Abstract
This study explores how norms on social network sites evolve over time and how violations of these norms impact individuals’ self-presentational and relationship goals. Employing Expectancy Violations Theory (Burgoon, 1978) as a guiding framework, results from a series of focus groups suggest that both the content of the violation and the users’ relationship to the violator impact how individuals react to negative violations. Specifically, acquaintances who engage in minor negative violations are ignored or hidden, while larger infractions (that could negatively impact the individual) result in deletion of the offending content and – in extreme cases – termination of the Facebook friendship. Negative violations from close friends (that did not impact participants’ goals) resulted in confrontations, while similar violations from acquaintances were often ignored by participants in an effort to ‘keep the peace.’ Furthermore, positive violations were more likely to arise from acquaintances than close friends.
Teenagers and young adults are among the most active users of social network sites (SNSs) such as Facebook (Lenhart et al., 2010). While users do not appear to be highly concerned about potential negative repercussions of sharing so much content with a large audience (Barnes, 2006; Dwyer et al., 2007), recent media accounts have provided numerous reasons for why they should be concerned. A 2008 study of the top 500 US colleges by Kaplan found that 10% of admissions offices checked applicants’ SNS profiles, and 38% of those saw information that negatively impacted the applicants’ prospects for admission (Hechinger, 2008). More recently, a college student was cited for underage drinking after campus police found pictures on Facebook of the student holding a beer (Lang, 2009).
Despite these risks, SNSs have grown tremendously in a short period of time, and sites such as Facebook now support hundreds of millions of users. Because of the speed at which these sites have evolved, however, an established set of social norms guiding users’ behavior has been slow to follow. Furthermore, when behavioral norms are ambiguous, it becomes more difficult to both establish a formal set of norms and to respond to perceived norm violations. Online etiquette – also known as netiquette – is difficult to define in these spaces because both the environment in which users interact and the composition of their social network are constantly evolving.
With so many groups interacting on Facebook, there is a huge potential for norm variance, and thus norm violations are likely to occur. Expectancy Violations Theory (EVT; Burgoon, 1978, 1993) explains how people respond to unexpected behaviors (i.e. expectancy violations) and posits that individuals pay more attention to communication events when expectations are violated. Expectancy violations elicit either positive or negative reactions from receivers, depending on the situation and the relationship between the individuals involved. On SNSs, such violations may have a direct effect on users’ impression management goals, and thus reactions to violations may be unique due to the very public context.
The concepts of norm and expectancy violations and the public nature of SNSs raise a number of questions. First, in order to understand violations, we must identify the norms that guide individuals’ use of SNSs. With this information, the next question becomes what actions do users perceive as violations? EVT has explored evaluative and nonverbal reactions to violations, but it is also important to know how they handle violations on SNSs, which are characterized by verbal (i.e. text-based) interactions. Another feature that characterizes SNSs is that people use them to interact with a wide range of connections, including close friends, college classmates, family members, and acquaintances. Therefore, it is important to consider how violations – and reactions to violations – vary based on the relationship. Employing a focus group methodology, the present research sheds new light on these questions by focusing on the role of norms in college students’ use of Facebook and their responses to various types of violations.
Social norms online
Conceptualizations of social norms vary across and within fields. Broadly conceived, social norms can be considered rules that guide behavior, the ‘customs, traditions, standards, rules, values, fashions and all other criteria of conduct which are standardized as a consequence of the contact of individuals’ (Sherif, 1936: 3). In other words, social norms are a framework through which people determine what behaviors are acceptable and unacceptable. Researchers have identified several different types of norms. Of particular interest to the present research is the concept of implicit norms, which are defined as norms that are not written down but understood by the group in general (Burnett and Bonnici, 2003).
Development of and agreement upon a set of norms is a critical component of group formation because it provides guidance to the boundaries of acceptable behavior for that specific group. Group norms are socially constructed over time, and conformity to a group’s norms should increase the longer the group interacts (Postmes et al., 2000). New members of a group may struggle to learn group norms initially, especially when those norms are not codified.
The social norms regulating behavior in an offline context tend to be well established, with many ingrained into children from an early age. In an online context, however, such norms often evolve with the technology. As the userbase increases, so too does the need for rules of conduct when multiple groups begin interacting (Preece, 2004). For example, Burnett (2009) examined interaction between three groups of users on an online music forum and found that because the perceptions and norms of use were different for each group, there was often disagreement over decisions made regarding the forum.
Research has demonstrated that norms that guide offline behavior hold in many online environments, even when users remain anonymous. In research on the soap opera discussion group rec.arts.tv.soaps, Baym (1998) found that a norm of politeness guided interactions among members, and violations of this norm led to quick, if gentle, reprimands. Likewise, research on The Sims Online found that norms of interaction reflected offline expectations of politeness (Martey and Stromer-Galley, 2007). If offline norms carry into online environments characterized by anonymity, they may be an even stronger force on sites where user identities are visible, such as SNSs; however, a number of other factors should also be considered.
The role of audience on SNSs – one of four properties that boyd (2008) notes differentiate mediated and unmediated publics – may help in understanding why norms regulating interaction on a site such as Facebook should differ from norms regulating offline interaction. While users can control their audience to some degree through privacy settings, norms of ‘friendship’ on the site also affect the makeup of the audience. Boyd (2006) notes that Friends (connections formally articulated in the system) on SNSs may differ greatly from friends in a more traditional (i.e. offline) space; the affordances of the technology allow users to easily maintain a much larger friendship network than would be possible through face-to-face interaction, and thus many users’ Friends lists contain people with whom they would not share information offline.
Findings from Facebook server-level data suggest that norms of sharing content are learned from observing the habits of Friends on the site (i.e. implicit norms); in other words, users contribute more content if they see Friends contributing content (Burke et al., 2009). However, norms of sharing content may also depend on other factors; for example, women tend to post more photos to Facebook than men, and older users (i.e. age 23 or older) tend to post fewer photos than their younger counterparts (Strano, 2008), suggesting that demographic variables impact the norms of use. These differences in perceptions of appropriate sharing behaviors may lead to violations of some groups’ norms when these groups are intermingled within an individual’s Friend network on the site. First, however, it is important to understand norms of use on SNSs and how these norms evolve over time, thus aiding in the interpretation of possible violations of these norms. Thus:
RQ1: What are the overarching norms of use for college-aged SNS users?
RQ2: How do users’ norms of behavior change over time, and what motivates this change?
Expectancy Violations Theory
Early social norms research posited that when norms are violated, some form of punishment must follow (Homans, 1950); however, more recent research suggests that for violations of norms established for a pair of individuals, context plays a role in the evaluation of the violation. According to EVT (Burgoon, 1978), when expectations are violated, the violation is judged as either positive or negative. When a friend surprises you with a gift, it is an expectancy violation, but one that is typically not negative for the recipient. One way the valence of an expectancy violation can be determined is based on the communicator valence: a violation carried out by a person the target perceives in a negative way is more likely to be perceived as a negative action than a violation carried out by a person the target perceives positively.
For this study, violations were divided into two types: norm violations and expectancy violations. Norm violations refer to violation of common rules within the context of Facebook, such as etiquette regarding acceptable posting habits on Facebook. Expectancy violations will be defined as behaviors that are unexpected based on the relationship an individual has with a communication partner, but that do not necessarily violate Facebook etiquette (e.g. receiving a message from an acquaintance that has been non-communicative for an extended period of time). Expectancy violations can be positively or negatively valenced, such as one Friend surprising another with a Facebook gift (positive) or a Friend posting an embarrassing story in a public forum (negative). Norm violations, however, defined as violations of etiquette, are always negatively valenced.
EVT states that a violation increases alertness and attention paid to the interaction and heightens attention to the characteristics of the communicator, the relational implications, and the meaning behind the violation. The overall attitude toward the violator is much more influential to the evaluation process when the violation is ambiguous but can play a role in any expectancy (norm) violation. For example, Burgoon and Hale (1988) found evidence that the valence of the communication partner (i.e. if the partner is perceived positively or negatively) only affected the evaluation of the target when the violation was ambiguous, thus demonstrating that communication partner valence has an effect on evaluations only under certain circumstances. Burgoon et al. (1992) found that the same behavior affected evaluations of communication partners in opposite ways for high and low valence partners – touch increased perceived task orientation for high-valence partners and decreased perceived task orientation for low-valence partners.
While EVT was developed in order to explain nonverbal behaviors such as personal distance while speaking (Burgoon, 1978), other contexts have also been explored. Burgoon and LePoire (1993) considered the violation of expectations regarding pleasantness and found that when expectancies were disconfirmed, they impacted post-interaction evaluations relative to the condition in which there was no expectancy disconfirmation but the same behavior was present. In other words, communication partners behaving pleasantly were evaluated more positively if their partner expected them to be unpleasant than if they expected them to be pleasant. Before exploring Facebook users’ reaction to violations, we must establish what constitutes violations on the site. Perceived violations on the site should differ somewhat from offline violations due to the decreased presence of nonverbal cues. Furthermore, we need to understand which types of violations are present in the online context before exploring reactions to these violations. Therefore:
RQ3: What norm and expectancy violations do college students observe on SNSs?
Responses to norm and expectancy violations
Expectancy violations have been demonstrated to operate differently depending upon the individual who violates expectations and the type of violation. EVT predicts that positively valenced individuals who violate expectations by decreasing involvement in an interaction will be met with compensation (Burgoon, 1978, 1993); in other words, interaction partners will increase their involvement in order to compensate for this decrease in involvement by their partner. For example, Bachman and Guerrero (2006) found that when romantic couples fought, more positive relationships resulted in more constructive communication and less destructive communication. However, other studies have failed to support this prediction. LePoire and Burgoon (1994) found that negative violations by positively valenced interaction partners were met with reciprocation; despite a previous positive relationship between interactants, individuals decreased their involvement in the interaction when their partner decreased involvement in the interaction.
To further explore reactions to violations, researchers have considered differences between positive and negative violations. Lannutti and Camero (2007) found that flirtatious behavior from physically attractive individuals was evaluated more positively than flirtatious behavior from physically unattractive individuals. Honeycutt (1991) found that positive expectancy violations resulted in more positive evaluations of an interaction partner than fulfilled positive expectations, while Koermer and Petelle (1991) had similar findings when looking at students’ evaluations of instructors. Floyd and Voloudakis (1999) found that positive expectancy violations resulted in more positive affect than before the violation occurred, and negative expectancy violations resulted in more negative affect than before the violation occurred.
While the majority of research on expectancy violations has focused on nonverbal behaviors, Rycyna et al. (2009) considered the impact of verbal deception (a negative expectancy violation) about participants’ weight. They found that minor deceptions about one’s weight were expected and therefore deemed to be acceptable, while more serious deceptions resulted in negative evaluations of the target. Research on deception in online dating had similar findings regarding daters’ height and weight (Toma et al., 2008) and indicated that, in some environments, a norm of exaggeration has developed such that ‘small lies’ are not just accepted, but expected (Fiore and Donath, 2004).
The modality of both initial meetings and subsequent interactions (online vs. offline) may impact individuals’ reactions to expectancy violations. For example, Ramirez and Wang (2008) found that partners who interacted for a prolonged period online before interacting offline rated each other lower than those who only interacted online or switched to offline interaction after a short period of time; the authors posited that after developing positive expectations over time, the face-to-face interaction violated expectations in a negative way. Walther et al. (2001) had similar findings when showing participants a picture of their partner after an extended period of text-only interaction.
When norms are violated in an online environment, a number of possible sanctions are available to both the group and the individual user. Discussion forums typically provide a set of explicit norms outlining acceptable behavior on the site; when users violate group norms, they are directed to the appropriate document (Burnett and Bonnici, 2003). With repeated violations, more serious sanctions may follow. For example, Baym (1998) found that members of one Usenet community were expected to use their real names or nicknames with identifying information; when this norm was violated, members typically requested this information before replying to a post.
On SNSs, general explicit norms exist regarding what the site defines as unacceptable content, and procedures exist to respond to norm violations, including account suspension. However, norms may also vary greatly across groups or individual friend networks, so milder sanctions – such as ‘hiding’ a user on Facebook, which removes their posts from the News Feed – enable individuals to better regulate content across hundreds of Friends. EVT predicts that reactions to violations will vary depending upon the interaction partner and the valence of the violation. Because SNSs center on interactions and those interactions occur with people across the entire spectrum of relationships, we would expect that individuals experience a variety of violations from users ranging from mildly (e.g. acquaintances) to extremely positively (e.g. close friends) valenced. Furthermore, research has demonstrated that different types of violations result in a range of attitudinal reactions (Bachman and Guerrero, 2006; Rycyna et al., 2009; Toma et al., 2008). To determine how responses vary based on the violation and the violator, we ask:
RQ4: How do college student SNS users react to norm and expectancy violations from differently valenced communication partners?
Methodology
Facebook is, first and foremost, a social medium; thus, it was important to employ a methodology that focused on interaction between users in order to draw out perceptions of norms. Focus groups enable researchers to derive the attitudes and perceptions related to a specific topic by observing interactions between users (Krueger and Casey, 2008). As discussed above, norms do not form in isolation; rather, they develop through day-to-day interactions between group members. While considering the research questions posed above, focus groups allowed us to fully explore the norms of Facebook use in a social, more natural setting than other research designs.
Participants were recruited from undergraduate classes at a large Midwestern university in exchange for extra credit. Five focus groups, each lasting approximately 90 minutes, were conducted in June 2010. Groups ranged in size from three to seven, and all but one group had at least five participants. Of the 26 participants, the majority were male (53.8%) and White (61.5%), and nearly all were entering their senior year at the university. After the completion of five sessions, saturation had been achieved and the researchers concluded that no further inquiry was necessary.
Groups were held on campus, video and audio recorded, and transcribed by the researchers. All names have been changed to protect the privacy of participants. Topics of discussion centered on the underlying theoretical concepts driving the research questions and included questions meant to derive responses about group norms and responses to violations from different types of Facebook friends. Following transcription, the data were analyzed using situational analysis (Krueger and Casey, 2008). In this process, we analyzed the transcripts in an iterative process in order to identify themes related to the research questions. Following Krueger and Casey (2008), we focused on seven factors in each of the transcripts: words; context; frequency, intensity, and specificity of comments; internal consistency; and ‘big ideas.’ The results synthesize the overarching themes that emerged from this process.
Results
Facebook norms
RQ1 addressed norms of use on Facebook. During this section of the focus groups, moderators prompted participants by asking them to talk about the ‘unspoken rules that guide your Facebook use’ in order to highlight implicit norms. Individuals spoke of norms regarding how much information they shared on their profiles and the norms of Friending on the site. In addition, participants discussed the types of content they shared publicly versus content they shared only privately and the norms that guided sharing information about their Friends (such as photo tagging). All norms discussed were implicit, and thus the learning process for these norms is discussed first, followed by a discussion of the norms of use.
Participants all seemed to indicate that the norms of use within Facebook are implicit (Kemper, 1968; Sherif and Sherif, 1964) and are therefore norms that are not written anywhere, despite various etiquette guides available online (e.g. Salam, 2007). Based on the consensus between the five groups, though, these norms are known among users. This finding is further validated by results from Cotter and Wanzer (2008), who reported similar norms to those described here.
Participants stated they had learned how to conduct themselves on Facebook by observing how others behaved as well as by applying the norms of everyday offline life (consistent with Martey and Stromer-Galley, 2007). For example, Paul compared Facebook to a family wedding, where various groups of people such as friends, family, and colleagues may be in attendance; therefore, he said, you need to take these varying groups into consideration when interacting to make sure that whatever you say is appropriate for everyone.
Almost all participants said they felt it was rude to ignore or deny a Friend request from someone they had met in person; for example, Jenna said, ‘Although I would accept acquaintances who I knew of but didn’t really know and now I’m starting to delete them, I just felt like it was mean to deny them.’ To further demonstrate how ingrained this norm is, Brian stated, ‘You don’t want to delete a person unless you really don’t like them.’ Other literature has discussed this phenomenon of liberal Friending practices (e.g. Donath and boyd, 2004; Tong et al., 2008), while some popular media outlets have developed informal guidelines for Friending (e.g. Salam, 2007). Participants also indicated they were uncomfortable accepting a Friend request from somebody they had not met, although many indicated they had done this when they first joined Facebook. These comments suggest that individuals are at least somewhat cautious about who they Friend on Facebook, and that caution has increased over time.
Norms of communication on Facebook differed by whether the communication occurred through a public or a private channel. Participants reported utilizing Facebook’s two primary features for public interaction – the Wall and status updates – to share information with members of their network, including videos, jokes, birthday wishes, and short messages to an individual. Private channels (i.e. Chat and Messages) were more likely to be chosen when participants wanted to start a discussion thread for a smaller group, organize an event, or give out personal information such as a home address or phone number. Participants also said it was courteous to utilize private messaging features for longer messages to an individual, or when they expected to exchange multiple messages with a contact, to avoid clogging Friends’ News Feeds.
Facebook is the most popular picture-sharing website in the world, and participants’ comments revealed norms related to posting and tagging pictures on the site. Most participants said they typically posted pictures of themselves and their friends from events (both formal and informal) or vacations, and some noted that their Facebook photos offered a skewed self-presentation of their social lives. For example, Lindsay commented on the types of pictures present on Facebook:
I think there’s probably a disproportionate amount of me, of pictures of me at parties … you know people aren’t gonna take pictures if you’re just hanging out and even though that’s much more my speed … if someone was only to look at my Facebook profile to get an impression of me, they would think that I go out way more than I do.
This comment carries a deeper meaning than merely describing norms of photo sharing on Facebook. When participants were asked how they would form an impression of a new interaction partner, most participants stated they would look at pictures to get a feel for the person. Therefore, it may be that when attempting to engender positive impressions through the Facebook profile, users need to be cognizant of the underlying meaning their Facebook photos convey and selectively choose which ones to make publicly visible.
Regarding expectancies for friends’ behavior, most participants said their friends were very considerate of their wishes and impression management goals. For example, John said, ‘My friends are considerate of me and I’m considerate of them, so I haven’t had any trouble with [Friends posting inappropriate content to my page].’ This statement, and several similar quotations, indicates a norm of consideration or privacy among friends. In addition, female participants reported tighter control over their online persona than their male counterparts, who were more relaxed in the types of information they shared and the people with whom they shared it. This finding is in line with previous research by Pempek et al. (2009), who found that females were more likely than males to untag photos of themselves. Based on these accumulated findings, it would seem that females are more concerned than males with self-presentation goals on Facebook.
Facebook norm evolution
RQ2 addressed how participants’ behavioral norms changed over time. The concept of multiple audiences on Facebook is more salient now than previously; for example, Lampe et al. (2008) found increases in college students’ perceptions of the number of audiences viewing their profile between 2006 and 2008, most likely due to the site’s reduced restrictions on membership and its increasing popularity among non-college student populations. Focus group participants’ comments suggest they became more skilled at navigating Facebook as time went on and, subsequently, adapted some of their Friending and content-sharing practices.
To demonstrate this learning process, all participants indicated that the content of their profiles had changed over time. Many stated that they increased their privacy settings once the media began publishing stories about negative consequences resulting from Facebook posts. In addition, many individuals indicated that they reduced the amount of personal information in their profile (e.g. contact information), but increased the number of pictures available over time. Others stated that while they had not taken the profile fields seriously when they first began using Facebook, they later returned and filled in more information once they better understood the site. Some participants said their posting habits became more relaxed after turning 21, because it was no longer a problem if Friends tagged them in pictures involving alcohol. It is important to note that the present study only included upperclassmen and women, many of whom were particularly aware of an impending job hunt, which may have impacted the degree of control they exerted in their self-presentations on the site. In previous work by DiMicco and Millen (2007), older users were more likely to have profiles aimed at impressing employers than younger users, and profiles were typified by less personal information, more job-related information on the profile, and more professional profile pictures.
Norm and expectancy violations on Facebook
RQ3 asked about observed violations on Facebook. The most frequently stated norm violation from participants was too many status updates, which many participants saw as both unnecessary and annoying. For example, John said:
I could change my status update when I get up, and then I could change my status update when I get in the shower, and then I could change my status update when I get out of the shower, and then after I get dressed, but I don’t think people care.
This comment demonstrates a desire to conform to the norms of the site and be considerate of other members, and to have others demonstrate the same respect by being conservative with status updates.
Another norm violation mentioned by participants was overly emotional status updates or Wall posts. Participants did not feel Facebook – as a public venue – was an appropriate place to air private or sensitive grievances. As Lana stated, ‘Some people will talk way too much about their feelings and I’m like, “you should probably keep that private. I don’t need to know every detail about what’s going on in your life.”’ Some participants stated that they believed many of these people posted emotionally laden status updates in order to receive support and comments from their friends, but participants wished that those Friends would look for comfort outside of Facebook.
Heated interactions, fights, and name calling through Facebook’s public features was another common norm violation discussed; as with above, participants voiced disapproval of conducting private discussions where everyone – including friends, family, colleagues, and potentially non-Friends – could view the entire interaction. John commented, ‘If I’m gonna scream at you or try to slam you or run you through the mud, I’m gonna do it in person or by other means, not through Facebook.’ Others expressed a similar desire to keep private feuds and discussions out of the public eye. This norm seemed to be out of a desire to be considerate of other users as well as the individual with whom the participant had a disagreement.
Although participants often spoke of ‘pet peeves’ when asked about violations (previously discussed violations are defined as pet peeves because they are bothersome to participants but have no direct effect on them), other comments suggest that posting or tagging pictures of them (the participants) when they were intoxicated, messy, or not looking their best was viewed as inappropriate. This included any type of Wall posting or picture that would create a negative impression for their friends, family, or prospective employers. For example, Lana spoke of an incident with a friend of hers:
One of my friends tagged a picture of another friend … and she was hammered and her skirt was up and her underwear was showing in a not-very-flattering way at all, and she tagged her. I’d be pissed if someone tagged a picture like that of me.
This and similar comments suggest that posts or pictures that might reflect negatively on an individual should not be tagged on Facebook, another implicit norm. Many female participants, in particular, stated that they untagged themselves from photos in which they were drinking or participating in embarrassing activities.
Because norm violations are, by definition, negative, there are no positive norm violations to be discussed. Regarding positive expectancy violations, individuals stated they had sometimes been surprised to find out that an acquaintance or former friend had similar interests to them or had participated in a cause that the participant thought was very worthwhile. This often led to conversations and rekindling of the friendship. Similarly, participants indicated they had received messages from or been tagged in pictures by childhood friends, and this expectancy violation had also resulted in a reunion. One participant related a story of reconnecting with an old friend through Facebook and, as a result, going to Las Vegas to visit her. While these examples of expectancy violations reflect making distant or weak ties stronger, some participants commented on expectancy violations involving closer relationships. For example, one participant told a story of a positive message received from a family member who was not normally particularly supportive. In this case, it may be that interactants are unable or unwilling to express themselves in person but find that the mediated environment of Facebook allows them to express themselves more accurately.
Reactions to expectancy violations
RQ4 addressed how individuals dealt with expectancy violations and norm violations. The above examples support EVT’s prediction that the same action will be evaluated differently based on relationship type (Burgoon, 1978; Floyd and Voloudakis, 1999; Honeycutt, 1991; Koermer and Petelle, 1991). Tagging a photo or sending a supportive message is only extraordinary within the context of fairly weak ties (or, in some cases, non-communicative individuals). For close, communicative friends, this behavior would merely be the fulfillment of positive expectations, but for acquaintances it is an expectancy violation that may positively impact the nature of a friendship (Burgoon, 1978, 1993).
Reactions to positive violations are in line with EVT predictions, but negative norm violations are somewhat contrary to those predicted. EVT predicts that when negatively valenced communicators negatively violate expectations, this behavior is reciprocated, whereas negative violations from positively valenced communicators would result in compensation (Burgoon, 1978). However, these results suggest that for mere acquaintances a negative norm violation can result in different responses depending upon the nature of the violation. Serious violations that have negative repercussions for the target are often met with deletions. For example, Joey related a story of a public post that could have severely damaged his friendships and image: ‘One guy thought he was being funny and posted on my Wall. He was trying to make some joke, but in it he was totally calling me out for being racist … so I deleted that, then deleted him.’ The Wall post clearly threatened Joey’s self-presentational goals and, because the relationship was mainly neutral to begin with, it was low-cost to remove this individual from his Friends list.
Other, less serious violations of Facebook norms, such as posting status updates too often or participating in too many applications, often resulted in participants ‘hiding’ these individuals from the News Feed, but not deleting them from their Friends list. This action allowed participants to continue to evaluate the norm violator in a relatively neutral way with little cost to the participants themselves. This action did, however, reduce the likelihood that these ‘hidden’ individuals would provide participants with any information that would result in a positive expectancy violation, since these individuals are likely to remain hidden and forgotten and would need to engage in directed communication (Chat, private message, Wall post, comment) in order for the participant to see the post.
Participants were also asked about negative expectancy violations regarding friendship or working norms that did not reflect negatively on the participant but that did impact their lives. The example given by researchers for this type of violation was not finishing group work on time because of partying, a discovery participants would have made by the group member’s Facebook photos or Wall postings. For acquaintances, this violation was dealt with extremely passively. As Jenna said, ‘I would keep my eye on it; if they keep flaking out, then you might need to say something, but otherwise, unnecessary confrontation is probably not the best option.’ Some participants even indicated they would do the work for an acquaintance in order to avoid a confrontation and keep the peace, while this was never mentioned for close friends. This coping strategy reflects an extreme sort of compensation, possibly out of a desire to ignore the negative information because there is no valuable relationship to save; therefore, the best option for these participants was to keep the peace and ensure that their goals (a good grade in the class) were met.
In contrast, this same expectancy violation from a close friend resulted in completely different responses from participants. For example, Luc said, ‘I would scream at that person, if they would do that thing to me, I would walk to their house or their dorm and be like, “you’re gonna sit there and do your work. You’re not gonna go anywhere.”’ This indicates a reaction similar to reciprocation (Burgoon, 1978, 1993), likely based on a desire to solve the problem before it damages a rewarding friendship. Furthermore, participants indicated that if they knew the person better, they would be better able to determine if this was a habit or an isolated incident, and thus their reaction would depend on that. Lindsay commented,
You know it would probably depend on the person. If I knew that they were someone that consistently did that sort of thing then maybe, but if you know typically they’re a dependable person then, like I said, I think it would be a judgment call.
To summarize the results regarding this particular expectancy violation, a confrontation is much more likely for close friends than for acquaintances, and compensation is much more likely for acquaintances, a finding in contrast to the typical EVT predictions.
A confrontation of some sort was also likely to occur between positively valenced expectancy violators and participants for norm violations that threatened the participants’ ability to control their impressions online (such as the tagging of inappropriate pictures). This type of violation does not impact work-oriented goals (e.g. completion of a class project), but rather self-presentational goals. When pictures were tagged that the participants viewed as inappropriate, they indicated that they would not only de-tag the picture but would also ask the individual to remove the picture from Facebook. Although participants stated that their true friends were willing to take the pictures down once asked, the few incidences where this was not the case resulted in a further negative expectancy violation that was much more likely to affect the friendship than the initial tagging. As discussed above, this same violation from a more neutral target could result in defriending. Participants did not discuss how they reacted to more minor norm violations from their friends, likely because this was much more common among their acquaintances.
Discussion
The present findings suggest that, in line with Richardson’s (2010) findings, social etiquette is alive and well in the SNS world. Implicit norms are in place to guide what information is shared about one’s Facebook Friends in addition to more general norms that ensure that Friends’ News Feeds are not clogged by overzealous Facebook users. These general norms have been made explicit in various informal netiquette guides, but the more important, impression management norms have yet to be published. By understanding these norms better, it is possible to make these norms explicit and reduce the blunders that have become common in media reports.
These norms have evolved as more and more groups of individuals join Facebook and as the users themselves move through life. As demonstrated in Paul’s wedding example, the wide variety of individuals on Facebook means that public postings need to be tailored to be acceptable for a wide variety of individuals. Recent research by Marwick and boyd (2010) examined how Twitter users manage multiple audiences on the site, but little research has considered the impact that maintaining an increasingly diverse Facebook Friend network has on site norms and self-presentation strategies. As expected, the life stage of individuals in this study impacted the norms discussed: because most participants were of legal drinking age and also on (or soon-to-be-on) the job market, the norms of use were different for participants than they were when they first began utilizing Facebook. This finding points to the need to expand studies of Facebook behaviors to populations beyond college undergrads, as users in different life stages (e.g. high school students, parents, older adults) may be using the site for various purposes and have different sets of norms guiding their behaviors.
Results suggest that norm violations can occur based on simply annoying behaviors or posts that can negatively impact an individual’s self-presentational goals. Expectancy violations, however, were mostly positive. Positive expectancy violations were noted most frequently among acquaintances, likely because positive actions from positively valenced individuals are likely to be a fulfillment of expectations, rather than an expectancy violation (Floyd and Voloudakis, 1999; Honeycutt, 1991; Koermer and Petelle, 1991). In addition, negative violations of any kind tend to result in a confrontation of some sort for positively valenced figures, a trend that seems to be a form of reciprocation aimed at ensuring that the relationship continues by ‘clearing the air.’ In contrast, neutral or negatively valenced figures engaging in negatively valenced violations are either ignored (if the infraction is small enough) or removed from the individual’s Friends list. This provides an interesting extension of EVT in the area of ‘networked publics,’ a term boyd (2008) coined to describe online environments such as SNSs that are characterized by the persistence, replicability, and searchability of information, as well as ‘invisible audiences’ which describes the ability to view content and interaction without being ‘present.’ Because of the wide variety of violations and the unique context, a wider range of violation types and reactions become apparent. Further research should explore these violations to determine if this is a consistent pattern. Our research suggests that compensation/reciprocation reactions may be a function of more than just communicator and violation valence, but also the type of violation. To further explore this relationship, an experiment could manipulate communicator valence and violation type (e.g. usage-based norm, self-presentation norm, or expectancy).
In addition, this study provides more insight into the phenomenon of negatively valenced norm violations for positive figures and provides an interesting perspective on the findings of Bachman and Guerrero (2006) regarding fighting couples. In both cases, it seems that a confrontation was beneficial to the relationship because the confrontation decreased the likelihood that the individual would violate the norm again and ensured that the relationship could continue to be a mutually beneficial one.
Limitations
Because this was an exploratory study, the research provided a large amount of in-depth data from the participants in the focus groups, but generalizability is lowered because results are based on a small subset of a population – juniors and seniors at a US university, nearly all of whom were over 21. Future research should determine if these results hold for younger college students, who are not about to enter the job market and who are more at risk for negative consequences related to underage drinking. Furthermore, research should be conducted to confirm the current findings utilizing quantitative methods in order to increase external validity.
Conclusion
This study has explored the norms of Facebook use and the consequences of norm and expectancy violations on Facebook. Overall, our results suggest that EVT may operate somewhat differently in Facebook: although there are few negatively valenced individuals in this context, acquaintances (who are probably fairly neutral in evaluations) seemed to elicit a wide variety of coping mechanisms when they negatively violated expectations (many passive), while positively valenced Friends tended to be met with confrontation if they negatively violated norms or expectations.
The norm of passivity among acquaintances on Facebook seems to have impacted the response to expectancy violations. When violations occurred that negatively impacted a participant’s ability to achieve self-presentational goals, results suggested that reciprocity by withdrawal occurred for acquaintances, but reciprocation by a conversation or confrontation occurred for friends. When violations occurred that negatively impacted a participant’s work-oriented goals, however, acquaintances received a sort of extreme form of compensation, while friends were met with a confrontational form of reciprocation aimed at clearing the air. Overall, these results suggest that the relationship between expectancy violation type and valence of the communication partner may be more complex than previously thought, and responses to a violation may be a product of the context and the goals that are threatened by such a violation. By further exploring the nature of these violations, researchers can begin to understand the contrasting findings of LePoire and Burgoon (1994) and other EVT researchers, as well as identify how expectancy violations function in a somewhat different – but frequently used – environment than the original theory described.
Footnotes
Author biographies
Caitlin McLaughlin is a doctoral candidate at Michigan State University, where she studies the effects of new communication technologies, including social network sites, on advertising outcomes and people’s communication patterns.
Jessica Vitak is a doctoral candidate in the Telecommunication, Information Studies, and Media department at Michigan State University. Her research focuses on the social impacts of new communication technologies, specifically on outcomes associated with use of social network sites.