An empirical analysis of users’ privacy disclosure behaviors on social network sites

Abstract

Users’ privacy on social network sites is one of the most important and urgent issues in both industry and academic fields. This paper is intended to investigate the effect of users’ demographics, social network site experience, personal social network size, and blogging productivity on privacy disclosure behaviors by analyzing the data collected from social network sites. Based on two levels of disclosed privacy sensitivity information, the textual information of a user's blog postings can be converted into a 4-tuple to represent their privacy disclosure patterns, containing the breadth and depth of disclosure, and frequencies of highly and less sensitive disclosures. Collections of a user's privacy disclosure patterns in social network sites can effectively reflect the user's privacy disclosure behaviors. Applying the general linear modeling approach to blogging data converted with a coding scheme, we find that males and females have significantly differentiated privacy disclosure patterns in dimensions related to the breadth and depth of disclosure. In addition, age has a significant negative relationship with the breadth and depth of disclosure, as well as with highly sensitive disclosure. We also find that social network site experience, personal social network size, and blog length are not significantly related to users’ privacy disclosure patterns, while blog number always has positive associations with privacy disclosure patterns.

Introduction

Social media are a group of Internet and mobile based applications build on Web 2.0 technologies in which people can create, share, or exchange user-generated content. Currently, social media is becoming increasingly important in our daily life and has received substantial scholarly attention [46]. Social media outlets rely on the Internet and mobile technologies to provide interactive platforms for information dissemination, content generation, and interactive communications [65]. An array of Internet and mobile-based applications define the way social media functions. Examples include weblogs, microblogs, online forums, wikis, podcasts, life streams, social bookmarks, web communities, social networking, and avatar-based virtual reality [2], [18], [69]. Based on these applications, social network sites (SNS) have gained tremendous momentum, revolutionizing the way individuals build and maintain interpersonal relationships [1], [70].

Because people are increasingly incorporating SNS as a part of their routine social activities, the number of SNS users has grown exponentially in recent years. Every minute, terabytes of user-generated content are posted by millions of users on various social network sites, such as LinkedIn, Facebook, Twitter, QQ, etc. In this way, SNS, underpinned by social media, become versatile resources for both industry stakeholders and scholars to study the enriched and dynamic data contributed by a wide range of users in the social network sphere [34].

In the past decade, research has focused largely on the adoption and usage of SNS [33], [71], as well as the management of social relationships on SNS [17]. Limited work has covered the topic of privacy disclosure. In fact, privacy disclosure on SNS is becoming one of the most important and active research issues in the information systems arena [25], as the influx of user-generated content into various SNS has resulted in major concerns over the misuse of these data. Meanwhile, the Web 2.0 era demands data openness for all kinds of innovative online businesses, and as a result, increasing privacy information disclosure. Openness would essentially help to reduce the uncertainty of interactions, legitimate access to a person in an online group, and ultimately promote online business [39], [62].

Personal information exchange between users on social network sites allows these people to maintain relationships with friends, develop new friendships, and find support and information [62]. Privacy disclosure on SNS has a negative side, however; greater disclosure may correspond also to information theft, trafficking, and privacy invasion. Gross and Acquisti [21] found that users effectively place themselves at a greater risk for cyber and physical stalking, identity theft, and surveillance when they disclose personal information on SNS. As such, concern over the negative effects of privacy disclosure has a major influence in users’ adoption and routine use of SNS [71]. Therefore, this paper is intended to examine the predictability of privacy disclosure behaviors on SNS in terms of the privacy sensitivity levels of information to be disclosed (in short, we will use “sensitivity levels of information” implying the context of privacy). We will investigate the differences between users’ disclosure of highly sensitive information and less sensitive information in relation to SNS structure and mechanisms. These are particularly important questions for SNS industry service providers; hence providing insight into this theme can prove useful in practical applications such as interface design and SNS privacy policy.

Our study contributes to the literature by focusing on predictors of SNS privacy disclosure. Communication Privacy Management theory is introduced as a framework to clarify the influence of users’ gender, age, social network site experience, personal social network size, and blogging productivity on their privacy disclosure behaviors. Specifically, privacy disclosure is divided into two dimensions: breadth and depth. Highly sensitive disclosure and less sensitive disclosure are also distinguished in the study. To test these models, we collected practical data from one of the most popular social network sites. The results clarified various predictors of SNS privacy disclosure and offered insights into the social implications of SNS.

The remainder of this paper is structured as follows. Related literature about privacy disclosure on social network sites is reviewed to provide the theoretical background and foundation for our study in Section 2. We describe the research methodology of this paper in Section 3, including data, variables, and models. Empirical results are presented in Section 4. Finally, we conclude the paper and suggest future research directions.