MANAGING PRIVACY SETTINGS AND INLINE FEATURE CONTROLS IN SOCIAL NETWORKING APPLICATIONS

Abstract

Prior research has found that users of Social Networking Sites (SNSs) are often concerned about their online privacy; however, most users rarely make use of the privacy settings provided by such platforms. There are inconsistencies between privacy attitudes and observed privacy behaviours due to a lack of understanding, lack of control, abstract interfaces, and concerns about current privacy settings. Studies show that users still struggle to comprehend and control their privacy settings. In this thesis, a novel model called Privacy Settings Model (PSM) that assists users to understand, control, and update their privacy settings on SNSs is proposed. The model would enhance users’ privacy behaviours and thereby reduce their privacy risks. The research comprised of three major phases. A mixed-method study (Phase 1 and Phase 2) was conducted to explore users’ behaviours with respect to various privacy settings and the factors that impact users when learning and applying a particular privacy setting. In the first phase, a qualitative study of 22 SNS users was conducted to explore the behaviours and factors associated with privacy settings. Using thematic analysis, factors related to users’ concerns and behaviours according to the PSM components were obtained. After collecting, transcribing, and analyzing the data using the thematic analysis, 15 factors that influenced users’ behaviours when understanding and changing the settings, and receiving or finding new privacy updates were identified. Furthermore, it was observed that SNS users often rely on external resources for help and accurate information on how to understand and change their settings. In the second phase, a large-scale quantitative study of 101 users to validate the obtained factors that impact their behaviours from the qualitative study was conducted. Based on the findings in Phase 1 and Phase 2, significant design guidelines to assist in the design of usable SNS settings that meet users’ expectations were derived. In the third phase, using the design guidelines, a proof of concept application (PrivSet) was developed to help SNS users learn and configure the settings and new updates efficiently and effectively. To evaluate the system, a mixed-method study was conducted to compare the PrivSet application with the default settings in SNSs with respect to efficiency, effectiveness, and users’ satisfaction. The results of the evaluation showed that the SNS users who utilized the PrivSet application spent less time and successfully completed the process of understanding and changing the settings in comparison to the users who did not use the PrivSet application (i.e., they only used SNS settings). Additionally, the PrivSet application was well received by the users and offered promise to facilitate the steps required to achieve PSM components.