Research

In a connected world, privacy becomes a group effort

"... most people think of privacy as being individualistic, but privacy is no longer just about the individual, it's also a collaborative and coordinated process," said Haiyan Jia, a postdoctoral scholar in information sciences and technology.  Credit: © iStock Photo koya79. All Rights Reserved.

SAN JOSE, Calif.  -- As the world grows more social and connects more online, privacy management is becoming more collaborative, according to Penn State researchers.

"This is a paradigm shift, in a lot of ways, because most people think of privacy as being individualistic, but privacy is no longer just about the individual, it's also a collaborative and coordinated process," said Haiyan Jia, a postdoctoral scholar in information sciences and technology.

In a study, the researchers found that social media users act autonomously on some privacy issues, but are interdependent when information is co-owned by multiple users. As soon as individuals share information with their online social networks, they no longer have sole control over the information, but must rely on others for privacy protection, said Jia.

They add that participants in the study relied on several strategies to deal with privacy rights for group content, such as group discussions and sensitive photos. Among other strategies, participants collectively specify who can access certain content, share information only within a specific group and carefully add only the people they trust to private groups, according to Jia.

"They're recognizing that some content, which could include articles that are shared in a group or a picture with multiple people tagged in it, is co-constructed," said Jia. "So, in other words, this private information can also be co-owned."

Despite the growing importance of shared data, current social media sites lack tools for collaborative privacy management, according to Jia, who worked with Heng Xu, associate professor of information sciences and technology.

"The tools that manage privacy that we reviewed were very individualistic," Jia said. "There didn't seem to be tools that were managing privacy collaboratively."

The researchers, who present their findings at the ACM Conference on Human Factors in Computing Systems today (May 11), suggested that social media developers may want to explore new ways to make content ownership more effective and transparent, as well as consider current collaborative privacy management models and tools.

"Given users' need for collaborative privacy management, next-generation social networking platforms should consider built-in mechanisms to support group effort," said Jia. "A technological solution could include offering smart suggestions for privacy settings, which balance the preferences of individual members and the cohesion of the group."

The researchers also found that people who shared more were more concerned with group management.

"If you are in a group, but don't share much, chances are you're not too concerned with these issues, but, the more you share, the more you think about co-ownership," said Jia.

The researchers recruited 304 people with an average age of 36 from Amazon Mechanical Turk, a crowdsourced online work site. The participants were asked to take a survey that lasted about 20 minutes.

To make sure the first sample was representative across different populations, researchers also recruited a second group of 427 undergraduate students to take the same survey. The undergraduate participants appeared less likely to limit who can access their shared content, which the researchers said might be a generational difference.

The researchers are planning laboratory experiments to see how people collaborate on privacy issues behaviorally.

"We're also hoping these findings raise awareness on this issue, so that other researchers begin to explore it, as well," said Xu. "Collaborative privacy management will be an even more pressing issue as the web becomes increasingly social and as big data technology becomes more prevalent."

The National Science Foundation supported this work.

Last Updated July 28, 2017

Contacts