By Alyse Kupsis
Senior, Public Relations

Department Assistant Professor Lorraine Kisselburgh was curious how people with high privacy preferences handled living in a society where it is as easy as going to a Web site to find a person's address and view their home 24/7.

This past summer, Kisselburgh studied 54 Purdue students, testing their attitudes toward privacy and using data about their friendship networks on Facebook, the popular social networking site, for analysis.

She found that a typical friendship network on Facebook included anywhere from 100 to 800 friends and up to 20,000 relationships. But she found those students who indicated high privacy preferences were more likely to keep their friends compartmentalized in distinct groups. Kisselburgh said that perhaps a person had one group of friends from high school they kept separated from their college friends, for example. On the other hand, people with low privacy preferences tended to have friendship networks that were more wholly connected and were not easily separated into such distinct groups.

She said her research suggests that assumptions that people with high privacy preferences seclude themselves from others are incorrect. These individuals are socially connected to others, but they appear to use strategies that prevent their friends from mingling with one another.

The study has several implications, Kisselburgh said.

First, developers of social networks can use this information to cater to individuals who are looking for multiple ways to manage their online identity. Developers need to be able to provide features that allow individual control so that these online communities remain attractive to people who have concerns about protecting their privacy.
 
"Developers will say, 'If we don't provide that feature, they're not going to subscribe to this community,''' Kisselburgh said.

Also, the study will help in understanding human behavior offline as well as online.

Kisselburgh said her next study will collect more data about individual attributes of the participants, such as gender, major, and age, in order to draw more nuanced conclusions about how privacy influences social connectedness.