Friendship and online filtering: The use of social media to construct offline social networks

Friendship and interpersonal social networks

Many researchers have found that strong friendships and stable social networks are key to the success of college students (Raacke and Bonds-Raacke, 2015; Schuh et al., 2015; Spencer and Pahl, 2006). Friendships also play a role in the formation of individual identities (Adams and Allan, 1999; Johnson et al., 2009; Luyckx and Robitschek, 2014). Thus, networks and college culture can impact the formation of the self and result in long-term changes in individual and community behavior (Schuh et al., 2015). Friendships and social media form the basis of social networks among college students, with negotiations of meaning playing a significant role in the experience (Lambert, 2013; Standlee, 2016; Trottier, 2012). In addition, peer associations play a role in the development of life-long social networks which shape access to social capital, the internalization of social norms, and exposure to and persistence of new or emerging ideas (Antheunis et al., 2016; Su and Chan, 2017).

Recent work suggests that the intertextual and multimodel structure of the technology allows for social integration, centralization, and shared imaginaries that foster a sense of group solidarity (Raacke and Bonds-Raacke, 2015; Thomas, 2007). Scholarship on shifting social practices has detailed how communication technologies are increasingly embedded in social life (Halberstam and Knight, 2014; Masip et al., 2017). Scholars have identified significant consequences to this including: increasing focus on speed and efficiency in relationships (Bakardjieva, 2014), and transformations in the experience of privacy and intimacy (Lambert, 2013). Scholarship on the impact of social media on diversity in social networking is contested (Kim and Kim, 2017; Lönnqvist and Itkonen, 2016; Miller and Slater, 2000), but appears to have impacts on offline relationships as well. Further work suggests that while interests and attitudes generally bind people closer or pull them apart in face-to-face interactions, the amount of effort that is necessary to identify such similarities or differences results in a median level of social intimacy, while techno-mediated friendships are constructed more efficiently and with controlled levels of interpersonal connection (Boase and Wellman, 2006; Miller and Slater, 2000). Among the most visible implications is the increasingly partisan nature of information consumption, sometimes referred to as echo chambers or filter bubbles (Flaxman et al., 2016; Masip et al., 2017; Sunstein, 2009, 2017).

Research from across disciplines has identified the tendency for technologically mediated communications to filter information, allowing individuals to limit their exposure to diverse information (Pariser, 2011; Sunstein, 2009, 2017). This pattern may impact the very foundation of human social patterns, including knowledge consumption, social attitudes, and even the foundations of democracy. Political and social issue partisanship has been one of the area’s most profoundly impacted by this tendency toward social filters (Flaxman et al., 2016; Sunstein, 2017). Some scholars argue that filtering practices may be reshaping how democracy and political identity affiliations work (Boutyline and Willer, 2016; Spohr, 2017; Sunstein, 2017). Others contest this issue, arguing that offline practices, rather than online practices, are most central to this issue (Haim et al., 2017). This article considers the way in which the use of social media contributes to social filtering on the basis of shared social and political views among college students, by discussing the connections between offline and online communications.

Research design and methodology

This project includes data from two primary field sites, a private university in the northeast and a public college in the southeast. I collected data on 52 participants, between the ages of 18 and 24. The average age of my participants was 20 years. In all, 50% of the respondents were female and 46% male, with 2 respondents identifying as non-binary; 33 participants indicated that they were White, 10 Black or African American, 4 as Latino(a)s, 2 Asian American, 2 Native Hawaiian/Pacific Islander, and as 1 Middle Eastern American.

Research was conducted in the hybrid offline/online space inhabited by participants. In order to access this setting and navigate it, I had to understand how participants define concepts such as “Internet” and “digital culture” as a part of their cultural world. For the purposes of this study, the Internet is best understood to be a global network of data transfer (Thurlow et al., 2004). However, digital culture can better be understood as an interactional technology which provides a means of social interaction and representation that is experienced as inextricably connected with the offline social world, and exists within everyday spaces (Miller and Slater, 2000). I gathered data about everyday activities, meanings and attitudes, experiences, and behaviors of participants, within this environment.

I recruited participants through announcements made in various classes at the two institutions. I provided basic information and invited students to provide email addresses for more information. I emailed interested students a project description and a consent form. Potential participants were given about a week to consider. I emailed them again and, if they agreed, set up an interview time and date. No financial compensation for participants was provided, however, several professors routinely offered “extra credit” for participation in faculty and student research projects as a part of their course policies. I asked interviewees for recommendations of persons who might participate, at the end of the interviews. Participants in the online observation portion of this research project were recruited from among interviewees. Interviews were transcribed and coded using grounded coding, allowing trends and themes to emerge from the transcribed documents. Analysis was largely inductive, with findings and theories being developed after observation.

I used traditional face-to-face interviews as the primary data source and supplemented with online ethnographic data collection methods. Questions were open-ended and encouraged participants to tell their own stories and express their own meanings. Examples of topics included: daily Internet use, the value of online technologies, and the importance of technologies and common practices among peers. To maintain privacy, all names used in this work are pseudonyms. The interviews were between about 45 and 90 minutes and 1 hour on average. The digital ethnography generated several hundred pages of postings and field notes. While these data were informative, they were not comprehensive, as privacy settings available in Facebook limited my ability to view all of their postings and data.

Each of the 23 individuals I followed on social media gave permission after an initial interview. Participants, when pressed, requested that I observe them via Facebook, which many explained was both more visible and less private than other platforms. Each participant had to accept the friend request in order for me to view their postings. Once I was given access to their online environment, I collected data by saving postings and completing field notes, which are used for this project. I specifically do not use searchable postings in this article to avoid compromising the privacy of participants.

It is worth noting that while I was able to collect rich, in-depth data, it is not randomized nor is it generalizable. The student experiences I highlight reflected overall trends I was able to identify in the data, but as with any non-randomized ethnographic data this work reflects only the contextual and located experiences of individuals.

Doing your homework: making friends made easier

During the college years, a time of transformation, friendships can be both emotionally intense and socially desirable. Friendships are important to the success of college students (Friedlander et al., 2007; Kim and Kim, 2017; Raacke and Bonds-Raacke, 2015). Yet, the ability to make and keep friends in the midst of social changes and individual life upheaval is challenging. As Olivia explains, social media gives her the ability to create and maintain social relationships, because social meaning has a certain permanence, in her view:

… things happen … people move and things like that. But, with Facebook, if you move, you still have your Facebook. You don’t have to actually delete it and take it down, whereas, … you may have to change your number or get a new phone … or you lose your contacts … with Facebook, it’s just like one of those … it’s like one of those set-in stone, things.

This sense of permanence and consistency plays a role in how college students utilize social media in general. The ability to keep track of friends and to get significant amounts of information about them is very important to participants. The sense of permanence that participants attribute to social media networks also shapes the symbolic logic of using social media as a tool for offline social network construction. There is a belief that it is possible to get a sense of someone’s values and beliefs, over a long period of time, with minimal social investment, via social media. From the perception of participants, filtering potential troublemakers from one’s social world means avoiding those with different social and political attitudes. Among participants, heterogeneous social and political attitudes within an offline social network is viewed negatively, even as being unsafe. Therefore, participants focus on avoiding such social attitude diversity by evaluating online social behavior in an effort to create homogeneous and harmonious offline social networks. For some participants this filtering behavior is overt and acceptable, while for others such practices are somewhat taboo. In order to better understand how effective this behavior is in creating homogeneity in offline social networks, it is essential to understand the social acceptability of the practice.

The degree to which such behavior is socially acceptable relies on individual meanings that are attached to the practice of filtering its self. For just about half of those who consider background checks important, demonstrating too much knowledge about an individual from his or her social media, too early in the relationship is taboo. Emma explains that this practice has become a normalized expectation and is an unspoken assumption about social interactions.

Because if you’re at a social event, and your best friend is like, “Oh, here …, here’s my friend.” I guarantee you every person goes back and checks social media five seconds later, so they can see what they’re all about … like if they are well … bad. Or they post bad stuff.

As Emma describes, there is a presumption that every person is conducting social background checks, and thus everyone has a shared base of knowledge. This knowledge is often unspoken, however, as Emma explains that filtering should only occur without fully acknowledging that the social rejection is the result of conflicting values.

Another participant, Aisha identifies how choices about postings shape the way in which people are judged online. For Aisha, the nature of an individual’s posts can be directly attributed to his beliefs. Thus, the meaning of posts, and their social and political alignment, reflects the poster’s beliefs, thus allowing her to “know” something about this poster.

He’ll put it all on there, every single day he puts his self to the side, but not … because well he chooses what to post right? So, you know, you can guess about … what he posts, says about his stuff … his beliefs. I know, that about him. What he posts.

Aisha goes on to point out that while it is socially acceptable and even necessary to collect data via online, in order to maximize the efficiency of the process, it not entirely acceptable to be too explicit. She notes that it is “… rude to say you cut someone off for their posts.” Oscar agrees and notes that background checks are essential but not entirely acceptable as a social filter, explaining …“you make the decision to avoid the person based on what they post, but you can’t say that’s why you are doing it, right?”

Among a portion of participants, the practice of checking for shared social and political self-presentation in social media was understood explicitly as a means to ensure the efficient progress to friendship, but implicitly the idea of filtering to avoid conflict and “bad people” was also present. One consequence is the decrease in weak ties within networks (Granovetter, 1973). Without the diversity of ties acting as bridges between social networks, offline networks can develop into filter bubbles in ways similar to online groups (Flaxman et al., 2016). This process is somewhat hampered by the idea that being explicit about filtering is not socially acceptable. However, not everyone agrees that such practices are problematic. For some of the participants, the creation of social filters is entirely the point, as a means to avoid “bad” people.

You gotta know: filtering as a practical matter

Tim also feels that doing background checks is important, like studying for a test. However, unlike the previous group, he and a slight minority of participants are more explicit about the need for social filtering and reject any stigma. He explains:

[You] can really shape a friendship or relationship with the way they say things, or the way they view things may be different than what you, how you feel, and that can cause a big problem, and cause some stress levels to go high, blood pressures of people. You wanna know that stuff, in advance.

Ava is also explicit about the use of social media as a filter. She argues “You gotta know it, right!? You know their attitudes, look what they post and repost? So, if they repost some crazy shit, like haters, you know … you see that too.” Oliver also feels quite strongly about the importance of evaluating a person’s social media representations, as he explains:

… right so if they are really into politics, or social justice or sports or whatever, they post and repost that stuff. You need to be careful, because if someone posts something … nasty … or not even nasty, just not what you believe, what you think, then you need to know. Not to move the relationship forward. If they are against something, or for something … like race or religions you know. You can figure it out, and then … decide if you want to hang with them or not … right?

The decision to use evidence of undesirable or “nasty” posts as a means to filter relationships is common. Individuals address the nature of “nasty” posts, and warn about the need to avoid “haters”. In doing so they are explicitly attributing specific meanings to certain kind of posts through the process of symbolic interaction. These meanings shape the behavior of participants, leading to the practice of filtering. Like Oliver, Anthony approaches background checks as a practical means to avoid conflict. For him, a means to identify social attitudes easily and to make decisions about the relationships he wants to engage in allows him to avoid “bad” people and find and connect with good ones:

You can find out how a person is through posts and stuff—what they post up, what they talk to friends about, who they hang out with. So, you can get a feel of who this person is—are they a good person, bad person.

For Anthony, friendships with people who have shared interests are desirable and friendships without shared interests, beliefs, or morality are less desirable. This is a profoundly practical practice, for Anthony and other participants, a logical way to find information and to make informed decisions about offline social practices.

Kevin explains that for him, other people doing background checks is freeing. This is because this process is two-sided, involving both the individual doing the research and making the decision, as well as the posters’ choices for online identity. He sees posting potentially controversial material as a means to avoid offline social conflict:

You can post things that other people don’t like. Subtweeting. Reposting. People may take it personally and think it’s about them or just hate it if they disagree. They would be upset in person and you’d have to deal with that. [This way] Probably, they just avoid you, which, fine. Whatever.

Kevin frequently posts in ways that violate social norms among his peers. During my time following him online, he regularly posted political and social commentary online that denigrated racial and sexual minorities, Democrats, and other “liberals” as well as women. Often framed as a kind of “dirty joke,” or a rejection of political correctness, Keven claimed that it kept people who were “too sensitive” from being friends with him offline. Like other participants, Keven viewed online social media profiles as being a practical and legitimate tool to navigate social network establishment.

The sense of taboo evidenced by the previous group of respondents is less evident in this second group of participants, and the connection between the meanings attached to postings and the behavior associated with filtering is more explicit. Still, ideological limitations on the degree to which it is appropriate to evaluate someone online still exist. For the final group I identified, the knowledge gained and ability to filter social networks to limit exposure to controversy and political difference is so important that it overwhelms any concern about being considered a “creeper.”

Creepers and filters: safety and morality

Creeping is a more intense form of background checking that involves silently following an individual on one or more social media outlets without posting or commenting and doing expanded research on the person by following their social media friends and family members online. Among the participants who viewed social media profiles as a necessary part of friendship, six participants defined themselves, jokingly or otherwise, as “creepers.” These participants embrace the label of creeper, despite the identity having negative connotations, because the need to filter undesirable individuals out of their social network is simply too important. These participants are willing to continue creeping because they understand it to be an absolutely necessary. One self-defined creeper, Grace, explains:

Creeping is where you just sit online and for no reason you just stalk people, you look at everything. And serious, I will admit it, I creep … you see what people are doing. You read posts from other people. You see what party they went to, their posts about social issues they care about … Or they post about a group, or even an event or news story … you know that about them. You know all about them.

Grace explains that she uses creeping to feel powerful as she explicitly filters social interactions to avoid people who she sees as a bad fit for her interpersonal needs. Jake, a fellow creeper, touches on the importance of being able to gain social power, since he views creeping as a means of controlling his interactions and relationships:

All this personal information, what they like and who they are, and the things that their friends have posted, which is pretty powerful … I feel like it’s a powerful way to judge someone … [To find out] What do they believe? You know to, if someone seems like a good person, or maybe a friend, somebody you could trust, but then you look online and they are a hater, like a neo-Nazi or a confederate flag person, or just you know, a hater, then you can know and not … not be friends with the person. You don’t want to be around the negativity …

The power of having knowledge allows creepers and background checkers to “judge someone” easily. This gives creepers, many of who describe feeling socially awkward or adrift, a kind of control and sense of power that goes beyond the practical approach others describe. Another participant and creeper, Lisa, goes on to describe the sense of power she feels:

… people I haven’t talked to, or people I don’t like—I can view everything they’re doing—every conversation, every news article, every photo, every party they went to, and they will never know. And I think that’s the greatest aspect … being able to see what people are doing, and not having them know. Its powerful, and you decide if you want to get to know them, or not … because if they post bad stuff … personal stuff … just bad politics or whatever.

Participants in this section discuss how having this power, this special knowledge, helps guide them in social situations and shapes choices they make about who to associate with and who to avoid. For these participants the ability to attach meanings to individual posts, and then use those meanings in a way to filter social networks, is about the power to control the nature of their own social network.

For a small number of participants in this study, the intensive and potentially invasive process of creeping is essential to social interaction. Brad explains that for him creeping is necessary to “get at the truth” of people around him, in order to feel safe in social interactions. Like many other participants, Brad associates safety with similarity of attitude, but he also explicitly discusses issue of potential harm in the offline world. He provides an example:

There are cases where someone will say something online that they won’t say offline. So … it’s a way to get to know a person, safely, it’s not risky. Sometimes it’s okay for someone to say things online that they can’t say offline, right, because of the community that they live in. For me, I went to a private Christian high school, and their definition of a Christian was extremely narrow, so the things we talked about online were things we couldn’t talk about offline. We were all … scared, we were afraid that teachers would lower grades if the heard about what we talked about, or what people said, or other kids wouldn’t talk to us. I creeped, because I had to be sure, that the people I was talking to online, social media, that it was safe. Because if my parents or teachers found out, if I said something, that was it. Game over.

Brad provides an in-depth description of the importance of intensive background checks. He acknowledged that in high school, and even now, some people believed him to be a creeper, but he does not care. He explains that he always looks closely at any new acquaintances’ social media accounts before he gets to know them offline. While at one level, Brad’s experience demonstrates the ways in which social media can function as a tool to maintain an individual’s personal or emotional safety, at another level it demonstrates a growing force for social fragmentation. Brad and others can control and limit their social networks, offline and online, to individuals who share traits or beliefs with them.

While it has long been understood that offline social networks and friendship groups are often based on shared attitudes and traits, including race and gender (Cohen, 1977; Henrich et al., 2000), the time it takes for relationships to develop allows for each relationship partner to gain a more comprehensive picture of the complexity of each individual’s personhood. In contrast, the practice of choosing to reject the possibility of social ties, based on the fragmented and partial representation of an individual through their social media posting, has the potential to truncate fulfilling relationships.

Critiquing the filter

The ability of social media to help them better find and keep people in their social network, or explicitly share their views, values, and even, in some cases, politics, is generally seen as a valuable and positive function by the majority of participants in this study. However, some acknowledge the potential risks of using social media as a means to filter friendships. Jane interjects a cautionary note, highlighting the risks for filtering too much based on online behavior:

I guess sometimes you can get the wrong idea because of stuff people say or post … they might not mean it the way as somebody else takes it. So that can become a problem. Like if you go ahead and judge them.

David calls into question the practice of filtering friendships based on social media postings at its core, though he also reinforces the idea that finding people who agree with him, or at least are sympathetic with his views is important:

I got into some bad things because of what I posted about and some other people disagreed with it. So, I can see how people can take things out of context or take things the wrong way because we aren’t face-to-face. Still, it also helps in the same sense because I can connect with people who agree. Who don’t take things personally. They allow me my own opinion or agree.

David is ambivalent about the practice of filtering social networks, in part because of his own social positions. He explains that he holds views about race, gender, and sexuality that are unpopular among his peers, and that at times people view his opinions as offensive, and then avoid contact with him or challenge him in online spaces. He explains that he feels persecuted by people who take his comments or news posts “out of context,” but at the same time, it allows him to engage in filtering himself, by avoiding or rejecting people who disagree with or challenge him online, in offline contexts.

One additional critique related to the use of social media profiles as filters that emerged in the research is related to the structure of the technology. Some social media sites allow for the restriction of larger and smaller groups of contacts, allowing the participant to customize and control the visibility of posts and the strategic nature of technological visibility. Brad, who discusses his struggles in high school and with family above, provides us with an example of this issue:

A lot of my family and some people I knew from high school they say things, post things you wouldn’t believe. About black people, or gays or in support of groups. Bad groups. I know, I can’t talk to them, so they go on a list, and I mostly just creep or post sports scores whatever.

The participants in this study understand the strategic and partial nature of information portrayed online very well. Tyrone notes that:

Anybody can lie and say they’re interested in something or believe one way through social media when that’s actually not the case. To many people think … I saw it on the internet … this is true.

However, the strategy involved in constructing an online profile is part of the justification for their use as a filter. Participants know that when they go online to look at social media profiles they are getting carefully curated and hand-picked surface versions of the person and their beliefs. As Ellie explains:

I think it becomes important because you can tell a lot from a person by what they tag and what they choose to post and repost. They choose what to be connected too, choose it on purpose, and choose to make it visible to everyone, right? That is a statement. You can get an idea, so it probably is pretty important.

While several of the participants in this study acknowledged the problematic nature of making decisions about offline friendships based on the partial and fragmented self-representation presented in social media, the majority still view it as a desirable and even necessary social practice. In addition, even those who identify the problematic nature of the practice tend to engage in the behavior at least occasionally. However, this social practice has potential for complicating an already difficult social arena. The creation of social networks is one of the most important and far-reaching aspects of human socialization (Granovetter, 1973). Networks shape everything from emotional support to career success, and increasingly they also shape individual perceptions of truth and reality, which has the possibility for long social consequences.