The Persuasive Effect of Privacy Recommendations for Location Sharing Services

17 Pages Posted: 23 Feb 2014

See all articles by Bart Knijnenburg

Bart Knijnenburg

Clemson University

Hongxia Jin

Samsung Research America

Date Written: October 6, 2013

Abstract

Several researchers have recently suggested that in order to avoid privacy problems, location- sharing services (LSS) should provide finer grained methods of sharing one’s location. This may however turn each “check-in” into a rather complex decision that puts an unnecessary burden on the user. This paper presents two studies that explore ways to help users with such fine-grained location-sharing decisions. In study 1 (N=100), we show that although users differ wildly in their preferred sharing actions for any given scenario, users’ evaluation of the activity described in the scenario is a good predictor of the sharing action they choose. In study 2 (N=368), we develop several “privacy recommenders” that tailor the list of sharing actions based on the user’s evaluation of the activity. Testing these privacy recommenders against a static system, we find that they have a strong persuasive influence on the user, and that a recommender that shows a short list of recommended actions and hides all other actions is perceived to be helpful. We also find, however, that a system that does not ask the user to evaluate the activity is ultimately more satisfying.

Keywords: privacy, information disclosure, decision-making, personalization, recommender systems, user experience

Suggested Citation

Knijnenburg, Bart and Jin, Hongxia, The Persuasive Effect of Privacy Recommendations for Location Sharing Services (October 6, 2013). Available at SSRN: https://ssrn.com/abstract=2399725 or http://dx.doi.org/10.2139/ssrn.2399725

Bart Knijnenburg (Contact Author)

Clemson University ( email )

101 Sikes Ave
Clemson, SC 29634
United States

Hongxia Jin

Samsung Research America ( email )

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
129
Abstract Views
709
Rank
397,606
PlumX Metrics