Abstract

Keywords: social network sites, self-disclosure, privacy settings, adults, adolescents

doi: 10.5817/CP2012-1-3

Introduction

The evolution of social media into a mainstream phenomenon constitutes one of the most important evolutions in the short history of the Internet thus far. For many Internet users, social network sites (SNS) have emerged as a preferred way of communicating with others. The inherent social function of SNS is often associated with positive outcomes. The use of SNS offers opportunities for consolidating ties in networks of close friends and families, for offering and obtaining emotional support or, in the terms of Putnam (2000), for developing bonding social capital. Further, by opening their profiles to a broader audience and by being searchable and reachable to users outside their direct networks, users can enhance their bridging social capital or weak ties (e.g. Putnam, 2000; Ellison, Steinfield, & Lampe, 2011). These expanded opportunities for people to use SNS to maintain and extend their social networks have been associated with elevated levels of self-esteem and experienced well-being (Pelling & White, 2009). Self-esteem is particularly likely to be fostered when SNS users provide each other with positive feedback on posts made on their walls (Valkenburg, Peter & Schouten, 2006).

Privacy at stake

Despite the tremendous opportunities that SNS offer for connecting and bonding, privacy issues emerging from their use are currently receiving increasing media attention (e.g. Bilton, 2011; Helft, 2011). These privacy issues have emerged largely because many SNS platforms encourage their members to disclose lots of personal information on their personalised pages (‘profiles’). Moreover, SNS generate their revenues largely through business models that include the processing of this disclosed personal information for commercial purposes (e.g. micro-targeted advertising) (Thelwall, 2009).

Privacy and other issues related to SNS have become a much-publicised and controversial topic in the media. For adolescent SNS users, personal information disclosed on SNS can be used for purposes of identity theft, stalking, cyberbullying or sexual harassment by online predators. Even adult SNS users can be affected in their sense of online privacy, as illustrated by media reports of cases involving intrusions into the family or professional lives of adult users. Examples include burglaries in the homes of people who had made imprudent holiday announcements (e.g. Britten, 2010) or employees who have lost their jobs because they had posted inconsiderate comments about their employers (e.g. Smith & Kanalley, 2011).

The increased attention in the popular press to risks associated with the use of SNS by adults and adolescents has also increased attention from academia. Increasing research is being devoted to SNS use and the risks associated with potential misuse of the information divulged on the online profiles of both adolescents and adults (e.g. Livingstone, Haddon, Görzig & Olafsson, 2011; Palfrey, 2008; Vandoninck, d’Haenens, De Cock, & Donoso, 2011; Ybarra & Mitchell, 2008).

In the context of this study, the concept of information privacy is important. According to Westin (1967, p. 7) this concept refers to ‘the claim of individuals, groups, or institutions to determine for themselves when, how, and to what extent information about themselves is communicated to others’. In addition to the initial decisions of users regarding whether to divulge personal information, information privacy includes also their possibilities to control the extent to which disclosed personal information is made public.

In this regard, users can implement two main strategies in order to manage their privacy on SNS.

First, users can limit the amount and type of information they disclose about themselves online. In the context of SNS, self-disclosure has been defined as ‘the amount of information shared on user’s profile as well as in the process of the communication with others’ (Krasnova & Veltri, 2010, p. 1). Upon opening accounts on SNS or when updating their profile information, users can decide whether they are willing to enter specific types of personal information in the entry fields provided. It is important to note, however, that SNS are designed to encourage users to disclose and update personal information and share their experiences and, by doing so, adding to the site’s vibrancy.

In addition to decisions regarding data disclosure on SNS, Internet users can also apply fine-grained privacy settings to determine who will have access to these data, ranging from a broad audience (e.g. ‘all users’) to a narrower group of users (e.g. ‘friends and their friends’, ‘only friends’, ‘restricted to myself’ or even specific users).

Aims

This study has two objectives. First, we aim to provide further insight into the ways in which adolescents and adults manage their personal information on SNS and how they apply privacy settings to the information they are willing to disclose. Second, we aim to compare an adult and an adolescent sample with regard to predictors of the two SNS privacy-management strategies described above. Specifically, we attempt to verify whether the predictors of SNS disclosure and the use of privacy settings overlap, or whether their relative importance differs for the two age groups. In the following paragraph, we discuss why we consider it important to compare both SNS privacy-management strategies of these two age groups.

Why compare the privacy strategies of teenagers and adults on SNS?

According to popular belief, adolescents tend to ‘overshare’ online. Many adults report that they would not disclose specific information online, as they consider it too private or too sensitive (e.g. Hoffman, 2012; Holson, 2010). Nevertheless, comparative research remains scarce. Early research has shown that, compared to adults, adolescents do indeed report higher levels of personal information disclosure and lower levels of privacy settings use (Christofides et al., 2011). Adolescents have also been shown to be more willing than adults are to disclose personal information in other contexts as well (e.g. online marketing), and they are more inclined to respond positively to data requests in exchange for free gifts (Earp & Baumer, 2003; Turow & Nir, 2000; Walrave & Heirman, 2012). This finding corresponds to other, more general research into the decision-making processes of adolescents in risk situations. The findings of such research show that the decision-making processes of adolescents tend to be driven more by rewards and less by risks than are those of adults (Furby & Beyth-Marom, 1992). Younger SNS users could thus be less concerned about the possible consequences of their online data disclosure, as they tend to concentrate more on the social rewards they receive when disclosing information with their online networks. Correspondingly, research into online privacy concerns shows that adolescents are less concerned than adults are about possible privacy-related risks, including identity theft and loss of control over personal data (Earp & Baumer, 2003). Moreover, privacy concern has been found to increase with age. Research has shown that senior citizens are more concerned about their online privacy than are either adolescents or young adults (Fox et al., 2000; Hoofnagle, King, Li, & Turow, 2010; Zukowski & Brown, 2007).

In summary, given early research showing that adolescents are generally more inclined than adults are to engage in risky behaviors, disclose more readily personal information and are less concerned about privacy risks, other factors could affect online data disclosure and the application of SNS privacy settings to manage profile access. The identification of these predictors could help in the development of intervention efforts based on the most relevant predictors for different age groups. The present study therefore aims to determine whether adolescents differ from adults with regard to predictors of SNS privacy management. With this study, we hope to provide policymakers with better information regarding intervention approaches that could be implemented for SNS users of both age groups.

Hypotheses

In order to identify the variables that should be included in the questionnaire and subsequent analyses, we conducted a literature review concerning the experiences of Internet users with SNS. We included important demographic (gender, age), use (frequency, motives) and psychological variables (privacy-related concerns, contact-related concerns, trust, susceptibility to peer influence).

In the following section, we formulate hypotheses regarding the expected effects of each of the discerned variables on the SNS disclosure and privacy settings usage of adolescents and adults.

Demographic factors

Previous research suggests that the ways in which Internet users decide to use SNS is partially determined by demographic factors.

Gender

With regard to gender differences in personal data disclosure, Lenhart and Madden (2007) report that male teenagers using SNS were more eager to share their telephone numbers and home addresses than female teenagers were. Tufekci (2008) explains this finding by arguing that males are more inclined to use SNS to meet new people and to engage in new romantic relationships. Females use SNS primarily to consolidate existing relationships with friends instead of to establish new social ties. According to several studies, these gender differences appear to persist throughout adulthood (Hoy & Milne, 2010; Fogel & Nehmad, 2009). The findings of a study by Christofides and colleagues (2011), however, reveal no gender differences in the level of personal data disclosure, although female users do appear significantly more likely to apply SNS privacy settings than male users are (Christofides et al., 2011). The latter finding has been echoed in other studies (e.g. Hoy & Milne, 2010; Livingstone et al., 2011a; Tufekci, 2008; Raacke & Bonds-Raacke, 2008).

In short, previous research indicates that females are generally more protective of their online privacy with regard to the amount of data disclosed on SNS and the level of access to these data. We therefore hypothesize that:

H1a: Both adolescent and adult females disclose less personal information on their profile pages than do males.

H1b: Both adolescent and adult females apply more restrictions to their profile data access than do males.

Age

Within the adolescent age group, differences in SNS disclosure level have been observed between younger and older teenagers. For example, Lenhart and Madden (2007) observed that older teenagers were more likely to disclose photographs and the names of their schools, although no significant differences were found for other types of data. Livingstone and colleagues (2011b) report that older teenagers are more prone to disclose their telephone numbers, schools or addresses than are younger teenagers. These findings correspond with more general research on self-disclosure by adolescents outside the context of SNS. Based on a review of 50 studies addressing self-disclosure by children and teenagers, Buhrmester and Prager (1995) report an overall increase with age in self-disclosure to peers. Given that adolescence is a period of identity exploration, self-disclosure has the potential to help adolescents to develop their own identities by clarifying their own self-images and by generating external feedback from their peers (Buhrmester & Prager, 1995; Steinberg, 2011).

As adolescents age, however, their competence in managing their SNS privacy settings increases. According to one European study, over half (56%) of users aged 11-12 years claim to be able to manage their privacy settings, as compared to more than three quarters (78%) of those aged 13-16 years (Livingstone, Olafsson, & Staksrud, 2011). Interestingly, this knowledge also seems to be put into practice. Children and younger teens (9-12 years) were found to be slightly more likely to have public profiles than were older teens (13-16 years) (Livingstone et al., 2011b). This corroborates the findings of Christofides and colleagues (2011) that, as adolescents grow older, they are more inclined to restrict access to their profiles.

In short, early studies paint a somewhat mixed picture of the way in which older adolescents approach privacy protection: although they disclose relatively more data, they apparently apply stricter privacy settings than young teenagers do.

Differences have also been found among adults with regard to privacy-related concerns and protective behaviors. Although similarities have been found between younger and older adults regarding some privacy-related issues, a study among a large sample of Americans (18-65+) also revealed many discrepancies (Hoofnagle et al., 2010). The older adults in this study reported being more concerned about privacy issues on the Internet, and they were more likely to refuse to disclose personal information to businesses. Moreover, young adults overestimated the extent to which privacy laws protect individuals, both offline and online. Other studies have shown that privacy concerns increase with age. Elderly Internet users are more concerned than others are, and they are more eager to control the personal information that is collected about them (Fox et al., 2000; Zukowski & Brown, 2007).

Maaβ (2011) argues that senior citizens tend to adopt a ‘minimax’ strategy in their Internet use: they seek to maximize the rewards of their Internet use, in terms of connection and support, while minimizing risks. Senior citizens who use SNS could thus be more inclined to apply privacy-protective strategies (e.g. disclosing fewer personal details and restricting access to their profiles).

We therefore hypothesize that:

H2a: As adolescents grow older, they disclose more personal data, whereas adults disclose less personal data as they age.

H2b: As SNS users grow older, they are more likely to restrict access to their profiles.

Use Variables

Frequency of SNS use

Research has revealed a positive correlation between the frequency of online communication and self-disclosure (e.g. Valkenburg & Peter, 2009). Within the context of SNS, a longitudinal study by Trepte and Reinecke (2011) shows that users who are prone to engage in self-disclosure are more inclined to use SNS. Over time, frequent SNS users exhibit a greater tendency towards self-disclosure. This process is reinforced by the amount of social rewards that users receive. Given that self-disclosure can stimulate social interaction, it can lead to feedback or other actions that can be perceived by users as social rewards. It thus follows that the more users receive what they perceive as social rewards (in terms of bonding with other SNS users), the more they tend to self-disclose. Further, by granting access to their profiles to a broader audience, users can enhance their number of weak ties (e.g. Ellison, Steinfield & Lampe, 2011). In summary, by disclosing personal information and relaxing their privacy settings, users can lower the thresholds to initial interaction with users outside of their actual networks, thus facilitating the formation of common ground (Ellison, Lampe, Steinfield & Vitak, 2010).

Based on the studies mentioned above, we hypothesize that:

H3a: For both adolescents and adults, the frequency of SNS use is positively related to the level of personal data disclosure.

H3b: Individuals who use SNS more frequently impose fewer restrictions on access to their profiles through privacy settings.

Motives for use

Several studies have demonstrated that people use SNS for a wide range of interpersonal motives. Some use these sites to maintain contact with friends they rarely see offline, while others use them to make new friends or to follow the lives of SNS friends by viewing pictures and status updates. Many people have also re-established contact with old friends through SNS (Ellison et al., 2007; Joinson, 2008; Raacke & Bonds-Raacke, 2008). In addition to these relational and friendship motives, additional categories of SNS motives can be discerned, including entertainment, information and the desire to escape the reality of everyday life (Lin & Lu, 2011; Papacharissi & Mendelson, 2011).

Previous studies have examined the impact that motivations for use have on self-disclosure through computer-mediated communication. For example, Cho (2007) reports that self-disclosure in online chatting is fuelled by a variety of motivations, with higher levels of disclosure for people seeking to develop interpersonal relationships than for those in search of information. In another study, Gibbs and colleagues (2006) establish that online daters with long-term relationship goals engage in higher levels of self-disclosure than do other users.

Moreover, Joinson (2008) reports a significant association between motivations for SNS use and the application of privacy settings, suggesting that the motivation to meet new people is an important motivation for making online profiles less private.

Based on the findings presented above, we hypothesize that:

H4a: A positive relationship exists between the motivation of SNS users to develop and maintain interpersonal relationships and the amount of personal information that they disclose on SNS.

H4b: The more people use SNS to develop interpersonal relationships with new acquaintances, the less they restrict access to personal information on their profiles.

Psychological variables

Trust

In order for SNS to function properly, users must be willing to share some personal information about themselves. Although such sharing is accompanied by at least some level of risk (e.g. identity theft) evidence suggests that millions of SNS members do not hesitate to share personal data and even intimate thoughts within these virtual networks (Acquisti & Gross, 2006). For this reason, Grabner-Kräuter (2009, p. 506) argues that ‘social networking takes place in a context of trust’. This context of trust has two dimensions. First, SNS members trust other community members not to misuse the thoughts, life experiences and other sensitive information that they share. Second, individual SNS members trust the providers of these sites to secure their personal data. In our study, we are interested only in exploring the predictive value of trust in other SNS members on the management of SNS privacy.

Metzger (2004) argues that the role of trust in online disclosure situations should not be underestimated, given the absence of physical contact in computer-mediated communication, which reduces the number of social cues available to help verify the trustworthiness of online partners. Although SNS users probably know most of the contacts in their friend lists, these friend lists consist of a wide variety of people (Lewis & West, 2009). Moreover, some users do accept friendship requests from total strangers (Patil, 2012). Especially in situations in which SNS users have little trust in some of the contacts included in their friend lists, they may feel less inclined to disclose personal information and more inclined to apply stricter privacy settings. In this regard, previous research has established a strong and significant positive correlation between information disclosure on SNS and the level of trust in other SNS members (Mital, Israel & Agarwal, 2010). According to another study (Christofides et al., 2009), lower levels of trust in other SNS members is a significant predictor of higher levels of information control.

Based on the studies mentioned above, we hypothesize that:

H5a: Trust in other SNS members is positively associated with the amount of information divulged on SNS.

H5b: The higher the level of trust in other SNS members, the less a user will be inclined to apply strict privacy settings.

Privacy-related concerns

Within the context of SNS, several studies have reported a negative correlation between privacy-related concerns and personal data disclosure (e.g. De Souza & Dick, 2009; Krasnova & Veltri, 2010; Nov & Wattal, 2009). Moreover, some studies (Utz & Krämer, 2009; Nov & Wattal, 2009) have revealed a positive relationship between privacy-related concerns and the application of stricter privacy settings. Stutzman and colleagues (2011) report that concerns about information leakage are particularly likely to be associated with the application of stricter privacy settings.

Based on these findings, we expect that:

H6a: Privacy-related concerns are negatively associated with personal data disclosure on SNS.

H6b: As privacy-related concerns increase, SNS users are more inclined to restrict access to their profile data.

Contact-related concerns

In a study by Strater and Lipford (2008), respondents reporting events about unwanted and intrusive contacts made by strangers on SNS were more likely than other users to give thorough consideration to the information that they should disclose and the level of access they should allow to their profile data. More specifically, Debatim and colleagues (2009) report that SNS users who had personally experienced privacy invasions are more likely to modify their privacy settings. In summary, concerns about contact-related risks can be fuelled by personal experiences, as well as by other variables (e.g. media coverage concerning risks associated with SNS use), which could subsequently influence privacy management in SNS (Debatim, Lovejoy, Horn, & Hughes, 2009; Utz & Krämer, 2009).

We therefore hypothesize that:

H7a: Concerns related to contact risks are negatively associated with personal data disclosure.

H7b: As concerns related to contact risks increase, SNS users apply greater access restrictions to their profiles.

Peer influence

The ways in which peers influence the decision-making of adolescents in their daily lives have been extensively documented (for a review, see Rubin, Bukowski, & Laursen, 2011). Peers are considered important agents of consumer socialisation, especially during adolescence, due to their influence on many adolescent behaviors (Ekström, 2010). More specific to the context of social media, adolescents use SNS to communicate with peers (Lenhart & Madden, 2007; Wiley & Sisson, 2006), and this fosters the development of individual identity, the quality of peer relationships and self-esteem (Valkenburg, Peter, & Schouten, 2006). The use of SNS may provide adolescents with a venue for receiving peer feedback about themselves. This is especially useful for adolescents, who are highly susceptible to peer influence. Although empirical research on the relationship between peer influence and SNS disclosure is scarce, previous studies have suggested peer influence as a possible venue for future investigations on the amount of data that young people disclose on SNS (Acquisti & Gross, 2006). More specifically, adolescents who are more susceptible to peer influence and who have a greater need for popularity among their peers are assumed to provoke relatively more peer feedback by disclosing more and subsequently applying fewer restrictions to the disclosed information (Christofides, Muise, & Desmarais, 2009).

Fewer studies have addressed the topic of peer influence on adults. In general, adolescents are seen as more susceptible to the influence of their peers than adults are (Steinberg & Monahan, 2007). According to an experimental study by Gardner and Steinberg (2005), the peer effects on risk-taking and risky decision-making are stronger among adolescents than they are among adults. Gardner and Steinberg (2005) suggest that the psychosocial capacities needed to resist peer pressure may continue to develop throughout late adolescence and into early adulthood.

Based on the studies mentioned above, we hypothesize that:

H8a: High susceptibility to peer influence is positively associated with the level of SNS disclosure.

H8b: SNS users who are highly susceptible to peer influence are less inclined to use SNS privacy settings to restrict access to their profile data.

Methods

Procedure and participants

We conducted an online survey in collaboration with MSN in Belgium. The study was presented as an investigation of SNS usage. In order to avoid priming the participants, the invitation to participate in the survey contained no allusion to privacy issues. Individuals with SNS profiles were invited to participate. The invitation was presented on the MSN starting page.

The total sample consisted of 1484 respondents. The youth sample consisted of 343 respondents (aged 10-19, M = 16.2, SD = 2.2) including a balanced mix of boys (45%) and girls (55%). The adult sample (aged 20-65, M = 40.0, SD = 14.1) consisted of 54% women.

Measures

Frequency of SNS use: The frequency of SNS use was measured by asking respondents how often they consulted their profile sites (six-point scale ranging from ‘once a month or less’ to ‘several times a day’).

Personal data disclosure: In order to assess their disclosure of personal information, respondents were confronted by a list of 18 types of personal data. The respondents were subsequently asked whether they had divulged each type of information on their profiles. The scale was found to be reliable (α = .90). A summative index was used to measure the level of data disclosure.

Application of privacy settings: We assessed the application of privacy settings by asking respondents how they managed the access to the personal data divulged on their profiles. For each piece of information disclosed on their profiles, respondents could indicate their level of access by choosing one of the following four categories: ‘all users’, ‘friends and their friends’, ‘only friends’ or ‘restricted to myself’. Raw scores were aggregated, with higher values indicating a higher level of restriction on access to profile data through the application of privacy settings. This scale was found reliable (α = .89).

Motives: In order to gain insight into the motives that the respondents had for using SNS, the questionnaire included a series of seven items regarding the importance of specific contact-related motives (based on Subrahmanyam et al., 2008), measured on a five-point scale (ranging from ‘not important at all’ to ‘very important’).

Concerns: Two types of concerns were measured. To probe the level of privacy concern, respondents were presented with a series of nine statements, based on, amongst others, Milne and Culnan (2004). The following is an example: ‘I am concerned about what websites do with my personal data’. Responses were measured using a five-point Likert scale anchored by 1 (‘totally disagree’) and 5 (‘totally agree’). Factor analysis identified a single factor (see Appendix), consisting of seven items, and the scale was found reliable (α = .76). Raw scores were summed, with higher values indicating a higher level of privacy concern.

We then measured the concerns held by the respondents with regard to the possible negative consequences of SNS use, especially contact-related risks (e.g. ‘I am concerned about the possible misuse of my profile to bully me’). Four items were measured using a five-point scale, and they formed a single factor (α = .71, see Appendix). Subsequently, a summative index was constructed, with higher scores on the scale defining higher levels of SNS contact-related concerns.

Trust: A scale was constructed to assess the level of trust in other SNS users. These three items were measured using a five-point scale, and they formed a single factor (e.g. ‘People I meet on SNS are trustworthy’, α = .77, see Appendix).

Susceptibility to peer influence: Peer Influence was assessed using items from Bearden, Netemeyer and Teel’s (1989) and Mangleburg and Bristol’s (1998) scales for measuring the extent to which individuals are inclined to follow advice from their peers in consumption situations. A single factor was discerned. This seven-item scale was found reliable (α = .88, see Appendix).

Results

Descriptive findings

The results of our study show that the most popular SNS among the adolescent respondents was Facebook (60%), followed by the Belgium-based Netlog (55%). Facebook (55%) was also the most popular SNS among adult respondents. Although the difference in popularity between Facebook and Netlog was relatively small among adolescents, we found a greater discrepancy in this regard among adults, with only 20% of the adult respondents using Netlog. This is not surprising, as Netlog is targeted primarily at young people between the ages of 14 and 24 years (Donoso, 2011). Other SNS were used by a small minority of both adults and adolescents: Hyves (adolescents: 6%; adults: 4%), LinkedIn (adolescents: 0.5%; adults: 5%), MySpace (adolescents: 7%; adults: 7%) and Twitter (adolescents: 7%; adults: 2%; other SNS: adolescents: 10,5%; adults: 5%).

Both of the most popular SNS in Belgium offer granular privacy settings that allow users to restrict the access of other users to their personal data (Fluckiger, Grehan & Donoso, 2011; Walrave, Taddicken & Donoso, 2011).

Our results reveal a major difference between adolescents and adults with regard to the frequency of SNS usage. Whereas a majority (54%) of the adolescent respondents reported checking their profile at least once a day, only 45% of the adult respondents did so (χ² = 13.98, df = 1, p < .001). The same holds for posting status updates: while 17% of teens posted messages at least daily, only 12% of adults did so (χ² = 7.16, df = 1, p < .001).

Comparison of the subsamples of teenagers and adults reveals significant differences with regard to the disclosure of data on SNS and the application of privacy settings. On average, adolescents were inclined to disclose 13 of the 18 requested pieces of personal information (M = 13.4, SD = 4.0), whereas adults were willing to disclose significantly less information (M = 12.0, SD = 4.6, t(1002) = 3.11, p < .05). Moreover, teenagers scored significantly lower on the scale measuring the application of privacy settings, which indicates that they allow a broader audience access to their profile data (adolescents: M = 35.5, SD = 13.2; adults: M = 39.6, SD = 14.2, t(954) = 5.92, p < .001).

We found no significant difference between the two subsamples with regard to the level of trust in other SNS users (adolescents: M = 3.0, SD = 2.6; adults: M = 3.1, SD = 2.5, t(2054) = -.34, p > .05). Interestingly, trust in SNS users was quite low. An additional finding is that the adolescent respondents were less concerned about their online privacy than the adult respondents were (adolescents: M = 27.6, SD = 4.7; adults: M = 29.1, SD = 4.4, t(1323) = -8.29, p < .001). The same holds for concerns about contact-related risks (adolescents: M = 16.4, SD = 4.9; adults: M = 17.9, SD = 5.0, t(888) = -5.92, p <.001). Finally, adolescents scored significantly higher on susceptibility to peer influence (adolescents: M = 11.8, SD = 5.8; adults: M = 10.3, SD = 6.1, t(942) = 4.87, p < .001).

Predicting SNS privacy management

As mentioned before, the second objective of this study (in addition to providing additional insight into the ways in which Internet users approach disclosure and the application of SNS privacy settings) is to compare whether the two SNS privacy-management strategies are fuelled by the same predictors for adolescents and adults. To compare possible predictors of data disclosure and the application of privacy settings, separate hierarchical regressions were performed for each age group.

In the sequential regression equations, gender and age were entered first, followed by the frequency of SNS usage and motives. Variables measuring trust in other SNS users, general online privacy concerns and specific concerns related to online contact risks were entered in subsequent equations. The final regression equation included also the susceptibility of respondents to peer influence.

Predicting disclosure: Adolescents versus adults

The regression equation accounted for 16% of the variation in data disclosure by adolescents and for 12% of such variation among adults. The final models are summarised in Table 1.

The main predictors of data disclosure differed between adolescents and adults.

For adolescents, the use of SNS to communicate with schoolmates, gender and susceptibility to peer influence were significant predictors of SNS data disclosure. The three most important predictors for adults were the frequency of SNS usage, level of trust in other SNS members and using SNS to find a partner.

Our analyses show that the most important predictor of SNS disclosure by adolescents is the motivation to communicate online with schoolmates outside of class hours. Female adolescents proved less inclined to disclose personal data, and adolescents who were more susceptible to peer influence were inclined to disclose more personal information. Adolescents motivated to use SNS in order to reconnect with distant friends were also inclined to disclose more personal data.

The analyses further identify the frequency of SNS usage as the most important predictor of the disclosure of personal data by adults. Other significant predictors include trust and concerns related to privacy risks. Our results reveal a positive relationship between the level of trust that adults have in other SNS users and the amount of personal data that they disclose on their profiles. Conversely, a negative relationship exists between the level of concern for privacy on the part of adults and the extent to which they disclose data on SNS. The results further identify three contact motives for using SNS as significant predictors of disclosure: communicating with colleagues, engaging in communication with family members and searching for a partner.

Although we found no significant gender difference for adult SNS disclosure, we did find that age has a significant influence, with older adults being less willing to disclose personal data on their profiles.

Table 1. Predictors of data disclosure (final models).

Finally, the results offer support for the following hypotheses: H1a (partially supported for adolescents), H2a and H3a (for adults), H4a (partially supported for adolescents and adults), H5a (for adults), H6a (for adults), H8a (for adolescents). The results indicate that H7a should be rejected.

Predicting the application of privacy settings: Adolescents versus adults

The results reveal differences between adolescents and adults with regard to the type and strength of predictors for the application of privacy settings.

Although adolescent females were more likely to apply privacy settings in order to restrict access to profile data, gender was not a significant predictor for the information-control measures used by adults. Interestingly, we also found that teenagers become less inclined to restrict access to their profiles as they age, while older adults proved more likely to apply stricter privacy settings.

Adolescents using SNS to connect with friends with whom they had lost contact, those who were eager to meet new people online and those using SNS to converse with schoolmates were more inclined to relax their privacy settings. Adults using SNS to make new acquaintances, find a partner and contact distant friends were more likely to open their profiles. In this case as well, increased frequency of SNS usage apparently makes adults more inclined to open their profiles to a broader audience.

Adults concerned about contact-related risks tend to place greater restrictions on access to their profile data. The results further reveal similarities between adolescents and adults with regard to the influence of privacy concern, with greater privacy concerns leading to stricter privacy settings.

The final model (see Table 2) accounted for more variance for teenagers (adjusted R²: 24%) than for adults (adjusted R²: 22%).

Table 2. Predictors of the application of privacy settings (final models).

In short, our analyses provide support for H1b (for teenagers), H2b (for adults), H3b (for adults), H4b, H6b and H7b (for adults). The results indicate that H5b and H8b should be rejected.

Discussion

Although research and policy have devoted extensive attention to possible risks related to the disclosure of personal data, few comparisons have been made between adolescents and adults with regard to SNS disclosure and the application of privacy settings (Christofides et al., 2011). The overarching goal of the present study is therefore to explore and compare possible predictors of both privacy-management strategies between adolescents and adults.

In contrast to the results of previous studies (e.g. Christofides et al., 2011; Joinson, 2008), gender was not a significant predictor of privacy settings use in our adult subsample. Our results do show, however, that female adolescents were more protective of their online privacy by disclosing less information and by placing greater restrictions on access to their profiles. The latter is in line with the findings of Christofides and colleagues (2011) concerning adolescents’ use of privacy settings. As suggested by Staksrud and Livingstone (2009), teenage boys apparently pay less attention to online risks than teenage girls do. This result could be explained in part by the fact that teenage girls report that their parents impose relatively more rules regarding the disclosure of personal information than do boys (81% versus 74%) (Livingstone et al., 2011a). Teenage girls could thus be more sensitized about online privacy protection.

Interestingly, our results indicate that adolescents tend to open their profiles as they age, while an opposite relationship between age and profile access was observed among adults. This result could be explained in part by the way in which teenagers develop during adolescence. Throughout this period, teenagers find it increasingly important to be in contact with their peers (Brown & Klute, 2003). In order to optimize their chances of being contacted by their peers, therefore, adolescents could deliberately foster access to their profiles by applying more lenient privacy settings.

Conversely, older adults are more inclined to restrict access to their profiles, as privacy-related concerns have been found to increase with age (Fox et al., 2000) and senior citizens using the Internet have been found to be more sensitive to privacy than younger adults are (Zukowski & Brown, 2007).

Our results identify relational motives as important predictors of both data disclosure and the application of privacy settings. According to our analyses, adolescents and adults who use SNS to contact their schoolmates and colleagues are more likely to disclose data on SNS. In this context as well, bonding with existing offline contacts is associated with higher levels of self-disclosure. As observed by Burke and colleagues (2010), active contributions to SNS and direct communication can enhance the social capital of users.

The desire to meet new people online and to contact distant friends stimulates both adolescents and adults to open their profiles. This need for bridging social capital could encourage SNS users to increase their searchability and accessibility by relaxing their privacy settings in order to lower the threshold for contact. However, adults motivated to use SNS in order to search for a partner are inclined to disclose more information and to relax their privacy settings more than other adult users are. One possible explanation is that increased self-disclosure leads to greater perceived relational success (Gibbs et al., 2006).

Another finding of the present study is that trust in other SNS users has a positive influence on the personal data disclosure of adults. This finding is in line with previous research among adults (Dwyer et al., 2007; Christofides et al., 2009; Utz & Krämer, 2009). It could be that trust reduces the risks perceived in relation to the disclosure of private information (Metzger, 2004). Nevertheless, trust levels were found to be very low among both adolescents and adults, thus suggesting an overall critical attitude towards, amongst others, information posted by others on SNS.

In contrast, our results indicate that privacy-related concerns affect the disclosure and profile-access restrictions of adults, while such concerns affect only the application of privacy settings by adolescents. This contrasting result suggests possible differences between the ways in which adolescents and adults perceive privacy and related concerns. While adolescents are more likely to communicate certain types of personal data freely, their sensitivity to privacy is more likely to focus on who is to have access to these data (Livingstone, 2008). Adults were found to be more concerned about the contact-related risks of SNS, and this influenced their profile-access management as well. In summary, the results indicate that adolescents were less concerned about SNS-related contact risks. Moreover, this concern did not influence their level of disclosure and profile-access restrictions.

In line with previous studies, the susceptibility of adolescents to peer influence encourages them to disclose more. Teenagers who are more sensitive to feedback from their peers could be encouraged to disclose more, which in turn generates feedback (Christofides, Muise & Desmarais, 2009).

One of the central findings of this study is that, in addition to disclosing more personal data than adults do, adolescents also apply less restrictive privacy settings to their data, thus corroborating the findings reported by Christofides and colleagues (2011). One possible explanation for this finding could be related to adolescents’ development. Adolescence is a life stage in which individuals discover and try to understand who they are as individuals, and in which they develop a coherent sense of self, including gender, sexual, social and cultural facets. Although identity changes can occur throughout the lifespan, adolescence is the first life stage in which individuals become more self-conscious and capable of appreciating potential changes and imagining possible selves (Steinberg, 2011). During this process of identity development, adolescents strive to gain more autonomy, especially from their parents.

As the influence of parents decreases, however, peers become increasingly important. Among peers, adolescents develop a broader range of relationships, including close friendships and intimacy with romantic partners (Steinberg, 2011; Subrahmanyam & Smahel, 2011). In no other period of life is the need for social interaction, identity experimentation and peer feedback as important as it is during adolescence (Harter, 1999). The activities of adolescents on SNS can play an important role in this regard. More specifically, SNS offer adolescents the opportunity to choose how they wish to present themselves in words and pictures. The ability to manage self-presentation enhances individuation (Schmitt et al., 2008) and facilitates further self-disclosure (Valkenburg & Peter, 2009), thereby leading to reciprocity and closer relationships (Burhmeister & Prager, 1995). The need to belong and the need for popularity, which are particularly important during adolescence (Santor et al., 2000), have been positively associated with online self-disclosure (Christofides et al., 2009).

In addition to the possibility of managing self-presentation, SNS offer adolescents the possibility of controlling the users with whom they will share specific information, as well as to contact distant friends and engage in online contact with other users who share the same interests or interrogations. During this important life stage, the opportunities offered by SNS may partially explain the engagement of adolescents in SNS and, more specifically, their greater inclination to disclose sensitive information to peers. Moreover, given their involvement in identity experimentation and self-reflection, adolescents could be more inclined than adults are to share and try out facets of their personalities that they would like to explore or to which they aspire to assume (Subrahmanyam & Smahel, 2011).

The present study is subject to several important limitations. First, the data were drawn from a convenience sample, although respondents were not primed to privacy issues prior to their participation. Second, the significance of several independent variables approached the threshold level. Third, the variables addressed in this study explained only 12% of the variance in the SNS disclosure of adolescents and 16% of adults’ data disclosure, and they explained a relatively small portion of the variance in the application of privacy settings by adolescents (24%) and adults (22%). A large portion of the variance in both dependent variables thus remains unexplained. The inclusion of additional psychological constructs (e.g. personality traits, self-esteem and need for popularity) may help to explain a greater portion of the variance. Additional factors that could explain differences in online profile-access management include the level of knowledge that users have regarding privacy settings and their level of self-efficacy with regard to protecting their privacy online.

The most important results of this study have several implications. Some of the predictors could be used to inspire further awareness-raising efforts focusing on SNS users. Privacy-related concerns emerged as one of the most important predictors of the application of privacy settings by both adolescents and adults, in addition to predicting disclosure by adults. Whereas concern about contact-related risks influenced adults to restrict their profile audiences, such concerns had no effect on the privacy-management strategies of adolescents. Concerns about contact risks could therefore be triggered in order to encourage adults to make greater use of their privacy settings.

For both adolescents and adults, sensitizing campaigns could be deployed to increase concern about privacy issues, thus encouraging both groups to become more conscious about audience management on SNS. We recommend that such sensitizing initiatives should be accompanied by practical training on how users can deploy the complex and increasingly granular privacy settings in order to enhance the protection of their online profiles. Such training is apparently necessary, as a recent study found that Internet users (both adolescent and adult) can experience difficulties finding and understanding information about privacy settings on most SNS (Donoso, 2011). Moreover, given that self-reported competence in the application of privacy settings increases with age (Livingstone et al., 2011a), young SNS users may be more vulnerable due to their broadly accessible profiles.

In 2009, the major SNS providers committed themselves to implementing the Safer Social Networking Principles, in addition to other actions, including setting the profiles of minors to private by default. Since that time, however, only a few providers have actually honored their pledge (Donoso, 2011). The reasons for this lack of commitment call for thorough investigation; the implementation of all principles should be subjected to further close monitoring, and possible supplementary policy initiatives should be explored.

Moreover, awareness-raising efforts targeted at adolescents, parents and teachers should be strengthened. Although adolescents are increasingly confronted with education and awareness-raising efforts regarding e-safety, these programmes often focus on online predators, cyberbullying and other contact-related risks (Hoofnagele et al., 2010). Both adolescents and adults should be made aware that data disclosure has been related to increased likelihood of being confronted with these risks (e.g. Palfrey, 2008; Walrave & Heirman, 2011; Ybarra & Mitchell, 2008).

It has further been shown that adolescents who have had negative experiences on SNS are almost twice as likely as other users are to have public profiles (Lenhart et al., 2011). For this reason, deeper reflection on the disclosure of personal data, and especially the management of access to profile data, could be promoted as a strategy for possibly reducing online risks while enjoying the contact and other opportunities offered by SNS.

References

Appendix : Factor analyses of scales