Predicting Information Self-Disclosure on Facebook: The Interplay Between Concern for Privacy and Need for Uniqueness

Abstract

This study examined the overall relationship between information privacy concern, need for uniqueness (NFU), and disclosure behavior to explain the personal factors that drive data-sharing on Facebook. The results of an online survey conducted with 222 Facebook users show that among diverse data that social media users disclose online, four distinct factors are identified: basic personal data, private data, personal opinions, and personal photos. In general, there is a negative relationship between privacy concern and a positive relationship between the NFU and the willingness to self-disclose information. Overall, the NFU was a better predictor of willingness to disclose information than privacy concern, gender, or age. While privacy concern has been identified as an influential factor when users evaluate social networking sites, the findings of this study contribute to the literature by demonstrating that an individual's need to manifest individualization on social media overrides privacy concerns.

1. INTRODUCTION

There has been a widespread discussion around the concept and the uses of big data in various disciplines including business, communication, and information systems. Simply put, the term refers to a massive collection of complex data. While companies are accruing data at petabyte scale, the discussion on how the enormous amount of information collected via the Internet should be utilized to better communicate with individual users is still nascent. Online ads have been personalized at a crude level from saluting with a user’s name to a more sophisticated level showing ads based on the user’s “likes.” Since 2014, Facebook has been clear about gathering users’ Web browsing data beyond its own platform and using it for advertising [1], and Facebook’s business has expanded around users’ entire browsing history and information voluntarily provided by the users. This is an example of the current state of using big data; however, over the past few years, Facebook has been facing legal investigations regarding the collection and the usage of users’ personal data. Provision of personal data and privacy are two sides of a coin, yet how users’ balance concern for privacy and need for personalization is still unclear in the literature.

At large, two types of data can be collected on the Web. For one, users’ Web browsing data can be collected without the users’ explicit permission. This type of data is often used to monitor the actions of site visitors and the information collected is used to personalize content they are exposed to. In turn, consumers benefit by receiving content that better matches their personal needs, wants, and interests [2]. Another type of data is based on users’ voluntary input. Research has examined the degree to which Internet users reveal their personal information online and the motivation for voluntary self-disclosure. For example, people are more inclined to disclose information on commercial websites when they spend more time on the Web [3]. Both types of data may be used by websites to provide personalized service [2]. A downside of such service is that it may be perceived as a threat to privacy to the users. As social media is built on “the ideological and technological foundations of Web 2.0” that allow the creation and exchange of content [4], information sharing is inevitable and social media can thrive only when users voluntarily offer to supply content and exchange information with other users. Therefore, there lies a seemingly conflicting issue of information disclosure and maintenance of privacy.

The greatest advantage of personalization based on an individual’s information is that it increases personal relevance and a sense of uniqueness for that individual. Need for uniqueness is an individual trait [5], [6] that makes people want something different from others. The tradeoff of receiving something unique to one’s self, in this case, personalized content, is that individuals sometimes have to forgo their privacy. Therefore, the purpose of this study is to examine how an individual’s concern for privacy and need for uniqueness affect willingness to disclose various types of personal information on Facebook (name, hometown, gender, education, birth date, profession, feelings, political views, religious views, home address, purchase history, photos, etc.). The findings of this study will contribute to our understanding of the delicate nature of privacy issues on social media. Furthermore, insights on what drives users to share their information on social media are provided.

 

2. LITERATURE REVIEW

2.1 Information Self-Disclosure

Self-disclosure is defined as the act of revealing personal information to others that is generally unknown and not available from other sources [7], [8]. It is often times an integral part of online communication [9]-[12], especially on social media as the very basic requirement of using the platform is sharing some form of personal information with other users. The extent of the shared content can range from basic personal information such as name, age, and gender to highly private information such as pictures, personal opinions about sensitive issues, and religious views. While self- disclosure can take place interpersonally and via online, for an average individual, the stark difference between the two is that for the latter, the information shared publicly can be viewed by anyone with an access to the Internet unless one makes an adjustment to the default privacy setting. In addition, information disclosed online can often be traced and stored; thus, users typically face higher stake of tainting their self- image and higher risk of identity theft. Nevertheless, most people disclose information about themselves in exchange for maintaining relationships, facilitating communication, and convenient transaction.

Research has shown that computer-mediated communication encourages more intimate interaction among people due to lack of nonverbal cues and, therefore, individuals engage in deeper self-disclosure than in a face-to-face setting in order to reduce uncertainty [13] and increase intimacy [14], [15]. A user’s willingness to self-disclose is important to marketers as well when they attempt to develop and maintain relationships with their consumers [16]. The information gained can be used as a way to develop trusting relationships and to provide personalized interactions. As social media is a space to reveal and exchange personal information [17], [18], being comfortable with self-disclosure became almost a requisite and social media users tend to show a high level of disclosure behavior [3], [19]. Despite these circumstances that make self-disclosure an essential part of online communication, users still need to make a choice between being completely open to sharing and protecting their privacy.

 

2.2 Communication Privacy Management Theory and Concern for Privacy

Privacy is defined as a personal boundary regulation process to regulate the levels of privacy with others depending on the context [20]. In conjunction with self-disclosure, privacy concerns have been identified as key antecedents to online behavior [21]-[23] and online transactions, especially when personal information is used [24]. Communication privacy management theory proposes that privacy and self-disclosure is an individual’s balancing act where individuals develop their own set of rules to make decisions about revealing or concealing private information [25], [26]. One interesting aspect about concern for privacy and disclosure behavior is that individual’s rule may change with time, familiarity with the medium, and experience. For example, by using longitudinal data of 5,076 Facebook users, Stutzman and colleagues were able to show how Facebook users’ disclosure behavior changed between 2005 and 2011 [27]. Over time, users demonstrated more care about privacy by limiting personal data shared publicly while what users shared within their network increased. A number of studies have demonstrated that privacy concern is found to have a negative impact on self-disclosure. For example, when online privacy policy on the willingness to provide personal information on websites was examined, there was a negative effect of privacy concerns on willingness to provide personal information [28]. The more users were concerned for privacy, the less they were willing to disclose information on social networking sites [20] and when users cared for customizing privacy settings, they were less likely to disclose information on Facebook [29]. Thus, consumers who are concerned about their online privacy will be less willing to disclose personal information on Facebook.

H1: Concern for privacy will be negatively related to general willingness to disclose personal information on Facebook.

However, research has also demonstrated that concern for privacy might not guide our behaviors online as much as we think it does [12]. Despite concerns for privacy, individuals still reveal their personal information online by leaving traces behind through online activities and also by voluntarily providing information with purpose. This can be explained by impression management literature that suggests people engage in actions to create and maintain a desired image [30].

Therefore, information they provide on social media can be collectively used to present themselves in a desirable way.

 

2.3 Impression Management and Need for Uniqueness

How we view our own self is central to many of the choices we make. Research on the “self” has shown that people have varying levels of desire to be unique [5]. Individuals are known to vary in their need for differentiation and the expression of their own personal traits and attributes [31]. When it comes to consumption, studies have demonstrated that consumers’ preference for something (i.e., a product or a brand) always depends on need for uniqueness, which is the desire to be seen as different from others [6], [32]. Need for uniqueness is the internal drive to be different from others for the purpose of developing and enhancing one’s personal and social identity [6].

Naturally, people with higher need for uniqueness prefer products that are more scarce or unique, whereas those with a low need for uniqueness may make their choices in line with others [6]. In the consumer socialization process, need for uniqueness has a significant moderating effect on consumer evaluations; when others serve as a reference point, high- uniqueness consumers are less likely to be influenced by others’ opinions than low-uniqueness consumers [32]. Therefore, people with high need for uniqueness are less willing to conform and less likely to be affected by peer communication, whereas the direct effect of peer communication on product attitude is more evident for low- uniqueness consumers [33].

Similarly, one’s need for uniqueness can also affect willingness to receive personalized information based on one’s shared information. It is proposed that high-uniqueness users will appreciate personalized information unique to the individual more than low-uniqueness users.

H2: Need for uniqueness will be positively related to general willingness to disclose personal information.

 

2.4 Social Penetration Theory and Information Disclosure

To explain the procedure of how people reveal themselves to others, social penetration theory [34] is introduced. Social penetration theory sheds light on how people reveal themselves to get intimate with others. The degree of self-disclosure is adjusted by controlling the depth (the level of intimacy) and breadth (the amount of topics or dimensions revealed)—by exchanging more sensitive and larger amounts of information, people penetrate each other and develop intimate relationships. This can be applied to information shared on social media as users can decide the depth and breadth of information that they choose to disclose or keep private.

As reviewed previously, it is predicted that concern for privacy has a negative impact on general willingness to self- disclose while need for uniqueness is expected to have a positive impact on information self-disclosure. As these two constructs exert seemingly conflicting effects on information disclosure, the following research question is posed to examine the effect these two have on information disclosure on social media. In addition, two key demographic information, gender and age, that have been identified as factors that affect users’ willingness to share personal information are examined. A past study has shown that women have a stronger preference for privacy and are more likely to have private profiles (on Facebook) than men [35]. Similarly, men have been reported to exhibit higher willingness to disclose basic information about them than females [36]. When it comes to age, younger users are in general less concerned about privacy than older users and online privacy worries increase with age [28].

RQ 1: What is the strongest predictor (concern for privacy, need for uniqueness affect, willingness to disclose information, demographic information) of disclosing various types of information on Facebook?

 

3. METHOD

3.1 Participants and Procedure

Using Amazon Mechanical Turk, an online survey was conducted with 222 adults in the U.S. who are users of social media. There was almost an even split between female respondents (55.9%) and men (44.1%), and ages ranged from 18 to 77 years (M = 33.05, SD = 11.96). The majority of the participants were White, non-Hispanics (70.3%), 11.7% were African Americans, 7.7% Asian American, 5.4% Hispanic, 5.0% others. About 94% of the respondents have used the internet for more than 7 years. The vast majority of the participants reported that they use Facebook at least once a week (93.2%), 58.4% use Twitter at least once a week, 52.5% use Pinterest at least once a week, 47.0% use Instagram at least once a week, and 38.6% use LinkedIn at least once a week.

First, participants were asked a series of questions about their Internet usage behavior. To measure information disclosure intention on social media, the participants were asked to imagine they are on Facebook and indicate their general likelihood of disclosing given list of information (i.e., First name, Hometown, Gender, Last name, Education, Birth date, Profession, Home address, Purchase history, Browsing history, Current location, Political views, Current emotions, Feelings, Religious views, Thoughts, Habit, Interests, Photos of self and friends, Photos of self and family, Photos of just oneself).

 

3.2 Measures

The key constructs of the study were measured using scales adopted from previous literature.

 

3.2.1 Information privacy concern:

A 4-item scale measured on a 7-point scale (1 = strongly disagree to 7 = strongly agree) was used: It bothers me to provide personal information to so many companies; When companies ask me for personal information, I sometimes think twice before providing it; It usually bothers me when companies ask me for personal information; companies should take more steps to protect personal information against unauthorized use [37] (α = .86). 3.2.2 Need for uniqueness: A 5-item scale measured on a 7- point scale (1 = strongly disagree to 7 = strongly agree) was used: I collect unusual products as a way of telling people I’m different; I have sometimes purchased unusual products or brands as a way to create a more distinctive personal image; I often look for one-of-a-kind products or brands so that I create a style that is all my own; I actively seek to develop my personal uniqueness by buying special products or brands; The products and brands that I like best are the ones that express my individuality [6] (α = .92).

 

3.2.3 Likelihood of disclosing information on Facebook:

Participants were asked to rate the likelihood of disclosing following information on Facebook using a 7-point scale (1 = very unlikely to 7 = very likely): First name (real name), last name (real name), hometown, gender, education, birth date, profession, feelings, current emotions, thoughts, habit/interests, political views, religious views, home address, current physical location, mobile phone number, purchase history, photos of self and friends, photos of self and family members, and photos of just oneself. These items were adapted from previous studies that have measured information sharing behavior on social media [36]. For the ease of analyses, a factor analysis was conducted to combine the items to smaller categories.

 

3.2.4 General willingness to self-disclose:

As a control variable, general willingness to self-disclose was measured. A 5-item scale measured on a 7-point scale (1 = strongly disagree  to 7 = strongly agree) was used: It is difficult for me to talk about myself (reversed); I prefer that people know only a little bit about me (reversed); I sometimes find myself telling casual acquaintances things about myself; There are many things about myself that I would rather not talk about with other people (reversed); I will not talk about personal matters unless someone else does so first (reversed) [12] (α = .74).

 

4. RESULTS

To categorize the comprehensive list of information that individuals typically disclose on social media, principal axis factor analysis with varimax rotation was conducted. KMO measure of sample adequacy was .93, above the recommended value of .6, and Bartlett’s test of sphericity was significant (χ2 (300) = 5397.4, p < 0.001. All items had primary loadings over .5. “Interests” was eliminated because it cross-loaded (> .5) with two factors, basic personal information and personal opinions. Using an eigenvalue cut-off of 1.0, four factors were extracted with a cumulative variance of 68.6%. Based on examining the items of each factor, the following four factors were identified: Basic personal information, personal opinions, private information, personal photos. With the rotated solution, the first factor explained 24.0% of the variance, the second factor 18.1% of the variance, the third factor 15.4%, and the fourth factors explained 11.1% of the variance. The factor loading matrix is presented in Table 1 and the descriptive statistics of the factors are in Table 2.

Bivarite correlations were examined to see the overall relationship between information privacy concern (IPC) and need for uniqueness (NFU). There was a moderate negative relationship between concern for privacy and general willingness to disclose personal information (r = -.30, p < 0.001) and a weak positive relationship between need for uniqueness and general willingness to disclose personal information (r = .20, p < 0.001) (Table 3).

 

Table 1. Factor loadings and communalities

 

Table 2. Descriptive statistics for the willingness to disclose four types of  information on social media

 

Table 3. Correlation, means, and standard deviations of measures

 

Table 4. Impact of need for uniqueness, willingness to self-disclose, and information privacy concern on intentions to reveal information on social media

 

To test the proposed hypotheses and answer RQ1, gender, age, information privacy concern, need for uniqueness, and general willingness to self-disclose were entered in a regression model to predict the likelihood of revealing each type of information on social media. See Table 4 for details of the results.

For all information types, the two demographic variables were not significant predictors of intentions to reveal information. For basic personal information, individual’s general willingness to self-disclose (β = .15, p < .05) was the sole significant predictor. For disclosing private information, information privacy concern (β = -.14, p < .05) and need for uniqueness (β = .20, p < .01) were significant predictors. As predicted, IPC had a negative effect on disclosing private information whereas NFU had a positive effect. For sharing personal opinions, NFU (β = .25, p < .01) had a significant positive effect and for sharing personal photos, NFU (β = .17, p < .05) also had a positive effect. For both sharing personal opinions and personal photos, privacy concern had a negative effect but the results were not statistically significant.

 

5. DISCUSSION AND IMPLICATIONS

The results of this study showcase the overall relationship among IPC, NFU, and disclosure behavior. The more concerned people are about their information privacy, the less they are willing to disclose personal information. Conversely, individuals who score high on need for uniqueness in general exhibit higher willingness to disclose personal information. The first key contribution to the literature is that information shared on social media is treated differently by users. How privacy concern and need for uniqueness influence information disclosure depends on the type of information an individual is asked to share on social media.

As the type of information collected may guide one’s willingness to provide personal data, a comprehensive list of various types of information collected online were formed. Factor analysis revealed that these can be classified to four categories: Basic personal information, private information, personal opinions, and personal photos. In addition to the two key concepts examined in the study, gender and age were included in the regression model to test their effects on information self-disclosure of the four types of information. Findings showed that different factors affect willingness to disclose each type of information.

Individual’s general willingness to self-disclose was the main driver for disclosing basic personal information such as the first and last names, hometown, gender, education, birth date, and profession. IPC had a negative and NFU had a positive effect on willingness to disclose private information such as home address, current location, purchase history, and browsing history. For revealing personal opinions such as political and religious views, current emotions, feelings, thoughts, and habit, NFU was the significant predictor. Lastly, NFU predicted willingness to share personal photos including photos of one’s self, friends, and family. Neither gender nor age was significant predictors of all types of information including private information.

The second key contribution to the literature is that NFU is identified as a social influence construct that affects an individual’s self-disclosure. Social penetration theory suggests that by exchanging more sensitive and larger amounts of information, people penetrate each other and develop intimate relationships [34]. In the current study, the receiver was not presented to the respondents. However, because the participants were instructed to imagine they are on Facebook, it is likely that they viewed themselves disclosing information to their close in-networks. The results imply that sharing personal opinions and photos are outcomes of individuals’ need to differentiate themselves from others. Indeed, these types of information are typically unique to individuals and they are able to use them to display their characters and personalities.

A moderate negative relationship was detected when examining bivariate correlation between privacy concern and overall information disclosure. An unexpected finding, though, is that in most occasions, IPC served as non-significant predictor of disclosing information. This finding supports a stream of research that revealed absence of a direct influence of privacy concerns on disclosure [38]. Privacy concern had a negative influence on willingness to provide personal information; however, the relationship was not strong. It seems that when individuals share content largely created by themselves (e.g., opinions, photos), concern for privacy is not an issue that deters them from sharing.

Some research has demonstrated that information disclosure and information control are not negatively correlated [39]. In other words, similar to current research, concern for privacy does not necessarily discourage individuals from disclosing information. The notion of “privacy paradox” that suggests consumers’ privacy might not be such a hindering issue for companies as we think [40] gives insight into this phenomenon. Although Internet users cognitively acknowledge their concern for privacy, behaviourally, they are not too guarded or careful about revealing information about themselves. Especially when using social media, the issue of privacy might seem not so critical because when an individual chooses to follow a brand or a company on Twitter or “like” something on Facebook, there is an implicit agreement that he or she is choosing to receive personalized messages and some form of tailored information based on their personal information (i.e., their online behavior). Therefore, the privacy issue often raised regarding the collection of personal information on the Web might not be of a concern when people are on a social media platform.

While privacy paradox phenomenon explains why users might share information despite privacy concerns, this study revealed that when it comes to sharing private information, concern for privacy does have a negative influence on individuals. Private information in this study included home address, current location, and purchase and browsing history. This implies that such information that can be used beyond digital sphere and disturb individuals face-to-face or information that are typically used for commercial purposes do concern social media users. Nevertheless, the beta value for NFU was slightly higher than IPC. This implies that if this is of a concern to marketers, they may be able to highlight how the information can be used to increase one’s NFU in order to diminish the impact of privacy concerns.

 

6. LIMITATIONS AND FUTURE RESEARCH

As previous studies have shown, participants of this study might have had varying motivations for using social media [21], [36]. For example, Facebook may be used mostly to satisfy social motivation whereas Twitter may be used for information seeking purpose and for professional uses. One limitation of the findings is that the scope of this study is limited to Facebook— because various types of social media platforms are used for different reasons [36], [41], how users feel about IPC and NFU on each platform are likely to be context specific. Based on the motivation that drove users to use a particular site and their own motivation for using a particular platform, users may take a different stance about disclosing certain piece of information. Future studies need to measure motivations and/or present different scenarios of how the information might be used to parse out what drives users to share various types of content and when.

Second, this study merely asked the participants to report their general likelihood of revealing information on Facebook. The instruction did not specify what the purpose of sharing content is and who can view the shared information. Social distance, which is a representation of the distance between self and another entity, can be measured by using two separate dimensions: generality (vague vs. specific) and psychological distance (close friends vs. distant others) [42]. This framework can be used to categorize the potential viewers of personal information to observe greater impact of information privacy concern on distant and vague others. To enhance practical implications, examining a condition where marketers are the receivers of the information can be useful. Extending upon the results of the current study, future studies can conduct experiments by providing personalized content based on the each of the four categories of information identified in this current study (basic personal information, private information, personal opinions, personal photos) and measure participants’ reactions to the personalized information and privacy concerns to confirm varying levels of concerns for privacy. In some instances, consumers do show concerns over privacy that lead to falsifying at least a part of personal information [43].

Lastly, there are other factors that have been identified to influence willingness to disclose intimate information about one’s self that have not been measured or were beyond the scope of the study to control for. For example, a recent study examining users’ online privacy literacy revealed that general level of knowledge and application of privacy-enhancing tools were low to moderate [44]. Users may be proficient at using social media for personal use but may not have an understanding of to what degree personal information is collected by companies and used and, thus, do not fully recognize the risks of foregoing privacy. Therefore, measuring privacy literacy and examining its relationship with need for uniqueness may further provide insight into privacy paradox phenomenon.