As more and more researchers in English studies take Internet texts as the focus of their work, new issues about working with texts vs. working with their authors have come to the fore. Please distinguish, though, research USING the Internet from research ABOUT the Internet. Asking people to complete your web-based survey on-line, for instance, constitutes using the Internet. Extracting data from conversations in a chat room, however, falls into the category of research about the Internet, and these studies still pose substantial issues, primarily because of subject identifiability and informed consent.
On the one hand, most Internet postings on discussion forums and stable sites are contributed by writers with a particular identity (even though it may be a pseudonymous identity). Within the Internet community, singling out that writer could cause embarrassment or other harm. So the protection provided to other known subjects, critics argue, should be provided to pseudonymous posters on the Internet. As the IRB explains:
“A human subject is a living individual, ABOUT WHOM the researcher obtains information, either through interaction/intervention or identifiable private information. In this case, [the graduate student] isn't interacting/intervening (she's passively collecting), but it is unclear with Internet stuff what is considered identifiable, as well as what people consider "private." If one is looking at people's postings to the message board that they consider constrained to their own membership (whether or not that is logical)...they might consider it private, even though it is in fact open to the entire Internet world. And then how identifiable would it be...if the community is small, how many "Jane Doe’s" could there be...or even without names, email@example.com or an IP address is pretty identifiable... and just removing names doesn't necessarily anonymize it...how many Black women full professors are there at CSU?”
On the other hand, some researchers argue that writers posting on the Internet realize that their forums are not private, and they freely contribute their writing to these community forums. How are these postings any more protected than notes tacked to a public bulletin board, which can be observed and analyzed by researchers without informed consent? In other words, as Sarah Sloane explains, researchers don’t see themselves either “interacting with or intervening with Web-based writers of what are in essence fully public and publicly accessible materials. In addition, it is clear that these writers fully expected their ‘conversations’ to be public and not private, in part because the loosely defined membership of these [electronic discussion forums] actively solicits participation from others outside their immediate community.”
The debate can become much more detailed but at its core it involves whether writers posting on the Internet should be considered subjects (with identities to protect) or authors who eschew such protections. A particularly useful resource on this issue is the Association of Internet Researchers, a multi-disciplinary group that has promulgated ethical guidelines specifically for researching Internet sites and discourse at http://aoir.org. Just click on "ethics guide" under “documents” on their home page, and you'll get a PDF file of the 30+ page document accepted by the group in 2002.
An earlier paper (1999) from the American Association for the Advancement of Science also takes up questions of human subject protection in Internet research: http://www.aaas.org/spp/sfrl/projects/humansubjects/pub
At the moment, CSU’s IRB is taking the position that such research “may be public observation of behavior (in an electronic setting), but that, too, is human research,” and an eProtocol application may be needed. Stay tuned as this discussion evolves.