Tuesday, June 15, 1999

On-Line Research With Human Subjects Deserves More Scrutiny, Scientists Say By FLORENCE OLSEN Washington On-line studies with human subjects involve risks and benefits that committees overseeing universities' research policies need to take into account, participants said at a science workshop here last week. "I'm astonished there have been so few problems with Internet research," said one sociology professor. Federal regulations written in the 1980s may need to be revised because the Internet offers social-science researchers new venues, they said. At the very least, the committees -- known as institutional review boards -- may need information-technology checklists to help them review the growing volume of Internet-based research proposals demanding their attention, the scientists said. The American Association for the Advancement of Science and the Office for the Protection from Research Risks of the National Institutes of Health sponsored the two-day meeting, whose purpose was to identify ethical and legal concerns for human-subject researchers in cyberspace. Mark S. Frankel, director of the Scientific Freedom, Responsibility, and Law Program at the A.A.A.S., said the association would publish a full report on the topic in the fall. Melinda S. Bier, a program director for the James S. McDonnell Foundation in St. Louis, argued that proposals for conducting human-subject research in cyberspace must disclose the potential benefits and risks to human subjects. Researchers who use the Internet "are obliged to educate themselves regarding these issues," she said, so they can insure participants' privacy and informed consent. The McDonnell Foundation gives grants for biomedical and behavioral-sciences research. As universities acquire data-warehousing and data-mining technologies, research data may wind up being used in ways participants and researchers never imagined, said John Kennedy, director of the Center for Survey Research at the University of Indiana at Bloomington. Human beings could find themselves "unwitting subjects of research" to which they never consented, he said. Amy Bruckman, an assistant professor in the College of Computing at the Georgia Institute of Technology, proposed that researchers apply the same rules of confidentiality to Internet pseudonyms that they use with real names. To the individuals who created them, pseudonyms are oftentimes as "real" as their actual identities -- and professional researchers should respect that, Ms.Bruckman said. "I'm astonished there have been so few problems with Internet research," said Jim Thomas, a sociology professor at Northern Illinois University. Human subjects in cyberspace are not the only people put at risk by such research. Social-science researchers themselves can be the targets of libelous attacks published on listservs, said Craig Childress, president-elect of the International Society of Mental Health Online, in LaVerne, Cal. "Subjects have your e-mail address and can slam you -- and even shut you down" with denial-of-service attacks, which broadcast a flood of messages that the recipient's e-mail system can't handle. Background story from The Chronicle: Head of NIH Seeks to Move Unit That Oversees Research With Human Subjects (6/11/99)