Teenagers and social network sites: Do off-line inequalities predict their online social networks?
First Monday

Teenagers and social network sites: Do off-line inequalities predict their online social networks? by June Ahn

This study analyzes a dataset of 701 U.S. teenagers (ages 12–18) that merges an online survey of social network site (SNS) preferences with administrative records from their public school districts. Using a multinomial logistic model, I examine whether off–line divides across gender, ethnicity, socioeconomic status, self–esteem, and social capital predict teenagers’ membership in the popular SNSs, Facebook and Myspace. The results show that the characteristics of teens that use Facebook, Myspace, or both SNSs show distinct differences, which illuminate questions of digital divide and complex adolescent social practices as they relate to online participation. The study offers two main contributions by providing an analysis of: (a) teenage SNS users, a population that is less examined in research on online communities; and, (b) the relationship between their off–line characteristics and online social networks.


Conceptual framework
Discussion and conclusion




Online social networks are currently a major way through which individuals interact on the Internet. Facebook now boasts over 500 million members (Facebook, 2011) and many Web services incorporate social network features to promote user interaction and participation (Livingstone, 2008). The use of social network sites (SNSs) has grown dramatically across the United States and the world. National and global surveys in 2010 report that 46 percent of the U.S. population now uses social network sites (Pew Global Attitudes Project, 2010). Among teenagers, the uptake of SNSs is even more widespread. Of teenagers who utilize the Internet, 73 percent are members of some SNS community (Lenhart, et al., 2010).

While survey data suggests that the adoption of SNSs is becoming ubiquitous among younger populations, the statistics may mask finer grained social divides. Young people across many factors (gender, ethnicity, socioeconomic status etc.) appear to participate in online social networks. However, questions remain concerning how individuals participate, in what communities, and with whom. Do social divides persist in membership into online social network communities? Understanding this question is vital for many areas such as the digital divide, media effects, social computing, and digital youth research.

In this article, I examine whether systematic differences exist in preferences for Facebook and Myspace (or both) in a sample of U.S. teenagers. First, the article outlines how past and current research has examined the uptake of social network tools in adolescent and young adult populations. Second, I work from these theoretical frameworks to ask whether there are systematic differences in SNS preferences among a sample of U.S. teenagers. This study utilizes survey data of 701 teenagers merged with administrative records provided by their public school districts. I use multinomial logistic regression analysis to examine whether there are significant differences in social characteristics between users of Facebook, Myspace, and those who are ambidextrous; or participate in both SNS communities.

Third, the findings of the study suggest distinct differences in the makeup of Facebook users, Myspace users, and youths who utilize both social network sites. For example, Facebook teenagers came from more privileged backgrounds compared to Myspace members. Teenagers in the sample also divided by factors such as ethnicity and prior academic achievement in their preferences for Facebook and Myspace. Finally, I offer a discussion of the findings and its implications for future research in online communities. In particular, the findings help focus researchers on the social and cultural factors that determine participation in online communities. In addition, the results help to shed light on the unequal distribution of individuals across different online networking communities, and offer hints towards areas of digital divide that persist despite the growing ubiquity of Internet use among young people.



Conceptual framework

The need to understand the SNS preferences of users is vital for several areas of research in online communities. First, digital divide research typically examines whether there are systematic inequalities of access to new technologies (Norris, 2001). Such a perspective compels one to consider whether there are unequal levels of participation across factors such as gender, ethnicity, and socioeconomic status. The picture surrounding digital divides has evolved tremendously over the past decade. Early studies of Internet or computer access typically found stark disparities across many demographic indicators. For example, minority populations and individuals from lower socioeconomic backgrounds were much less likely to have access to technology compared to their white, or wealthy peers (U.S. National Telecommunications and Information Administration, 2000). Early studies of youths also consistently found that males were more likely users of technology than females (Volman and van Eck, 2001).

The current picture of the United States is of an evolving country where access to the Internet, inexpensive hardware devices, and tools such as social media are becoming widespread. For example, recent surveys find that 93 percent of teens ages 12–17 go online (Lenhart, et al., 2010). American families with children are also highly connected, with 93 percent owning a family computer and 94 percent of households having at least one member who regularly accesses the Internet (Wellman, et al., 2008). As technology becomes a more ubiquitous element of life, the idea of digital divides shifts to questions of inclusion and participation. Warschauer (2002) observes that access to technology is not a binary variable, but rather there are nuanced gradients to how individuals access and use the Internet or computers. Furthermore, Internet researchers consistently observe a second–level divide of participation and technology skills. For example, Hargittai and Hinnant (2008) found significant differences in technology literacy skills and the types of websites individuals visit, across a sample of young adults. Factors such as gender and technology access predicted technology skills, while indicators such as education level were significantly correlated to what types of sites an individual frequented.

This nuanced understanding of a second–level digital divide suggests that there is tremendous variance in the types of sites individuals may visit, the interactions they have, and the participation patterns they enact. Similarly, researchers find large variance in the types of technology skills, or literacy skills, that users of the Internet possess (Hargittai, 2002, 2005). Across factors such as age and gender, researchers find tremendous variability in abilities to search for information, assess the credibility of resources online, or use new technology tools (Druin, et al., 2010; Hargittai, et al., 2010). The emerging understanding is that mere access to technology, as a binary definition, is becoming less of an issue as tools become more readily available. However, divides continue to exist concerning types of access, to different tools, and for various personal uses and goals.

As social media becomes a ubiquitous part of teenage life, it is prudent to also consider whether youth access to these technologies is also unequal. Social tools such as SNSs are central gathering spaces for teenagers to develop new media literacy skills (Jenkins, 2006). For example, Jenkins (2006) notes that various skills become vital in networked spaces such as the ability to: collaborate with others, adopt changing identities as one navigates through different communities, or explore new knowledge domains when one has access to vast repositories of information. Early ethnographic studies of young people and social media support these theories of new media skills. Researchers have found that teenagers find creative ways to gain access to new technologies, participate in various online communities that help them learn new skills, and delve into deep learning on topics that are personally interesting to them (Ito, et al., 2010). Teenagers also utilize social network sites to provide social support to peers, share creative work, and network with others (Greenhow and Robelia, 2009). Participating in communities such as Facebook or Myspace provide new venues through which individuals learn these new media skills. Systematic differences in access to these online communities may thus reflect opportunities, or lack thereof, for particular youth populations to cultivate these literacy skills.

In addition to a concern for digital divides affecting the participation of young people in SNSs, studies that examine these differences also enlighten other research concerns. For example, many of the popular questions surrounding youths and SNSs ask what effects these technologies have on outcomes such as academic achievement or the development of social relationships (Ahn, 2011a). Social network sites may assist young people in developing better relationships, or social capital, with their network (Ellison, et al., 2007; Valenzuela, et al., 2009). Conversely, social media tools may also facilitate negative and dangerous interactions for young people such as breaches of privacy and cyberbullying (Palfrey, et al., 2009). Researchers have also asked questions such as whether the use of SNSs affects academic achievement (see Hargittai and Hsieh, 2010; Pasek, et al., 2009).

Questions of media effects, or what consequences arise from the participation of young people in SNSs, are difficult to examine without considering the characteristics that youth bring into a respective online community. Self–selection looms large when examining behavior in social network sites, but little is known about how individuals self–select into particular communities. Early studies of online communities suggest that factors such as ethnicity and gender define members of different SNSs (Hargittai, 2007). Thus, there is a vital antecedent question of whether individuals bring different traits to their participation in respective online communities, and if these systematic differences exist between users of different sites. This study contributes to the literature by considering a sample of young people (under 18). In addition, the analyses presented here offer insights into the social divides which persist in the membership of teens in social network sites.

Member preferences for social network sites

There have been few studies that consider systematic differences in user characteristics of SNSs, and even less that focus on youth populations. In one study, Ahn (2011b) utilized a national 2007 survey of teenagers from the Pew Internet & American Life Project. Several findings in the study highlighted the changing face of digital divide, access, and technology use in the U.S. For example, black teenagers were 1.42 times more likely to be SNS members than their white peers. Female teens were also more likely than males to have a SNS profile. Furthermore, traditional indicators of digital divide such a socioeconomic status (SES) and place of Internet access had opposite than expected relationships. Parents’ education beyond a high school diploma, a common indicator of SES, did not have a significant relationship to teens’ use of SNSs. In addition, youths who primarily accessed the Internet away from home or school were most likely to be SNS members (Ahn, 2011b). The findings suggested that systematic differences in user characteristics existed, but no longer in traditionally expected ways as defined by the digital divide literature. However, this study did not differentiate users between particular SNS communities such as Facebook or Myspace, and only considered a binary variable of whether a teenager had an SNS profile or not.

Qualitative accounts also hint at various differences in teenage preferences for SNSs. In the current literature, boyd (2012) offers the most in–depth description of the complex cultural and social divides between teenage SNS members. She observed that Myspace and Facebook teens strongly identify their choice of SNS community with issues of race and class. For example, the teenagers in her ethnographic study used particular language and terms to describe peers who are Myspace or Facebook members. Young people described Myspace as a “ghetto,” signifying both the population of members who preferred the site (often minority and from lower SES backgrounds), and the types of activities and tastes that were privileged there (music, entertainment, etc.). Conversely, Facebook evolved as a community that was more college oriented, middle class, and white.

boyd (2012) highlighted how the adoption of SNSs for teenagers was very far from random. Rather the complex racial, socioeconomic, cultural, and relationship dynamics that characterize high school life played a large role in the online communities that young people choose. Early quantitative accounts also supported this emerging understanding that complex social divides persisted in the makeup of online social networks. The most relevant study comes from Hargittai (2007), which examined the SNS preferences of a sample of college students. Hargittai found that demographic indicators, such as ethnicity and SES, were not related to whether these college students were SNS members. However, female students were more likely than their male peers to be SNS users.

The results were striking when Hargittai (2007) disaggregated her analysis by particular SNS community. Factors such as race and socioeconomic status began to define members of respective communities. For example, Hispanic students were less likely to use Facebook and more likely to be Myspace members compared to their white peers. Asian students were less likely to be Myspace users but highly likely to be Xanga members. Finally, individuals from lower socioeconomic backgrounds were more likely to be Myspace members than any other SNS community. As Hargittai (2007) noted, the findings suggest that “different populations select into the use of different services, posing a challenge to research that tends to collapse use of all social network sites” [1]. There remains an unexplored opportunity to explore systematic differences in teenage preferences for SNSs using similar quantitative analysis. This study extends the literature to examine teenagers’ preferences for two popular social network sites: Myspace and Facebook.

In the following analysis, I consider several hypotheses that arise from the nascent research concerning youths and social network sites. The first hypothesis speaks to the relationship between demographic differences and SNS membership. In the studies mentioned thus far, researchers have found that factors such as age, gender, ethnicity, and socioeconomic status were highly related to individuals’ choices of SNSs. These considerations frame the first general hypothesis:

H1 — Age, gender, ethnicity and socioeconomic status will be predictive of teenagers’ choices of Facebook or Myspace.

Prior studies also offer expectations for which social network site a teenager might choose based on off–line social divides. For example, boyd (2012) and Hargittai (2007) found that ethnic minority groups are less likely to use Facebook. In addition, teenagers from lower SES backgrounds may opt to choose Myspace. The second hypothesis thus examines this proposition:

H2 — Teenagers from ethnic minority and lower SES backgrounds will be more likely to choose Myspace than Facebook as their primary SNS.

In addition to demographic indicators, prior research suggests that complex factors surrounding culture, taste, personal characteristics and community factors are related to SNS participation. I include several independent variables that may describe aspects of teenagers’ personal characteristics. First, one often–examined construct in the literature is self–esteem. Self–esteem typically describes one’s evaluation of self–worth, happiness, or satisfaction (Valkenburg, et al., 2006). Prior studies have found that individuals use social network sites in different ways, or accrue different social benefits from SNSs, based on their level of self–esteem (Ellison, et al., 2007; Valkenburg, et al., 2006). However, no studies to this author’s knowledge have considered self–selection, or whether users might differ in existing levels of self–esteem in different SNS communities.

Second, the concept of social capital is also highly related to behavior surrounding social network sites. Social capital refers to the idea that one’s relationships or networks provide benefits such as information, emotional support, or advice (Portes, 1998). Different theorists focus on disparate elements of social capital theory (i.e., Coleman, 1990; Putnam, 2000), and in this study I utilize measures that relate to bonding and bridging social capital. Bonding capital describes close relationships with others that often bring benefits such as emotional support and scarce resources. Bridging capital describes acquaintances that may not represent close relationships, but bring a different set of resources such as new information. As outlined below, I utilize measures of bonding and bridging social capital from the Internet Social Capital Scales (Williams, 2006).

The research literature has consistently found that the use of Facebook is related to higher levels of social capital (Ellison, et al., 2007; Valenzuela, et al., 2009). Most recently, researchers have found that Facebook use may be more related to bridging relationships, but that particular behaviors such as actively seeking out social information about peers can be related to bonding relationships (Burke, et al., 2011; Ellison, et al., 2011). What has not been considered thus far has been whether individuals with differing levels of social capital congregate in distinct SNS communities. In a teenage sample it may be that teenagers with more bonding capital may cluster into one SNS, while peers with more bridging social capital interact in another. Such questions begin to illuminate differences in the culture, interface, and interactions of various SNS communities. The studies cited above have typically only focused on the effects of Facebook, but in this study I consider the hypothesis that teenagers bring varying levels of social capital to different SNS communities. Combining this consideration of social capital with the discussion of self–esteem, the third hypothesis considers whether divides in psychosocial factors are related to teenagers’ preferences for SNSs:

H3 — Teenagers’ levels of self–esteem, bonding social capital, and bridging social capital will be predictive of their choice of Facebook or Myspace.

Finally, the dataset in this study is unique, combining survey data of teenagers’ SNS preferences with administrative records of academic performance provided by their public school districts. Thus, I also examine whether teenagers cluster into particular SNS communities based on their level of academic achievement. Academic achievement is a variable that describes a student’s performance in school. However, the variable can also be viewed in this analysis as an indicator of teenagers’ personal identity and social context. Being a high or low achiever in school is intimately related to a teenagers’ social life, network of friends, daily activities and personal identity. Thus, a measure of academic achievement may also predict the social network communities a teenager joins. Do high achieving students cluster into different online communities compared to their lower achieving peers, and vice versa? The fourth hypothesis suggests that varying levels of academic achievement are influential in teenagers’ preferences for SNSs.

H4 — A teenager’s level of academic achievement is predictive of their choice to use Facebook or Myspace.

Hypotheses 1–4 offer the prediction that offline social differences are highly related to teenagers’ choices to use Facebook or Myspace. However, it is also plausible that teenagers use multiple online communities simultaneously. No studies to date have considered the question of whether young people who participate in multiple communities, and navigate different social circles, exhibit different characteristics than their peers who identify closely with a particular platform. In early studies, the analysis has typically considered only members of one SNS community or grouped individuals into distinct SNSs without a consideration for multiple membership (i.e., Hargittai, 2007). Recently, scholars have begun to theorize and classify SNS members according to their diverse preferences. For example, Hargittai and Hsieh (2011) classify individuals as a devotee if they only utilize one SNS with any frequency. Conversely, omnivores are individuals who frequently use multiple SNSs. I contribute to the literature by examining the potential that young people who are ambidextrous in their SNS preferences — or use both Myspace and Facebook — exhibit different social characteristics than their peers who only utilize one SNS (Myspace or Facebook).

H5 — Teenagers who are ambidextrous in their SNS preferences (omnivores) will exhibit different social characteristics — along demographic, psycho–social, and academic factors — than their peers who are members of only one community (devotees).




Data for this study comes from a survey of high school students in two, urban school districts. The districts are located in two major cities in the western United States. All data was collected in Autumn 2009. With the participation of the districts, high school students were invited to participate in a Web–based survey. Teenagers who participated in the study also returned signed parental consent forms to their teachers, prior to accessing the survey. The survey first asked youths if they were members of Facebook and/or Myspace. The questionnaire also collected measures of self–esteem and social capital (outlined below). The data from the Web–based survey was then merged with administrative records provided by the school districts. Each student in the district had a unique, district–provided identification number (ID) that is used to track academic records. The students who returned their consent forms were entered into a database connected to the Web–based survey. These participants were then allowed to log in using their district ID numbers to complete the survey. The identification numbers allowed for the merging of the survey data with the students’ administrative records. These records provided key background variables including gender, ethnicity, age, prior academic achievement, and socioeconomic status. In this study, 701 students returned their consent forms and provided complete data for the analysis.

Data analysis and variables

This study utilizes a multinomial logistic regression model to predict teenagers’ preferences for Facebook (only), Myspace (only), or for both SNSs. The dependent variable considered four groups: Teenagers who did not use either Facebook or Myspace (coded as 0), members of Facebook only (1), members of Myspace only (2), and members who utilized both SNSs (3). The base, reference group were youths who did not use any SNS (0). The multinomial logistic model predicts the odds that a teenager adopts Facebook, Myspace, or both SNSs (compared to peers who do not use any SNS).

Hypotheses 1 and 2 suggest that teenagers who differ along demographic indicators — age, gender, ethnicity, and socioeconomic status — will segregate into distinct SNS communities. The demographic variables used in the model include: age, gender, ethnicity, and socioeconomic status. Gender was a categorical variable, with female as the reference group. For ethnicity, white teenagers were the reference group. This categorical variable included Asian and Hispanic teens. Black teenagers and those of other ethnic groups were too few in this sample (less than 20) and inclusion of such a small sample would bias the multinomial regression results. Thus, these participants were excluded from the analysis. Finally, the school districts provided the teenagers’ free–reduced lunch status, which is a common indicator of socioeconomic status. Teenagers who qualify for free–reduced lunch in school fall below a specified poverty index.

Hypothesis 3 posits that the psycho–social factors of self–esteem and social capital are predictive of teenagers’ SNS preferences. To measure self–esteem, I utilized nine items of the Rosenberg self–esteem scale, which has been used in previous studies of social network sites and user behavior (i.e., Ellison, et al., 2007). The scale asked participants various questions about how they feel about themselves and their capabilities. Cronbach’s alpha for the self–esteem scale was 0.843. To measure social capital, I utilized the Internet Social Capital Scales (ISCS) (Williams, 2006). Five items were modified from the ISCS to capture the bonding social capital teenagers’ have in their high school relationships. The items used were: (1) There are people in my high school I trust to help solve my problems. (2) There is someone in my high school I can turn to for advice about making very important decisions. (3) There is no one in my high school I feel comfortable talking to about my personal problems (reverse coded). (4) When I feel lonely, there are several people in my high school I can talk to. (5) If I needed an emergency loan of $500, I know someone in school I can turn to.

Students responded on a 1–4 scale that ranged from strongly disagree, somewhat disagree, somewhat agree, and strongly agree. Cronbach’s alpha for this social capital scale was 0.717. I also considered teenagers’ bridging, online social capital using five items from the ISCS. These items were: (1) Interacting with people online makes me interested in things that happen outside of my town. (2) Interacting with people online makes me want to try new things. (3) Talking with people online makes me curious about other places in the world. (4) Talking with people online makes me feel like part of a larger community. (5) Interacting with people online makes me feel connected to the bigger picture. For these items, the participants also responded on a 1–4 scale. Cronbach’s alpha for this scale was 0.888.

To measure prior academic achievement (Hypothesis 4), the participating school districts provided academic data from student records. I utilized the teenagers’ prior cumulative grade point average (GPA). This cumulative GPA is the average grade, calculated from all previous classes on the student’s academic record. The cumulative GPA ranges from a 0–4 scale. The average GPA in this sample was 3.07 with a standard deviation of 0.82. All descriptive statistics are provided in Table 1.


Table 1: Descriptive statistics.
Note: N=701.
VariableMean (std.)Min.–Max.
Age15.44 (1.36)12–18
Self–esteem28.41 (12.69)10–36
School social capital14.74 (3.01)5–20
Online social capital13.50 (3.66)5–20
Prior GPA3.07 (0.82)0.33–4.5
Free–reduced lunch26%
SNS membership 
Facebook only21%
Myspace only27%
Both SNSs39%


One should note that this sample is not nationally representative and represents a convenience sample of high school youths in two particular U.S. cities. In this sample, the majority of students were either White or Hispanic, with a third major population of Asian participants. Black, multi–racial, or Native American teenagers were much less represented here. Approximately 26 percent of the teenagers qualified for free–reduced lunch at their school. Finally, the breakdown of SNS preferences is also of interest. The majority (nearly 40 percent) of youths in this sample used both Facebook and Myspace. The second largest group consisted of teens that only use Myspace, with Facebook as the third most popular option. Finally, approximately 13 percent in this sample did not use either Facebook or Myspace (Table 1).




The results of the multinomial logistic model are presented in Table 2. Note that in place of beta coefficients, I present odds ratios to aid with interpretation. Values above 1.0 represent higher odds of choosing a respective SNS compared to the reference group. A value of less than 1.0 signifies lower odds compared to the reference group. For example, the relationship between being male and choosing Myspace is an odds ratio of 0.59. The result suggests that the odds of a male preferring Myspace as their SNS of choice are 41 percent less likely than a female choosing Myspace. A related interpretation is that male teenagers in this sample are less likely to be Myspace users, while their female peers are more likely Myspace members. The results in Table 2 provide consistent evidence that off–line divides along gender, ethnicity, social capital, academic achievement, and socioeconomic status are related to teenagers’ decisions to participate in specific SNSs.

Hypothesis 1 stated that demographic indicators — age, gender, ethnicity, and socioeconomic status — would predict teenagers’ preferences for social network sites. The results offer support for the notion that off–line social divides predict young people’s online communities as well. In this sample, older teens preferred Myspace while younger adolescents tended to choose Facebook as their SNS of choice. Female teenagers were more likely to choose Myspace or be ambidextrous in their SNS participation (i.e., used both Facebook and Myspace). Asian youths were less likely to use any of the SNSs in this study, while Hispanic teens showed higher odds of choosing Myspace as their primary social network site (Odds ratio = 1.84, ρ<0.10). Finally, young people living in poverty (free–reduced lunch) showed clear preferences to be Myspace members or were ambidextrous in their SNS membership. These results together highlight how off–line social divides across age, gender, ethnicity, and socioeconomic status were significantly related to young people’s choices of online communities. These off–line divides persist in teenagers’ online networks as well.


Table 2: Odds ratio of adopting an SNS.
Note: N=701; Pseudo r2=0.16; * ρ<0.10, ** ρ<0.05; (beta coefficients, standard errors) in parentheses; Beta coefficients converted to odds ratios.
 FacebookMyspaceBoth SNSs
(-0.19, 0.11)
(0.26, 0.11)
(0.04, 0.09)
(-0.25, 0.28)
(-0.52, 0.28)
(-0.53, 0.25)
(-0.79, 0.36)
(-1.12, 0.46)
(-0.52, 0.33)
(-1.96, 0.46)
(0.61, 0.34)
(-0.75, 0.33)
(0.01, 0.03)
(0.00, 0.03)
(0.01, 0.03)
Bonding social capital1.10**
(0.10, 0.05)
(0.18, 0.05)
(0.05, 0.04)
Bridging social capital1.14**
(0.13, 0.04)
(0.04, 0.03)
(0.11, 0.03)
Prior GPA1.38
(0.32, 0.22)
(-1.02, 0.20)
(-0.45, 0.18)
Free–reduced lunch0.83
(-0.19, 0.46)
(0.90, 0.36)
(0.65, 0.34)


The second hypothesis (H2) suggested that teenagers from ethnic minority groups and lower SES backgrounds would likely choose Myspace, compared to their peers who were Caucasian or from higher SES backgrounds. The results offer empirical support for boyd’s (2012) observation that factors such as race and class are related to young people’s decisions to participate in different SNS communities. As noted previously, Hispanic youths were highly likely to adopt Myspace as their SNS of choice. In addition, these youths were substantially less likely to adopt Facebook or be ambidextrous in their SNS preferences. For this sample of Hispanic teenagers, Myspace was the place to network and interact with friends online, despite the overall mainstream popularity of Facebook. Furthermore, lower SES teenagers overwhelmingly preferred Myspace or were ambidextrous in their SNS preferences. The findings in Table 2 suggest that Caucasian and higher SES teens preferred Facebook, while their Hispanic or lower SES peers chose Myspace as their online community. Such results mirror qualitative accounts of these complex patterns of community among teenage populations (boyd, 2012).

Hypothesis 3 explored the potential that young people bring differing levels of self–esteem and social capital to their SNS communities. The results (Table 2) show no support that teenage members of Facebook or Myspace differ in their average level of self–esteem. However, differences in social capital were intriguing across social network sites. Youth who had higher bonding and bridging social capital were more likely to choose Facebook as their primary SNS (controlling for other factors). However, teenagers who reported only higher bonding social capital were likely to be Myspace members. Finally, young people who reported only higher bridging capital were more likely to be ambidextrous in their SNS participation.

Such results lend support to Hargittai’s (2007) observation that constraints present in one’s off–line life are also present in online interactions as well. Teenagers in this sample brought differing levels of social capital to their membership in either Myspace or Facebook. How this social capital is enacted through interactions within those sites has been a topic of recent studies (Burke, et al., 2011; Ellison, et al., 2011). However, this analysis offers a foundation for further questions relating SNSs and social capital development. Perhaps different technical affordances in Myspace and Facebook were related to levels of bridging and bonding social capital. Did Facebook with its broad audience, standardized interface, and middle class image among teens attract individuals with different social capital than Myspace with its focus on entertainment, busy but expressive interface, and specialized member base?

Equally plausible is that the variance in social capital that teenagers bring to an SNS is tightly related to their social and cultural contexts. For example, ethnographic researchers have found that families from lower SES backgrounds use technology to develop closer bonds to local relationships, while those from higher SES families also use technology for bridging relationship development over wide distances (Ames, et al., 2011). Perhaps these sociocultural factors can help to explain why Facebook teens (who typically come from higher SES backgrounds) brought more bonding and bridging capital, while Myspace teens (typically from lower SES families) brought more bonding social capital. By including psycho–social factors such as social capital, this study highlights how different SNS communities may hold varying levels of potential social capital irrespective of the kinds of relationship building interactions its members may enact after joining.

The fourth hypothesis also relates to this notion that young people bring differing levels of capital and personal resources to their SNS communities. As Table 2 shows, teenagers who had higher GPAs (or were higher academic achievers) were substantially less likely to be members of Myspace or ambidextrous users. The inverse interpretation is that young people who did not use any SNSs and their peers who were devotees of Facebook exhibited higher academic achievement than their Myspace or ambidextrous peers. This clustering of young people along academic lines may be highly predictive of the types of interactions and experiences they have in the online community. For example, perhaps high school teens that are likely to attend college congregate in Facebook while their lower achieving peers network on Myspace. If one observed more posts and comments about academics or applying to college on Facebook, such results may be less likely a function of Facebook as a platform, but more likely a result of the characteristics of the members themselves. An attention to self–selection dynamics lends deeper insight into the types of interactions and effects one observes from a given platform.

Finally, Hypothesis 5 suggested that young people who were members of both Facebook and Myspace (omnivores) would exhibit starkly different characteristics than their peers who were devotees of only one SNS. The results (Table 2) support this hypothesis. There are clear differences in teenage members of respective SNSs along demographic and social divides. However, the differences and stories of these youths become clearer when one views these factors as a holistic whole. Thus, in the following sections I offer an overview of the types of teenagers who chose different SNSs in this sample.

Profile of a Facebook teen

The findings (Table 2) show that teenagers who adopted only Facebook as their primary SNS were distinctly different as a group than their peers. Overall, young people who were devotees of Facebook came from more privileged backgrounds. Facebook members were less likely to come from ethnic minority groups, or rather more likely to be white. In addition, Facebook members in this sample also reported higher social capital (both bonding and bridging) compared to their peers who did not use any of the SNSs. Facebook teens were higher academic achievers (and they did not differ in academic achievement from the reference group of peers with no SNS membership). Socioeconomic status was a non–significant predictor of adopting Facebook (compared to the reference group). However, Facebook members and peers who did not use either SNS, as a group, tended to come from higher SES backgrounds than Myspace or ambidextrous teens. Viewing all of these characteristics together, the profile of a Facebook teen in this sample was one of higher levels of privilege, connections, and academic achievement.

Profile of a Myspace teen

The average Myspace member in this sample was an older, Hispanic, female teenager. This young person was also a lower academic achiever and came from poorer family backgrounds. The differences between teens that adopt Myspace and their peers were stark. Along ethnic lines, Hispanic teenagers overwhelmingly adopted Myspace as their primary SNS. The odds of choosing Myspace was 84 percent higher for Hispanic youth than their white peers. Teenagers on Myspace were also lower academic achievers and overwhelmingly likely to qualify for free–reduced lunch at school. A teenager that qualified for free–reduced lunch at school was 2.45 times more likely to adopt Myspace compared to being a non–user. Such results begin to paint a picture of high school life for these students. Facebook members tended to come from backgrounds of higher privilege. Myspace teenagers were more likely to come from lower SES families.

Profile of an ambidextrous teen

Finally, the characteristics of teenagers who utilized both SNSs differed considerably than their Facebook–only and Myspace–only peers (H5). An ambidextrous teenager was most likely to be a White female, who had lower academic achievement and came from a low SES family. The odds of Hispanic teens being ambidextrous SNS members were 53 percent less likely than their Caucasian peers. The ambidextrous SNS user was most likely to be female (Odds ratio = 0.59, ρ<0.05). These teenagers had lower academic achievement (Odds ratio = 0.64, ρ<0.05). They also did not report higher bonding social capital compared to the reference group, while their Facebook and Myspace peers did bring higher levels of bonding capital. However, ambidextrous teenagers showed an inclination to develop bridging social capital (Odds Ratio = 1.12, ρ<0.05). Finally, if a teenager qualified for free–reduced lunch in school, they were more likely to be an ambidextrous SNS member.

The combination of these social factors offers intriguing questions for the social life of these ambidextrous SNS members. Did White, female youth from lower SES backgrounds relate to different social circles? Perhaps their higher SES, White friends were overwhelmingly on Facebook. Conversely, their lower SES friends with lower academic achievement (who would have likely been in similar high school classes) were clearly socializing on Myspace. Bridging across different, segregated social groups might have necessitated these teenagers to participate in multiple online communities.

Another compelling question is how these ambidextrous teens were connected to their local community. Youth who reported higher bonding social capital in their high school community were more likely to be Facebook or Myspace devotees. Comparatively, ambidextrous teenagers reported higher bridging social capital in their online activities compared to the reference group. Were these youth more apt to bridge across different social circles, but as a consequence were less likely to have many close friends at school? The patterns found in this dataset show unique social divides that warrant deeper study. Teenagers clearly cluster in online communities around gender, ethnicity, and indicators of privilege. There are also young people who bridge seemingly segregated social circles and exhibit very different social and cultural characteristics. The results of this study suggest that off–line divides do persist into these teenagers’ online communities and individuals who actively participate in multiple communities may be a particularly intriguing population to study. What are the social and cultural divides that these individuals bridge as they interact in different social network sites? Such comparative analysis and description may yield deeper insights into the cultural dynamics of different SNS communities and subsequently what they suggest for off–line inequality and cultural dynamics as well.



Discussion and conclusion

Several limitations of this study are salient when interpreting the findings. First, the data represents a convenience sample of 701 teenagers in two school districts. These districts are located in urban areas of the western United States. The social characteristics of these youths are very specific to this context. For example, the majority of teens in this sample were White, Hispanic, or Asian. It is highly likely that in other regions of the U.S., the ethnic breakdown of a student sample would differ considerably.

The findings presented here should not be interpreted as generalizable to all populations. For example, the findings in Table 2 do not suggest that all Asian youths are less likely to use SNSs compared to their peers. An equally plausible finding may be that Asian teenagers in other contexts may be more likely SNS users. In addition, Asian teenagers might cluster into other online communities that were not examined in this study. Nevertheless, the results of this study highlight how social divides within any context are highly predictive of teenagers’ participation in online social network communities. Future research that examines the unique patterns of online participation and social divides promises to contribute richer knowledge about these cultural phenomena.

Second, this study presents cross–sectional data and cannot imply causality. For example, bonding social capital in a teenager’s high school community was positively related to their adoption of Facebook and Myspace, but not related to ambidextrous adoption. In the context of this analysis, the interpretation is that young people who were highly connected off–line chose particular SNS communities, but not both. However, an equally plausible causal direction might be that teenagers’ use of Facebook or Myspace is related to higher social capital. The likely answer is that young people’s existing level of social capital is related to their choice of online community. Their subsequent activities in the social network site then help them maintain or further develop their relationships with others. Future longitudinal studies that control for already existing levels of social capital may then offer stronger claims to the effect of SNSs on relationship development (i.e., Burke, et al., 2011).

Finally, this research only considers teenagers’ preferences for two popular SNSs: Facebook and Myspace. The results show clear demarcations along social divides among users of these SNSs. However, Internet activity and participation in online communities is highly diverse. While 13 percent of teens in this sample were not members of Facebook or Myspace, they may participate in other, less popular online communities. Understanding of less mainstream patterns of online participation is also needed. Furthermore, as new social technologies develop, and teenagers shift their preferences for online participation over time, future studies that consider a more diverse set of media will develop a richer picture of teens and technology use.

This study offers one of the first analyses to consider the relationship between off–line social divides and online SNS participation in teenage populations. The patterns that emerge from this dataset offer potentially rich ways to identify teenage users of technology that goes beyond single indicators of SES, gender, or ethnicity. This data provides two critical contributions to the literature on social network sites and young people. First, prior ethnographic research suggests that teen adoption of online communities is intricately wrapped up in their developing identities and social life (boyd, 2012). Few studies have examined the social dynamics of why teenagers choose one SNS over another, but this study lends empirical support to emerging findings by researchers such as boyd (2012). The social divides that separate teenagers in their off–line lives, largely predict their choices to participate in online communities.

The second contribution of this study is to identify complex social patterns of participation in SNS communities. The typical Facebook teenager in this sample was middle class, white, and had higher academic achievement. These youth had higher quality relationships with peers in school and also used the Internet to develop broader relationships online. Conversely, Hispanic teenagers in this sample clearly preferred Myspace as their online community of choice. Myspace teenagers had higher quality relationships with peers in school, but did not show a higher preference to develop online relationships. Finally, lower SES, White females were most likely to be ambidextrous SNS users, and the reasons behind these trends warrant future study of similar sub–populations of SNS members.

These patterns illuminate how behaviors around SNSs are not monolithic, but rather quite diverse. Furthermore, teenagers’ choices to use SNSs are related to cultural and social factors such as ethnicity, SES, and community dynamics (i.e., social capital). The findings of this study compel deeper questions of how different teenage populations use SNSs, and how this use is related to their cultural and social lives. Did Myspace teens utilize the SNS to interact with known friends and develop stronger local ties? Did Facebook teenagers use the SNS to interact with their local friends, but also have numerous peers (perhaps in college) that required more online relationship development? The patterns of SNS preferences found in this sample of teenagers underscore the need to understand diverse user groups.

Overall, the analysis presented here pinpoints how social divides around SES, race, and high school achievement predict young people’s decisions to interact online. From a digital divide perspective, access and use of the Internet is becoming ubiquitous for teenage users. However, this study shows that participation in specific communities such as social network sites largely mirrors the social divides that segregate people in their off–line lives. Digital divides may continue to exist, but in a nuanced manner around individuals’ choices to participate in specific communities. The study illuminates new user populations that warrant richer, qualitative study and poses important questions about the effect of SNSs on young people. If off–line inequalities persist into online communities, what does that mean for the effects these tools have on social life? Further research in this area promises to yield deeper insights into how young people use new media tools and how new technologies are embedded within the personal, social, and cultural lives they lead. End of article


About the author

Dr. June Ahn is an assistant professor at the University of Maryland, College Park. He holds joint appointments in the College of Information Studies and College of Education. His research focuses on the impact of social technologies on young people’s social development, learning, and education. He also explores the social informatics of technology use in education settings, or how institutional policies, culture, and practices interact with new media tools to create technology–enhanced learning environments.
E–mail: juneahn [at] umd [dot] edu



1. Hargittai, 2007, para. 50.



June Ahn, 2011a. “The effect of social network sites on adolescents’ academic and social development: Current theories and controversies,” Journal of the American Society for Information Science & Technology, volume 62, number 8, pp. 1,435–1,445.

June Ahn, 2011b. “Digital divides and social network sites: Which students participate in social media?” Journal of Educational Computing Research, volume 45, number 2, pp. 147–163.http://dx.doi.org/10.2190/EC.45.2.b

Morgan G. Ames, Janet Go, Joseph J. Kaye, and Mirjana Spasojevic, 2011. “Understanding technology choices and values through social class,” CSCW ’11: Proceedings of the ACM 2011 Conference on Computer Supported Cooperative Work (Hangzhou, China, 19–23 March), pp. 55–64.

danah boyd, 2012. “White flight in networked publics? How race and class shaped American teen engagement with Myspace and Facebook,” In: Lisa Nakamura and Peter Chow–White (editors). Race after the Internet. New York: Routledge, pp. 203–222.

Moira Burke, Robert Kraut, and Cameron Marlow, 2011. “Social capital on Facebook: Differentiating uses and users,” CHI ’11: Proceedings of the ACM Conference on Human Factors in Computing Systems (7–11 May, Vancouver, B.C.), pp. 571–580.

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

Allison Druin, Elizabeth Foss, Hilary Hutchinson, Evan Golub, and Leshell Hatley, 2010. “Children’s roles using keyword search interfaces at home,” CHI ’10: Proceedings of the 28th International Conference on Human Factors in Computing Systems (10–15 April, Atlanta, Ga.), pp. 413–422.

Nicole B. Ellison, Charles Steinfield, and Cliff Lampe, 2011. “Connection strategies: Social capital implications of Facebook–enabled communication practices,” New Media & Society, volume 13, number 6, pp. 873–892.http://dx.doi.org/10.1177/1461444810385389

Nicole B. Ellison, Charles Steinfield, and Cliff Lampe, 2007. “The benefits of Facebook ‘friends’: Social capital and college students’ use of online social network sites,” Journal of Computer–Mediated Communication, volume 12, number 4, at http://jcmc.indiana.edu/vol12/issue4/ellison.html, accessed 26 August 2011.

Facebook, 2011. “Statistics,” at http://www.facebook.com/press/info.php?statistics#!/press/info.php?statistics, accessed 26 August 2011.

Christine Greenhow and Beth Robelia, 2009. “Old communication, new literacies: Social network sites as social learning resources,” Journal of Computer–Mediated Communication, volume 14, number 4, pp. 1,130–1,161.http://dx.doi.org/10.1111/j.1083-6101.2009.01484.x

Eszter 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 August 2011.

Eszter Hargittai, 2005. “Survey measures of Web–oriented digital literacy,” Social Science Computer Review, volume 23, number 3, pp. 371–379.http://dx.doi.org/10.1177/0894439305275911

Eszter Hargittai, 2002. “Second–level digital divide: Differences in people’s online skills,” First Monday, volume 7, number 4, at http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/942/864, accessed 26 August 2011.

Eszter Hargittai and Yu–li Patrick Hsieh, 2011. “From dabblers to omnivores: A typology of social network site usage,” In: Zizi Papacharissi (editor). A networked self: Identity, community and culture on social network sites. New York: Routledge, pp. 146–168.

Eszter Hargittai and Yu–li Patrick Hsieh, 2010. “Predictors and consequences of differentiated practices on social network sites,” Information, Communication and Society, volume 13, number 4, pp. 515–536.http://dx.doi.org/10.1080/13691181003639866

Eszter Hargittai and Amanda Hinnant, 2008. “Digital inequality: Differences in young adults’ use of the Internet,” Communication Research, volume 35, number 5, pp. 602–621.http://dx.doi.org/10.1177/0093650208321782

Eszter Hargittai, Lindsay Fullerton, Ericka Menchen–Trevino, and Kristen Yates Thomas, 2010. “Trust online: Young adults’ evaluation of Web content,” International Journal of Communication, volume 4, pp. 468–494, and at http://ijoc.org/ojs/index.php/ijoc/article/view/636, accessed 22 December 2011.

Mizuko Ito, Sonja Baumer, Matteo Bittanti, danah boyd, Rachel Cody, Becky Herr–Stephenson, Heather A. Horst, Patricia G. Lange, Dilan Mahendran, Katynka Z. Martinez, C.J. Pascoe, Dan Perkel, Laura Robinson, Christo Sims, and Lisa Tripp, 2010. Hanging out, messing around, and geeking out: Kids living and learning with new media. Cambridge, Mass.: MIT Press.

Henry Jenkins, 2006. Confronting the challenges of participatory culture: Media education for the 21st century. Chicago: John D. and Catherine T. MacArthur Foundation.

Amanda Lenhart, Kristen Purcell, Aaron Smith, and Kathryn Zickuhr, 2010. “Social media & mobile Internet use among teens and young adults,” Washington, D.C.: Pew Internet & American Life Project, at http://pewresearch.org/pubs/1484/social-media-mobile-internet-use-teens-millennials-fewer-blog, accessed 22 December 2011.

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

Pippa Norris, 2001. Digital divide: Civic engagement, information poverty, and the Internet worldwide. New York: Cambridge University Press.

John Palfrey, Dena T. Sacco, danah boyd, and Laura DeBonis, 2009. “Enhancing child safety and online technologies,” Cambridge, Mass.: Berkman Center for Internet & Society at Harvard University, at http://cyber.law.harvard.edu/pubrelease/isttf/, accessed 22 December 2011.

John Pasek, eian more, and Eszter Hargittai, 2009. “Facebook and academic performance: Reconciling a media sensation with data,” First Monday, volume 14, number 5, at http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/2498/2181, accessed 26 August 2011.

Pew Global Attitudes Project, 2010. “Global publics embrace social networking,” at http://pewglobal.org/2010/12/15/global-publics-embrace-social-networking/, accessed 26 August 2011.

Alejandro Portes, 1998. “Social capital: Its origins and applications in modern sociology,” Annual Review of Sociology, volume 24, pp. 1–24.http://dx.doi.org/10.1146/annurev.soc.24.1.1

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

U.S. National Telecommunications and Information Administration , 2000. Falling through the Net: Toward Digital inclusion. Washington, D.C.: National Telecommunications and Information Administration, at http://www.ntia.doc.gov/report/2000/falling-through-net-toward-digital-inclusion, accessed 22 December 2011.

Sebastián Valenzuela, Namsu Park, and Kerk 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

Patti M. Valkenburg, Jochen Peter, and Alexander P. Schouten, 2006. “Friend networking sites and their relationship to adolescents’ well–being and social self–esteem,” CyberPsychology & Behavior, volume 9, number 5, pp. 584–590.http://dx.doi.org/10.1089/cpb.2006.9.584

Monique Volman and Edith van Eck, 2001. “Gender equity and information technology in education: The second decade,” Review of Educational Research, volume 71, number 4, pp. 613–634.http://dx.doi.org/10.3102/00346543071004613

Mark Warschauer, 2002. “Reconceptualizing the digital divide,” First Monday, volume 7, number 7, at http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/viewArticle/967/888, accessed 26 August 2011.

Barry Wellman, Aaron Smith, Amy Wells, and Tracy Kennedy, 2008. “Networked families,” Washington, D.C.: Pew Internet & American Life Project, at http://www.pewinternet.org/Reports/2008/Networked-Families.aspx, accessed 22 December 2011.

Dmitri Williams, 2006. “On and off the ’Net: Scales for social capital in an online era,” Journal of Computer–Mediated Communication, volume 11, number 2, pp. 593–628, and at http://jcmc.indiana.edu/vol11/issue2/williams.html, accessed 22 December 2011.


Editorial history

Received 27 August 2011; accepted 23 December 2011.

Creative Commons License
This work is licensed under a Creative Commons Attribution–NonCommercial–ShareAlike 3.0 Unported License.

Teenagers and social network sites: Do off–line inequalities predict their online social networks?
by June Ahn.
First Monday, Volume 17, Number 1 - 2 January 2012

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

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