Archives

Yang Wang, Pedro Giovanni Leon, Kevin Scott, Xiaoxuan Chen, Alessandro Acquisti, and Lorrie Faith Cranor, Privacy Soft-paternalism: Facebook Users’ Reactions to Privacy Nudges

Yang Wang, Pedro Giovanni Leon, Kevin Scott, Xiaoxuan Chen, Alessandro Acquisti, and Lorrie Faith Cranor, Privacy Soft-paternalism: Facebook Users’ Reactions to Privacy Nudges

Comment by: Andrew Clearwater

PLSC 2013

Workshop draft abstract:

Anecdotal evidence and scholarly research have shown that a significant portion of Internet users experience regrets over disclosures they have made online. To explore ways to help individuals avoid or lessen regrets associated with online mistakes, we employed lessons from behavioral decision research and soft- paternalism to develop three Facebook interfaces that “nudge” users to consider the content and context of their online disclosures more carefully before posting. We implemented three nudging interfaces: profile picture, timer, and timer plus sentiment meter.

The picture nudge was designed to remind Facebook users of which individuals are in the audience for their posts. Depending on the particular post privacy setting, users were shown five profile pictures randomly selected from the pool of those who could see their posts. These profile pictures appeared under the status-updates and comment text boxes when users started typing. The timer

nudge was designed to encourage users to stop and think. The warning message

“You will have 10 seconds to cancel after you post the update” with a yellow background was displayed under the status-updates and comment text boxes when users started typing. After clicking on the “Post’” button, users were given the options to “Cancel” or “Edit” their post before it was automatically published after 10 seconds. The third nudge added a sentiment meter to the timer nudge, and the content of each post was analyzed by our sentiment algorithm. This nudge was designed to help make users more aware of how others might perceive their posts. For posts with a positive or negative score a warning message “Other people may perceive your post as {Very Positive, Positive, Negative, Very Negative}” was displayed during the countdown timer.

We tested these nudges in a 3-week field trial with 21 Facebook users, and conducted 13 follow-up interviews. By triangulating system logs of participants’ behavioral data with results from the exit survey and interviews, we found evidence that the nudges had positive influences on some users’ posting behavior, mitigating unintended disclosures. We also found limitations of the current nudge designs and identified future directions for improvement. Our results suggest that a soft-paternalistic approach to protect people’s privacy on social network sites could be potentially beneficial.

Pedro Giovanni Leon, Justin Cranshaw, Lorrie Faith Cranor, Jim Graves, Manoj Hastak, Blase Ur and Guzi Xu, What Do Online Behavioral Advertising Disclosures Communicate to Users?

Pedro Giovanni Leon, Justin Cranshaw, Lorrie Faith Cranor, Jim Graves, Manoj Hastak, Blase Ur and Guzi Xu, What Do Online Behavioral Advertising Disclosures Communicate to Users?

Comment by: Mary Culnan

PLSC 2012

Workshop draft abstract:

In this paper we present the results of a large online user study that

evaluates the industry-promoted mechanism designed to empower users to manage their online behavioral advertising privacy preferences.  700 [we expect about 1200] participants were presented with simulated behavioral advertisements in the context of a simulated and controlled web-browsing session. Subjects were divided into conditions to test two online behavioral advertisement disclosure icons, seven taglines (including no tagline), and five opt-out landing pages. Following this simulation, we surveyed the users about their understanding and perception of the OBA notification elements that they saw. Our [preliminary] results show that users often do not notice the icons or taglines, and that the industry-promoted tagline does a poor job of communicating with users. On the other hand, users found many of the opt-out landing pages to be informative and understandable.

Our results show various levels of effectiveness of disclosure taglines across three dimensions: clickability, notice, and choice. We found that only out about a quarter of participants ever recalled having seen the taglines, and that no tagline was effective at communicating all three concepts to users. We also found that “AdChoices,” the current tagline promoted by industry groups, was among the least communicative of the taglines we tested.  Conversely we found that the tagline “Why did I get this ad?” which has recently been adopted by the Google AdSense Network, performed well at communicating clickability and notice. Our work suggests that taglines that suggest an action are more effective at conveying the clickability of the link, which is a critical aspect of the disclosure, allowing users to seek more information or configure their OBA preferences. Furthermore, although none of the symbols was particularly effective at providing notice and choice, we found that the symbols are important at communicating clickability. In particular, the poweri symbol better conveyed clickability than the asterisk man symbol.

We tested the opt-out landing pages provided by AOL, Yahoo!,

Microsoft, Goolge, and Monster [and may test a few more].  All but the Monster Career Ad Network were perceived as informative and

understandable. AOL and Microsoft opt-out pages were shown to be more effective at encouraging users to opt out.