even conflicting political views and values (Joinson, et al., 2011). Here, the public display of personal political activity may result in interpersonal conflict and social disapproval due to likely disagreements (Quintelier, et al., 2011). This perspective associates political activities on SNSs with self–presentation. SNS users strategically manage their image online (Rui and Stefanone, 2013). Therefore, users with diverse audiences may feel pressure to refrain from publicly sharing information about their political activities to avoid negative consequences.
The stimulating and restraining effects of network diversity on political engagement mean that under certain circumstances network diversity may afford both benefits and burdens to SNS users regarding democratic participation. Thus, the purpose of this study is to examine the stimulating and restraining effects of network diversity on online political engagement. We suggest that SNSs allow two broad categories of political engagement behavior: public and protective political engagement. Public political engagement is operationalized as unconstrained communication behaviors stating individual political attitudes and activities, while protective political engagement behaviors are aimed at avoiding conflicts and social disapproval (Arkin, 1981).
Our argument is structured along two dimensions. First, we argue that network diversity has a systematic relationship with general political involvement. We operationalize political involvement in terms of political interest and political efficacy. General political involvement should be associated with public displays of political activities on SNSs. Next, we suggest that network diversity is related to strategic, protective political engagement online operationalized as self–censoring and restricting access to shared information It is clear that human behavior is shaped by a range of state– and trait–level variables. Drawing from literature on self–disclosure and self–presentation (Child and Agyeman–Budu, 2010; Hollenbaugh, 2011), we expect to discover how self–monitoring and disclosiveness affect protective political engagement online along with network diversity.
The paper is organized as follows. First, political engagement on SNSs and its implication for democracy is discussed. We then present arguments for and against network diversity in the context of online political engagement. This literature review is used as the foundation for the hypotheses presented herein.
Political engagement on SNSs
The successful application of social media during the 2008 U.S. Presidential campaign highlighted the importance of social media platforms for deliberation and political mobilization (Larsson and Moe, 2012). In authoritarian states, SNSs facilitate expression of alternative viewpoints and mobilization in social movements as demonstrated by the Arab Spring and Iranian election protests (Hamdy and Gomaa, 2012; Larsson and Moe, 2012). Given the popularity of political engagement on SNSs, scholars are interested in the association between online political engagement and off–line civic participation. For example, Valenzuela, et al. (2009) and Vitak, et al. (2011) found that political use of SNSs predicts students’ off–line political participation. The value of political SNSs use includes acquiring political knowledge, stirring political interest, and increasing political self–efficacy (Vitak, et al., 2011).
While the association between online and off–line engagement is worthy of investigation, it is equally important to understand antecedents of political engagement online. Internet use increases the chance of inadvertent encounters with dissimilar others (Brundidge, 2010), because ultimately SNS use weakens the boundaries separating different social spheres (Kim, 2011). Online social networks are characterized by the co–presence of geographically disperse people from different social spheres (Joinson, et al., 2011). Across different social spheres, one will find demographic differences, political and ideological division and variation in knowledge and experience. Increasing diversity in online social networks may either stimulate or restrain political engagement. Thus, the purpose of this study is to examine the role of social network diversity on online political engagement.
Benefits of network diversity
Diverse social networks can be beneficial for democratic participation. Scholars have argued that network diversity facilitates cross–cutting exposure in which individuals encounter new and different information (Mutz and Martin, 2001). On the individual level, exposure to different viewpoints can promote cognitive reappraisal of one’s pre–existing views, which in turn stimulates further information seeking and perspective–taking (Mutz, 2006). On the societal level, awareness of alternative viewpoints and tolerance for political differences can help develop consensus among politically heterogeneous groups (Kim, 2011). Cross–cutting exposure is especially important online. Sunstein (2007) warned that Internet applications filter information based on users’ preferences, resulting in selective exposure to information and viewpoints. As a consequence, individuals are exposed to a limited scope of information and perspectives, and new and different information that would otherwise interest and inspire users is filtered out. More broadly, this may result in opinion polarization and societal fragmentation (Sunstein, 2007).
Empirically, network diversity is beneficial because it is associated with increased general political involvement, operationalized herein as political interest and political efficacy. For example, racial, religious and political diversity of one’s social network stimulate political interest and awareness (Quintelier, et al., 2011; Scheufele, et al., 2006). Exposure to diverse political discussion networks is also related to higher levels of political efficacy, defined as belief in one’s ability to influence the government (Hayes, et al., 2006). Not surprisingly, increased political interest and efficacy are associated with higher levels of political participation like voting and membership in civic groups (Harell, et al., 2008). Increased political interest and political efficacy also positively predict online political engagement (i.e., online political discussion; Best and Krueger, 2005) and SNS use for political purposes (Vitak, et al., 2011).
Burdens of network diversity
Although network diversity is associated with general political involvement, political discussion via diverse social networks may pose risks because diverse networks are characterized by different and often conflicting norms, values and expectations (Joinson, et al., 2011). These risks are particularly salient during political engagement via SNSs for two reasons: individuals are concerned about their image when they use SNSs to share information with others, and the broadcast nature of communication via SNSs means that people are unlikely aware of the scope of their actual audience (Rui and Stefanone, 2013). Access by unwanted or unknown others challenges one’s ability to manage their social images online.
Self–presentation is the strategic revelation of personal information in ways that reflect how one wishes to be regarded by others (Johnson, 1981). The notion of self–presentation is rooted in Goffman’s (1959) concept of face. To protect one’s face, individuals tend to present themselves as socially desirable to others and adjust their social images to their audience’s expectations. However, diverse online social networks raise the problem of multiple audiences wherein users find it difficult to manage their diverse relationships through broadcast–level communication (Joinson, et al., 2011). Prior research has addressed this relationship between network diversity and self–presentation. In off–line interaction, network diversity is associated with intergroup anxiety (Levin, et al., 2003). In online communication, diversity of SNS networks, along the dimensions of demographics, private/professional relationship and kinship, predicts higher online tension (Binder, et al., 2009; Trepte and Reinecke, 2011). The multiple audience problem is especially salient in terms of political engagement on SNSs. Individuals construct their identities and build potential for social influence through political activities online (Gil de Zúñiga, et al., 2012). Disclosing political ideology, revealing one’s political activities and expressing political viewpoints can cause approval as well as disapproval in interpersonal interaction. These behaviors can enhance or compromise one’s image and standing in groups (Hayes, et al., 2006; Ikeda and Boase, 2011). Compared to political discussions via traditional online communication tools (e.g., discussion boards) in which identity can be concealed, SNS profiles are closely bound to actual identities. SNSs, such as Facebook and Google+, have real–name policies. It is also a norm to disclose real identity and communicate with actual friends on SNSs (Steinfield, et al., 2012). Therefore, shared political information online can propagate across an entire social network, affecting actual off–line relationships.
Prior studies have linked network diversity to political activity (see Harell, et al., 2008; Ikeda and Boase, 2011; Jun, 2012; Quintelier, et al., 2011; Visser and Mirabile, 2004). Here, network diversity has been operationalized as opinion diversity (Jun, 2012; Visser and Mirabile, 2004) and religious/language/socio–economic/political diversity (Harell, et al., 2008; Quintelier, et al., 2011). Although scholars have suspected that in political discussion, individuals tend to discuss matters that bring agreement and share moderate viewpoints to avoid conflict (Goel, et al., 2010; Harell, et al., 2008; Mutz, 2006; Visser and Mirabile, 2004), empirical findings on the relationship between diversity and political participation demonstrate a mixed pattern (Harell, et al., 2008; Quintelier, et al., 2011). Some studies show that network diversity dampens participation whereas other studies show a positive relation between diversity and political participation (Quintelier, et al., 2011).
Examining network diversity’s dual roles
Based on the literature review, network diversity likely has mixed effects on political engagement via SNSs. On the one hand, network diversity may increase general political involvement (i.e., political interest and political efficacy) which can stimulate political engagement online. On the other hand, people are likely to encounter disagreements regarding political views and values they share with diverse social networks. To avoid disapproval and conflict in interpersonal relationships, individuals are likely cautious in posting content related to politics.
Two types of self–presentation have been identified (Arkin, 1981). Engaging in acquisitive self–presentation, individuals project a positive self–image in order to seek approval. To avoid disapproval, people use protective self–presentation by withholding or otherwise limiting access to undesirable personal information. Strategies for protective self–presentation include conformity to social norms and avoiding radical viewpoints (Arkin, 1981).
Protective self–presentation as defined by Arkin (1981) is possible via SNS interaction due to user control over social cues available in computer–mediated communication (CMC). SNS users can strategically disclose socially desirable information while withholding undesirable bits. In addition, the asynchronous nature of CMC allows room for conscious self–editing and self–censoring (Walther, 1996; 2007). Protective self–presentation is also enhanced by adjusting privacy settings to personal needs (Trepte and Reinecke, 2011).
In this study, we operationalize political engagement on SNSs along two dimensions: public political engagement and its counterpart protective political engagement. Specially, public political engagement includes posting publicly about politics and revealing political identity on profile pages (profile disclosure); protective political engagement includes self–censoring of shared information (self–censoring) and limiting others’ access to shared information (access restrictions). Network diversity may affect both dimensions of political engagement.
To examine the stimulating effect of network diversity on political engagement, we postulated a set of hypotheses about the relationship between network diversity and public political engagement on SNSs based on the literature reviewed above. Specially, we postulate that
H1: Online network diversity is positively correlated with political interest and political efficacy.
H2: Political interest and political efficacy are positively correlated with posting publicly about politics on SNSs.
H3: Political interest and political efficacy are positively correlated with self–disclosure of political identity on SNS profiles.
To examine the restraining effects of network diversity, we focus on the relationship between network diversity and protective political engagement on SNSs. We postulate that
H4: Network diversity is positively correlated with self–censoring on SNSs.
H5: Network diversity is positively correlated with access restriction on SNSs.
Given that human behavior is shaped by a range of state– and trait–level variables, it should be noted that the restraining effect may vary for people with different psychological traits. A review of existing self–disclosure literature suggested that the traits of self–monitoring and disclosiveness likely interact with network diversity on political engagement (Arkin, 1981).
Self–monitoring describes the extent to which people pay attention to their image in social settings (Lennox and Wolfe, 1984; Snyder, 1974). High self–monitors can better comprehend social cues and show greater concern for social appropriateness (Douglas, 1983). High self–monitors can also adapt to situations by varying the levels of self–disclosure (Tardy and Hosman, 1982). In online interaction, high self–monitors engage in more strategic privacy management practices (Child and Agyeman–Budu, 2010). Particularly during conversations about politics, high self–monitors are more attentive to others’ views and may pretend that they share others’ views to be socially appropriate (Eveland, et al., 2011). As far as network diversity is concerned, high self–monitors should be more attentive to social pressures from diverse audiences, resulting in more protective political behaviors. Therefore, we postulated that
H6: The effect of network diversity on self–censoring is greatest for high self–monitors.
H7: The effect of network diversity on access restriction is greatest for high self–monitors.
In addition, disclosiveness is a person’s tendency to disclose personal information to others (Wheeles, 1978). People with higher disclosive tendencies more often reveal information about themselves and in more depth to general others (Wheeles, 1978). Not surprisingly, tendencies to self–disclosure generalize to online behavior as well. Research shows that highly disclosive individuals tend to tell more about themselves online, and the content of their disclosures also contain more private information (Hollenbaugh, 2011). This evidence suggests that disclosiveness likely counters the restraining effect of network diversity on political engagement. Thus,
H8: The effect of network diversity on self–censoring is lowest for people with high disclosure tendencies.
H9: The effect of network diversity on access restriction is lowest for individuals with high disclosure tendencies.
To address these hypotheses, 255 students enrolled in undergraduate communication courses at at the University at Buffalo were recruited to complete a survey. Participation was contingent on active use of SNSs such as Facebook and Twitter. All participation was voluntary and the university Institutional Review Board for Human Subjects approved all materials. About 53 percent of the sample was female and the average age of participants was 21 (SD=3.0) years. The majority of participants identified their ethnic background as Caucasian (approximately 63 percent). Although we used a convenience sample, college students are heavy SNS users. On average, the participants spent about 3.2 hours (SD=2.67) daily on SNSs.
Network diversity was operationalized along two dimensions: relational and political diversity. Political diversity was measured by asking participants to rate the perceived difference between themselves and their online friends in terms of general political viewpoints, favorite political party, favorite political candidate, and their position on one specific political issue: abortion. To cross–validate participant responses, we also asked participants how often they disagree with political viewpoints expressed by their friends online. Typical seven–point Likert–style responses ranged from one (strongly disagree) to seven (strongly agree). These items were assessed for reliability and condensed to create a political diversity scale (M=23.1; SD=8.78; α=.89), with higher scores representing greater diversity.
Relational diversity was operationalized as an additive index of different types of relationships (McCarty, et al., 2001). We considered relational diversity because political values are based, in large part, on social background (Quintelier, et al., 2011). Relations listed in the index include family, co–workers, school friends, neighbors, friends known through hobbies, and friends known through religious organizations, etc. (see Appendix for a full list). Participants’ SNS friends fell into an average of 8.84 categories (SD=2.88) out of 13 possible categories.
Network size was measured by asking how many friends participants have on SNSs. Although network size was based on participant recall, Stefanone and Lackaff (2009) found that young people are reliably able to recall the size of their online networks. Participants reported an average of 577 friends on SNSs (SD=245.26).
Table 1: Descriptive statistics and zero–order correlations for variables; Means (standard deviation) presented along diagonal.
Political engagement on SNSs was operationalized as public and protective political engagement on SNSs. Public political engagement on SNSs consisted of two variables: the level of profile disclosure (profile disclosure) and the extent of posting publicly about politics (public political posting). Profile disclosure was measured by asking participants, “Do you reveal your political views on your profile page?” Public posting about politics was measured by asking participants how often they engage in activities like political identity construction (e.g., joining political groups, etc.), political messaging (e.g., posting text/photo/video that are political in nature, discussions with online friends about political matters), and political mobilization such as RSVP to political events (see Vitak, et al., 2011). Protective political engagement on SNSs includes self–censoring and access restriction. The Facebook impression management scale (Smock, 2010) was modified to measure both self–censoring and access restriction. Participants were asked how often they delete or limit access to shared political information.
To validate our scales of public and protective political engagement on SNSs (see Table 2), we conducted an exploratory factor analysis (KMO=.91; Bartlett’s test: χ2=3979.63, p<0.001). Eight items addressing public political posting loaded together (eigenvalue=8.85) and were condensed into a public political posting scale (M=15.14; SD=9.24; α=.95). Higher scores indicate more frequent public political posting behavior. Next, five items addressing self–censoring loaded together (eigenvalue=2.66) and were condensed into a self–censoring scale (M=10.75; SD=6.60; α=.90). High scores indicate more frequent self–censoring of politically related information on SNSs. In addition, four items addressing access restriction loaded together (eigenvalue=1.63) and were condensed into one scale (M=8.76; SD=6.38; α=.96). Public posting about politics (M=1.12; SD=.22) and self–censoring (M = .96; SD=.25) as well as access restriction (M=.84; SD=.28) were skewed, so they were log transformed to correct the distribution.
Table 2: Component matrix for factor analysis of political engagement items.
Political self–efficacy was measured by a single item which addressed whether participants think they can influence government, following Gil de Zúñiga, et al. (2012). This single item measure was used by Anderson and Tverdova (2001). Political interest was measured by a single item asking participants to rate their agreement with the statement “I am interested in political issues.“ Single–item measure may have the potential for attenuation, but there is evidence that single items have comparable reliability and predictive validity as multiple–item scales when they are used to measure concreate constructs (Bergkvist and Rossiter, 2007; Wanous and Hudy, 2001). On average, most participants did not rate their political efficacy very high (M=3.40, SD=1.75) and political interest (M=3.45, SD=1.85).
Self–monitoring was measured using Lennox and Wolfe’s (1984) seven–point Likert scale. We selected five items with highest loadings from the original scale to increase its reliability (M=24.56, SD=4.65, α=.70). Disclosiveness was measured with Wheeless’ (1978) scale which measured amount, depth and honesty of participants’ self–disclosure in general social settings (M= 17.43, SD=5.34, α=.75).
Table 1 presents descriptive findings and correlations for the variables used in this study. To test H1, we conducted an ordinary least squares (OLS) regressions using political interest and political efficacy as the dependent variables. Political diversity, relational diversity, and demographic variables were entered into the model. The first regression model explained eight percent of the total variance in political interest, F(5,225)=4.72, p<.001. Males (β=.17, p<.05) and those with a politically diverse network (β=.19, p<.001) demonstrated higher levels of political interest (see Table 3). However, the model predicting political efficacy was not significant (p>.05). Thus, H1 was partially supported.
Table 3: OLS model predicting political interest.
To test H2, we used public posting about politics as the dependent variable (see Table 4). The model explained 24 percent of the total variance, F(5,225)=15.32, p<.001. Political interest (β=.45, p<.001) was positively correlated with public posting about politics, with non–white participants (β=-.13, p<.001) reporting more public posting about politics. H2 was also partially supported.
Table 4: OLS model predicting public political posting.
Next, we examined H3 which addressed the relationship between network diversity and identity disclosure on SNS profiles. Note that profile disclosure was measured as a dichotomous variable (0=non–disclosure, 1=disclosure of political identity on profile page), so we employed logistic regression. In the final model (see Table 5), χ2=39.8 (p<.001) and the Cox & Snell R2=.16 indicate that the model performed well. Political interest (β=.55, p<.001) was positively correlated with profile disclosure, although political efficacy was not associated with profile disclosure. Thus, H3 was partially supported.
Table 5: Logistic regression model predicting profile disclosure.
Next we turned to the second set of hypotheses regarding associations between network diversity and protective political engagement. We used self–censoring and access restriction as dependent variables, respectively. In both models (see Tables 6 and 7), possible moderators — self–monitoring and disclosiveness — were entered along with relational diversity, political diversity and network size. In addition, separate interaction terms for self–monitoring with political diversity and with relational diversity were entered in the model. Interaction terms for disclosiveness with political diversity and with relational diversity were entered in the next model.
When self–censoring was the dependent variable (see Table 6), the model addressing disclosiveness as a moderator explained 11.2 percent of the total variance, F(10, 220)=3.91, p<.001. As expected, political diversity (β=.15, p<.05) was positively correlated with self–censoring. However, contrary to our expectation, relational diversity (β=-.16, p<.05) was found to be negatively correlated with self–censoring. The model addressing self–monitoring as a moderator explained 10 percent of the total variance, F(10, 220)=3.41, p<.001. Similarly, political diversity (β=.14, p<.05) positively predicts self–censoring. Relational diversity (β=-.20, p<.05) was negatively correlated with self–censoring. Thus, H4 was partially supported. Political diversity and relational diversity were not significant predicators of access restriction (see Table 7), offering no support for H5.
Table 6: OLS model predicting self–censoring.
Table 7: OLS model predicting access restriction on SNSs.
Disclosiveness functioned as a moderator (β=-.14, p<.05) in the relationship between relational diversity and self–censoring (see Table 6). In order to better understand the interaction between relational diversity and self–censoring, we first re–centered disclosiveness variable and conducted OLS regression to compare the associations of relational diversity with self–censoring when disclosiveness was high and low. This association was only significant (β=-.25, p<.05) when disclosiveness was high. We then re–centered relational diversity variable and compared the associations of disclosiveness with self–censoring when relational diversity was high and low. This association was only significant (β=.28, p<.05) when relational diversity was low. Figure 1 was created to visualize this result. Although the moderation effect was present, its direction was contrary to our prediction. Thus, H8 was not supported.
Figure 1: Interaction of disclosiveness and relational diversity on self–censoring. Note: Figure X. relational diversity split at 11.3; disclosiveness split at 17.61.
In addition, we found no interaction effect for self–monitoring in the two models that predicted self–censoring and access restriction, respectively (see Tables 6 and 7), so H6 and H7 were not supported. We found no moderation effect for disclosiveness in the mode predicting access restriction (H9; see Table 7).
Additionally, self–monitoring (β=.15, p<.05) was positively correlated with relational diversity (see Table 8) consistent with extant research showing that high self–monitors are more likely to expand their social networks resulting in larger and more diverse social groups (Snyder, et al., 1983).
Table 8: OLS model predicting self–monitoring and relational diversity.
Discussion and conclusion
This study sought to explain the effect of social network diversity on political engagement via SNSs. The motivation of this study lies in the popularity of SNSs and the increasing use of these sites for political purposes. We examined the benefits and burdens of network diversity, operationalized as relational and political diversity respectively, on political engagement. Specifically, we investigated the relationship between network diversity and two political engagement behaviors on SNSs: public and protective political self–presentation. Our result sheds light on how network diversity may influence individual political behaviors on SNSs.
We first found political diversity in online networks positively predicts political interest. The finding supports the cross–cutting argument (Mutz, 2006). Individuals with friends from a broader political spectrum likely have more opportunities to receive alternative viewpoints, which sparks information seeking and discussions about politics. Ultimately, political information seeking and discussions may arouse more political interest. We further found that those participants with increased political interest were more likely to disclose their political identity and post publicly about politics. The finding is consistent with previous research linking increased political interest to more political participation both off–line and online (Harell, et al., 2008; Vitak, et al., 2011). Taken together, these results demonstrate advantages associated with online network diversity. The findings also highlight the role of political interest. Recent surveys show that college students are apathetic about politics, which is detrimental to political participation (Moeller, 2012). SNS use increases the diversity of one’s social network (Hampton, et al., 2011), and given the link between political interest and network diversity, students should be encouraged to use SNSs to interact with politically different others to reduce political apathy.
We also found that political diversity in online networks was associated with increased self–censoring behavior. Recall that individuals strategically present themselves for social approval (Rui and Stefanone, 2013). Discussion about politics, especially controversial political topics, often leads to disagreement (Quintelier, et al., 2011), and this risk is magnified in diverse networks. As a response, SNSs users delete posts or restrain from adding content about controversial topics. This behavior demonstrates that individuals yield to social pressure inherent in heterogeneous networks. Therefore, while network diversity may stimulate political interest and political engagement, it also likely burdens us with relational concerns, thus restraining carefree political expression.
Contrary to SNSs based on actual relationships, interest–oriented SNSs that connect strangers through shared interest may be an alternative venue for more candid and opinionated political discussion. Loose connections in interest–oriented online networks may exert less social pressure. It is important to distinguish interest–oriented SNSs from relationship–oriented SNSs in future research on network diversity and political behaviors.
Interestingly, relational diversity was negatively correlated with self–censoring. Although political and relational diversity tend to be correlated (Quintelier, et al., 2011), SNS users may not perceive different social ties as having disparate political viewpoints. Additionally, high relational diversity could result from indiscriminate friending whereby users add strangers to their online networks (Stefanone and Lackaff, 2009). Users are not invested in these relationships, which reduces the need for self–censoring. Future studies should distinguish between the two types of diversity in discussing the role of network diversity on political engagement.
We further found that the effect of network diversity on self–censoring varied depending on individual traits. People with high self–disclosure tendencies and low diversity networks are the most likely to self–censor. Extant research shows that people with high self–disclosure tendencies are more likely to use Internet applications to maintain relationships (Stefanone and Jang, 2007). Recall that we argued previously about higher relational diversity being a result of indiscriminate friending. It is likely that disclosive individuals with low diversity networks are those who use SNSs to maintain meaningful networks of actual relationships. As a consequence, these users are increasingly mindful about the content of their online communication.
Taken together, our findings show that network diversity plays contradictory roles in political engagement via SNSs and is context specific. On the one hand, political diversity increased political interest which was associated with political engagement on SNSs. Yet, political diversity was associated with increased self–censoring behavior. Thus, it is important for future research to further clarify the role network diversity regarding online political engagement.
We identify the following limitations with the current study. First, the hypothesized relationship between political diversity and political efficacy is not evident. The relationship may be indirect. Exposure to different political views may lead to increased political interest and political knowledge, which then makes individuals feel increasingly capable of influencing government decisions. Future research should address this possibility. Future research should also control for political knowledge which was not included in the current models. In addition, the sample demographic in this study is college students, and the majority of sampled students have never used SNSs for political engagement. These factors limit the generalizability of findings. Lastly, we used single items to measure political interest and political efficacy. However, single–item measures generate comparable reliability and validity as multiple item measures when used to measure concrete constructs (Bergkvist and Rossiter, 2007; Wanous and Hudy, 2001).
Future research should employ more sophisticated approaches to measure network diversity and political activity, as well. For example, instead of asking participants to recall the frequency of engaging in political activities online, participants or researchers could be asked to browse SNS profiles to report actual incidents of political activity. Off–line social networks may serve as an important antecedent to online political engagement, as well, and should be controlled for. In addition to the above limitations, we believe that future researchers need to distinguish between two types of network diversity: perceived and actual network diversity. Note that perceived network diversity was employed in the current study. With regard to the burdens of network diversity, perceived network diversity is relevant because people need to cognitively assess network diversity first in order to evaluate the risk of disagreement from friends. Yet, political interest might be stimulated by actual network diversity. In other words, perceived network diversity does not necessarily indicate opportunities for cross–cutting exposure.
In spite of these limitations, our study serves as a first step towards exploring the role of network diversity in influencing political behavior on SNSs. We revealed that network diversity plays contradictory roles in stimulating and restraining political engagement. The contradictory role sets agenda for future studies.
About the authors
Weiai Wayne Xu is a doctoral student in the Department of Communication at the University at Buffalo, The State University of New York. He is interested in political engagement on social networking sites and social impact of technology use.
E–mail: weiaixu [at] buffalo [dot] edu
Michael A. Stefanone is an Associate Professor in the Department of Communication, University at Buffalo, The State University of New York. His research interest focuses on the novel use of new communication technology and its impact on human relationships.
Jian Raymond Rui is a Ph.D. candidate in the Department of Communication, University at Buffalo, The State University of New York. His research interest focuses on the use and effect of new communication technology on the relationship and individuals.
C.J. Anderson and Y.V. Tverdova, 2001. “Winners, losers, and attitudes about government in contemporary democracies,” International Political Science Review, volume 22, number 4, pp. 321–338.
http://dx.doi.org/10.1177/0192512101022004003, accessed 3 September 2013.
R.M. Arkin, 1981. “Self–presentation styles,” In: J.T. Tedeschi (editor). Impression management theory and social psychology research. New York: Academic Press, pp.311–333.
L. Bergkvist and J.R. Rossiter, 2007. “The predictive validity of multiple–item versus single–item measures of the same constructs,” Journal of Marketing Research, volume 44, number 2, pp. 175–184.
S.J. Best and B.S. Krueger, 2005. “Analyzing the representativeness of Internet political participation,” Political Behavior, volume 27, number 2, pp. 183–216.
http://dx.doi.org/10.1007/s11109-005-3242-y, accessed 3 September 2013.
J. Binder, A. Howes, and A. Sutcliffe, 2009. “The problem of conflicting social spheres: Effects of network structure on experienced tension in social network sites,” CHI ’09: Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, pp. 965–974.
J. Brundidge, 2010. “Encountering ‘difference’ in the contemporary public sphere: The contribution of the Internet to the heterogeneity of political discussion networks,” Journal of Communication, volume 60, number 4, pp. 680–700.
http://dx.doi.org/10.1111/j.1460-2466.2010.01509.x, accessed 3 September 2013.
J.T. Child and E.A. Agyeman–Budu, 2010. “Blogging privacy management rule development: The impact of self–monitoring skills, concern for appropriateness, and blogging frequency,” Computers in Human Behavior, volume 26, number 5, pp. 957–963.
http://dx.doi.org/10.1016/j.chb.2010.02.009, accessed 3 September 2013.
W. Douglas, 1983. “Scripts and self–monitoring when does being a high self–monitor really make a difference?” Human Communication Research, volume 10, number 1, pp. 81–96.
http://dx.doi.org/10.1111/j.1468-2958.1983.tb00005.x, accessed 3 September 2013.
W.P. Eveland, A.C. Morey, and M.J. Hutchens, 2011. “Beyond deliberation: New directions for the study of informal political conversation from a communication perspective,” Journal of Communication, volume 61, number 6, pp. 1,082–1,103.
http://dx.doi.org/10.1111/j.1460-2466.2011.01598.x, accessed 3 September 2013.
H. Gil de Zúñiga, N. Jung, and S. Valenzuela, 2012. “Social media use for news and individuals’ social capital, civic engagement and political participation,” Journal of Computer–Mediated Communication, volume 17, number 3, pp. 319–336.
http://dx.doi.org/10.1111/j.1083-6101.2012.01574.x, accessed 3 September 2013.
S. Goel, W. Mason, and D. Watts, 2010. “Real and perceived attitude agreement in social networks,” Journal of Personality and Social Psychology, volume 99, number 4, pp. 611–621.
http://dx.doi.org/10.1037/a0020697, accessed 3 September 2013.
E. Goffman, 1959. The presentation of self in everyday life. Garden City, N.Y.: Doubleday.
A. Harell, D. Stolle, and E. Quintelier, 2008. “Network diversity and political participation: A complication or asset?” paper presented at the Youth and Politics Conference (Bruges, 3–4 July).
N. Hamdy and E.H. Gomaa, 2012. “Framing the Egyptian uprising in Arabic language newspapers and social media,” Journal of Communication, volume 62, number 2, pp. 195–211.
http://dx.doi.org/10.1111/j.1460-2466.2012.01637.x, accessed 3 September 2013.
K.N. Hampton, C.–J. Lee, and E.J. Her, 2011. “How new media affords network diversity: Direct and mediated access to social capital through participation in local social settings,” New Media & Society, volume 13, number 7, pp. 1,031–1,049.
http://dx.doi.org/10.1177/1461444810390342, accessed 3 September 2013.
A.F. Hayes, D.A. Scheufele, and M.E. Huge, 2006. “Nonparticipation as self–censorship: Publicly observable political activity in a polarized opinion climate,” Political Behavior, volume 28, number 3, pp. 259–283.
http://dx.doi.org/10.1007/s11109-006-9008-3, accessed 3 September 2013.
E.E. Hollenbaugh, 2011. “Motives for maintaining personal journal blogs,” Cyberpsychology, Behavior, and Social Networking, volume 14, numbers 1–2, pp. 13–20.
http://dx.doi.org/10.1089/cyber.2009.0403, accessed 3 September 2013.
K. Ikeda and J. Boase, 2011. “Multiple discussion networks and their consequence for political participation,” Communication Research, volume 38, number 5, pp. 660–683.
http://dx.doi.org/10.1177/0093650210395063, accessed 3 September 2013.
J.A. Johnson, 1981. “The ‘self–disclosure’ and ‘self–presentation’ views of item response dynamics and personality scale validity,” Journal of Personality and Social Psychology, volume 40, number 4, pp. 761–769.
http://dx.doi.org/10.1037/0022-35126.96.36.1991, accessed 3 September 2013.
A.N. Joinson, D.J. Houghton, A. Vasalou, and B.L. Marder, 2011. “Digital crowding: Privacy, self–disclosure, and technology,” In: S. Trepte and L. Reinecke (editors). Privacy online: Perspectives on privacy and self–disclosure in the social Web. New York: Springer–Verlag, pp. 33–45.
http://dx.doi.org/10.1007/978-3-642-21521-6_4, accessed 3 September 2013.
N. Jun, 2012. “Contribution of Internet news use to reducing the influence of selective online exposure on political diversity,” Computers in Human Behavior, volume 28, number 4, pp. 1,450–1,457.
http://dx.doi.org/10.1016/j.chb.2012.03.007, accessed 3 September 2013.
Y. Kim, 2011. “The contribution of social network sites to exposure to political difference: The relationships among SNSs, online political messaging, and exposure to cross–cutting perspectives,” Computer in Human Behavior, volume 27, number 2, pp. 971–977.
http://dx.doi.org/10.1016/j.chb.2010.12.001, accessed 3 September 2013.
A.O. Larsson and H. Moe, 2012. “Studying political microblogging: Twitter users in the 20 Swedish election campaigns,” New Media & Society, volume 14, number 5, pp. 729–747.
http://dx.doi.org/10.1177/1461444811422894, accessed 3 September 2013.
R.D. Lennox and R.N. Wolfe, 1984. “Revision of the self–monitoring scale,” Journal of Personality and Social Psychology, volume 46, number 6, pp. 1,349–1,364.
http://dx.doi.org/10.1037/0022-35188.8.131.529, accessed 3 September 2013.
S. Levin, C. van Laar, and J. Sidanius, 2003. “The effects of ingroup and outgroup friendships on ethnic attitudes in college: A longitudinal study,” Group Processes & Intergroup Relations, volume 6, number 1, pp. 76–92.
http://dx.doi.org/10.1177/1368430203006001013, accessed 3 September 2013.
C. McCarty, P.D. Killworth, H.R. Bernard, E.C. Johnsen, and G.A. Shelley, 2001. “Comparing two methods for estimating network size,” Human Organization, volume 60, number 1, pp. 28–39.
S. Moeller, 2012. “New study: College students show little interest in politics this election year,” Huffington Post (12 August), at http://www.huffingtonpost.com/susan-moeller/study-college-students-sh_b_1777322.html, accessed 3 September 2013.
D.C. Mutz, 2006. Hearing the other side: Deliberative versus participatory democracy. Cambridge: Cambridge University Press.
D.C. Mutz and P.S. Martin, 2001. “Facilitating communication across lines of political difference: The role of mass media,” American Political Science Review, volume 95, number 1, pp. 97–114.
E. Quintelier, D. Stolle, and A. Harell, 2011. “Politics in peer groups: Exploring the causal relationship between network diversity and political participation,” Political Research Quarterly, volume 65, number 4, pp. 868–881.
http://dx.doi.org/10.1177/1065912911411099, accessed 3 September 2013.
L. Rainie, A. Smith, K.L. Schlozman, H. Brady, and S. Verba, 2012. “Social media and political engagement,” Pew Research Center’s Internet & American Life Project (19 October), at http://pewinternet.org/Reports/2012/Political-engagement.aspx, accessed 3 September 2013.
J. Rui and M.A. Stefanone, 2013. “Strategic self–presentation online: A cross–cultural study,” Computers in Human Behavior, volume 29, number 1, pp. 110–118.
http://dx.doi.org/10.1016/j.chb.2012.07.022, accessed 3 September 2013.
D.A. Scheufele, B.W. Hardy, D. Brossard, I.S. Waismel–Manor, and E. Nisbet, 2006. “Democracy based on difference: Examining the links between structural heterogeneity, heterogeneity of discussion networks, and democratic citizenship,” Journal of Communication, volume 56, number 4, pp. 728–753.
http://dx.doi.org/10.1111/j.1460-2466.2006.00317.x, accessed 3 September 2013.
M.A. Stefanone and D. Lackaff, 2009. “Reality television as a model for online behavior: Blogging, photo, and video sharing,” Journal of Computer–Mediated Communication, volume 14, number 4, pp. 964–987.
http://dx.doi.org/10.1111/j.1083-6101.2009.01477.x, accessed 3 September 2013.
M.A. Stefanone and C.–Y. Jang, 2007. “Writing for friends and family: The interpersonal nature of blogs,” Journal of Computer–Mediated Communication, volume 13, number 1, pp. 123-140.
http://dx.doi.org/10.1111/j.1083-6101.2007.00389.x, accessed 3 September 2013.
C. Steinfield, N. Ellison, C. Lampe, and J. Vitak, 2012. “Online social network sites and the concept of social capital,” In: F.L.F. Lee, L. Leung, J.L. Qiu, and D.S.C. Chu (editors). Frontiers in new media research. New York: Routledge, pp. 115-131.
C.R. Sunstein, 2007. Republic.com 2.0. Princeton, N.J.: Princeton University Press.
A. Smock, 2010. “Self–presentation on Facebook: Managing content created by the user and others,” paper presented at the annual meeting of International Communication Association (Singapore).
M. Snyder, 1974. “Self–monitoring of expressive behavior,” Journal of Personality and Social Psychology, volume 30, number 4, pp. 526–537.
http://dx.doi.org/10.1037/h0037039, accessed 3 September 2013.
M. Snyder, S. Gangestad, and A. Simpson, 1983. “Choosing friends as activity partners: The role of self–monitoring,” Journal of Personality and Social Psychology, volume 45, number 5, pp. 1,061–1,072.
http://dx.doi.org/10.1037/0022-35184.108.40.2061, accessed 3 September 2013.
C.H. Tardy and L.A. Hosman, 1982. “Self–monitoring and self–disclosure flexibility: A research note,” Western Journal of Speech Communication, volume 46, number 1, pp. 92–97.
http://dx.doi.org/10.1080/10570318209374067, accessed 3 September 2013.
S. Trepte and L. Reinecke, 2011. “The social Web as a shelter for privacy and authentic living,” In: S. Trepte and L. Reinecke (editors). Privacy online: Perspectives on privacy and self–disclosure in the social Web. New York: Springer–Verlag, pp. 61–73.
http://dx.doi.org/10.1007/978-3-642-21521-6_6, accessed 3 September 2013.
S. Valenzuela, N. Park, and K.F. Kee, 2009. “Is there social capital in a social network site? Facebook use and college students’ life satisfaction, trust, and participation,” Journal of Computer–Mediated Communication, volume 14, number 4, pp. 875–901.
http://dx.doi.org/10.1111/j.1083-6101.2009.01474.x, accessed 3 September 2013.
P.S. Visser and R.R. Mirabile, 2004. “Attitudes in the social context: The impact of social network composition on individual–level attitude strength,” Journal of Personality and Social Psychology, volume 87, number 6, pp. 779–795.
http://dx.doi.org/10.1037/0022-35220.127.116.119, accessed 3 September 2013.
J. Vitak, P. Zube, A. Smock, C.T. Carr, N. Ellison, and C. Lampe, 2011. “It’s complicated: Facebook users’ political participation in the 2008 election,” Cyberpsychology, Behavior, and Social Networking, volume 14, number 3, pp. 107–114.
http://dx.doi.org/10.1089/cyber.2009.0226, accessed 3 September 2013.
J.B. Walther, 2007. “Selective self–presentation in computer–mediated communication: Hyperpersonal dimensions of technology, language, and cognition,” Computers in Human Behavior, volume 23, number 5, pp. 2,538–2,557.
http://dx.doi.org/10.1016/j.chb.2006.05.002, accessed 3 September 2013.
J.B. Walther, 1996. “Computer–mediated communication: Impersonal, interpersonal, and hyperpersonal interaction,” Communication Research, volume 23, number 1, pp. 3–43.
http://dx.doi.org/10.1177/009365096023001001, accessed 3 September 2013.
J.P. Wanous and M.J. Hudy, 2001. “Single–item reliability: A replication and extension,” Organizational Research Methods, volume 4, number 4, pp. 361–375.
http://dx.doi.org/10.1177/109442810144003, accessed 3 September 2013.
L.R. Wheeless, 1978. “A follow–up study of the relationship among trust, disclosure, and interpersonal solidarity,” Human Communication Research, volume 4, number 2, pp. 143–157.
http://dx.doi.org/10.1111/j.1468-2958.1978.tb00604.x, accessed 3 September 2013.
Received 2 August 2013; revised 24 August 2013; accepted 29 August 2013.
“The benefits and burdens of network diversity: Political engagement on social networking sites” by Weiai Wayne Xu, Michael A. Stefanone, Jian Raymond Rui is licensed under a Creative Commons Attribution 3.0 Unported License.
The benefits and burdens of network diversity: Political engagement on social networking sites
by Weiai Wayne Xu, Michael A. Stefanone, and Jian Raymond Rui.
First Monday, Volume 18, Number 9 - 2 September 2013