Adjustments, applications and other tools allow users to configure preferences and use services such as iGoogle and Yahoo to control and customize the news they consume online. These tools can make online experiences more efficient and productive, but they do not ensure that users will be consistently pleased with their selections, according to Penn State researchers.
"There is a major push toward customization in the marketplace because designers assume that more customization is better, but our research shows that only some users prefer customization," said S. Shyam Sundar, distinguished professor of communications, whose research with Sampada S. Marathe of the Media Effects Laboratory in the University's College of Communications appears in the current issue of Human Communication Research.
Two recent studies found that privacy concerns limit how both the most technology-savvy users and those with less expertise use opportunities to customize or personalize their online experiences.
For purposes of the study, researchers defined "customization" as a more proactive, highly user-driven practice, and "personalization" as having the system tailor content for users without active user input.
In the first study, the researchers discovered that power users -- those having higher levels of comfort with technology and interest in controlling their experiences -- found their visits to an online news site more enjoyable when they could customize the search process by defining search parameters or making changes to a website's settings themselves.
Conversely, nonpower users -- those less comfortable with technology -- enjoyed personalized experiences the most. Under those parameters, the site shaped the news it provided without any overt control by the users themselves. It offered news based on user behaviors while browsing and searching during a previous visit.
"Power users like to control their information universe," Sundar said. "So they like news when they customize it themselves. But regular or ordinary users of the Internet like it better when the system configures the news for them.
"We all like things tailored to fit us -- even our clothes -- but we'd rather not do the tailoring ourselves."
The follow-up study indicated that power users might prefer customization out of a concern for their privacy. As part of that second study, users were notified that the news site they visited either "may use" or "will not use" their browsing information to provide services they requested. This subtle difference in notification resulted in dramatic changes in user behaviors.
Specifically, in the high privacy environment, power users were more willing to cede control and have information tailored for them because they trusted the site and appreciated the convenience. In a lower privacy setting, those same users wanted more control.
"After our first study, we were wondering why power users were rejecting the convenience of the system personalizing news content for them," Sundar said. "Now we know that it is because personalization implies the system is snooping on you.
"With customization, users not only maintain control over the content they consume, but also control over how the system provides them the tailored content."
Nonpower users were more conscious and cautious of their browsing efforts in lower privacy situations. They were more likely to self monitor their news consumption and limit their online exploration as a result. But, they prefer content customized by them over that personalized by the system when assured privacy.
The concern for privacy has dogged many emerging communication technologies, including online social networking sites and several mobile applications.
"As the technology gets more intrusive, the savvy users are the ones who are able to negotiate the technology to their advantage," Sundar said. "Most of the rest of us either complain or reject the technology."
- S. Shyam Sundar, Sampada S. Marathe. Personalization versus Customization: The Importance of Agency, Privacy, and Power Usage. Human Communication Research, 2010; 36 (3): 298-322 DOI: 10.1111/j.1468-2958.2010.01377.x
Cite This Page: