Privacy in Context: Understanding User Privacy Behavior

Download data and study materials from OSF

Principal investigator:

Denise Anthony

Dartmouth College



Sample size: 1341

Field period: 2/12/2012-07/16/2012

Information technologies (IT) have rarely been designed with privacy in mind. Researchers have found that many users seem quite willing to give up privacy for access to relatively low value goods, whether grocery cards, social network sites or online products, suggesting that users do not value privacy. Some experimental research has shown that users seem not willing to pay very much to protect their information (Danezis et al 2005; Grossklags & Acquisti, 2007). Others, however, suggest that users will change behavior when the context of information sharing varies (Nissenbaum 2010). For example, studies of location tracking show that at least some users will vary their willingness to share depending on place and social context (Anthony et al 2007), as well as recipient (Consolvo et al 2005). Other studies of context show that users are more likely to reveal information when the reward from the exchange increases, but less likely to do so when risk of identity theft increases. Further, users were often willing to pay for privacy when given the opportunity to do so (Baumer et al 2005). Other studies find that users will utilize controls to limit information exposure when usable options are available (e.g., Caine et al 2010). This study proposes to examine how privacy behavior varies under conditions that test aspects of the context of information exchange. The study separates the choice to use a mobile-application from the choice to share information with the app.

Does willingness to use and/or share information with a mobile application-based service vary by type of privacy control policies (e.g., opt-in versus opt-out)?

Does subject level of resources effect willingness to share information?

Does willingness to share information vary by type of information to be shared (health versus location) with the mobile app-based service?

Does willingness to share information vary with the type of service (specifically who benefits, individual or public)?

Experimental Manipulations
Privacy Policy: None, opt-in, opt-out, Privacy-Rights Management

Amount of resources: None, Low, High

Type of information: health, location

Mobile application service: private benefit, public benefit

Use the mobile-application service (yes-no)

Share information with the mobile-application service (yes-no)

Summary of Results
The complicated design makes analysis also complicated. For the simplest condition (No resource), we generally find that Use and Sharing are higher for location compared to health information, and when the service has an individual versus a public benefit.
Generally, Sharing is lowest in the Opt-In policy condition. Use does not vary significantly by policy condition.

Additional Information
Analysis of the data is ongoing.


Sacrificing customers in the name of data? How online privacy policies affect purchase decisions," with Kelly Caine, Sameer Patil, Celeste Campos-Castillo. Paper presented at CITASA Symposium, Berkeley CA, August 2014.