How not to conduct research: Online ethics edition

Note: Everything in the following article and the provided links (at least at the time of posting) is work-safe, though some links may contain explicit language.  However,  please exercise caution in clicking other links found on the web pages referenced here! A quick Google search for recent Boston University grads Ogi Ogas and Sai Gaddam reveals buzz about their book, “A Billion Wicked Thoughts,” out today.  Ogas and Gaddam analyzed traffic to erotic web sites and millions of web searches to discover, in the words of Amazon’s product description, “ a revolutionary and shocking new vision of human desire that overturns conventional thinking.”  The New York Post, Huffington Post, and Newsweek among others have all picked up the story.  After all, who doesn’t love the science of erotica?

It turns out, some individuals are less than thrilled about Ogas and Gaddam’s research.  Allow me to awkwardly transition to the real focus of this post: the ethics of conducting research online.

The Internet is a vast, untamed wilderness of human experience. Both the mundane-- your former roommate’s Twitter update-- and the titillating-- and here I will refer you to the authors’ book-- exist in equal measure.  For psychologists, the Internet is a vast field site, the world’s largest and most permanent collection of case studies and large-scale observational studies just waiting to happen.

But any field site, whether in the African bush or a message board for disillusioned hackers, must be treated with caution and respect.  How does an ethnographer conduct his or her research without disturbing the existing culture or doing a disservice to its members?  The American Anthropological Association has one set of suggestions.  Various academics have suggested ways to bring research ethics into the digital age. Nonetheless, when it comes to studying the culture of the Internet (and if you don’t believe the Internet has a culture, lolcats and Rick Astley beg to differ-- and those are just the examples I can show you on a family-friendly site), researchers struggle with the most ethical approach.

While there are no simple answers, we do have a few examples of “What Not to Do.”  For better or for worse, Misters Ogas and Gaddam have provided one of the most enduring examples of online ethnography gone wrong.  That’s right... before Ogas and Gaddam presented us with every search engine vocabulary term you thought you wouldn’t read in Barnes and Noble, they had quite a different legacy: they were the perpetrators of SurveyFail 2009.

It began with a few innocent emails: Ogas contacted members of online fanfiction communities, particularly communities devoted to writing in the more explicit style popular in the genre. (Note that those links are Wikipedia).  The fanfiction writers responded with a “thanks, but no thanks:”

...Your project sound like ...one of the worst kinds of projects dealing with fandom and fanfiction...

...Studies of fans, particularly female fans, tend to follow in the long history of pathologizing women's behaviour and women's desire, the history of male scientists objectifying queer/female desires in order to subjectivize themselves...

...there is still a note – a note that is audible in your brief message, in your "fascination" with [our site] – of a nineteenth century scientist with a particularly interesting bug under the microscope. We're not interested in being your bug....

source.

Not to be deterred, Ogas and Gaddam nonetheless posted the survey, using advertisements on message board and online journal sites to attract members.  And here is where things began to go downhill.  Unwitting community members took the survey, and were angered by its content and the suggestions implied by the questions.  They warned friends not to take the survey, suggested that participants sabotage their own responses and, when Ogas and Gaddam continued to distribute the survey, got a little more creative, immortalizing the researchers in a, ahem, intimate fiction of their very own.

Finally, the researchers relented.  They removed the survey and posted an apology on their online journal before deleting the journal entirely (not that this could erase the damage-- you can still read about the whole event in more detail here, here and here, among other places).

Ogas and Gaddam clearly got more than they bargained for, since they abandoned the survey, and moved on to-- well, probably to their new book.  Unfortunately, not every community, either online or offline, would necessarily be able to mobilize in the way we saw here.   And this is why it is the researcher’s responsibility to, in the words of the AAA code of conduct “ do everything in their power to ensure that their research does not harm the safety, dignity, or privacy of the people with whom they work.”

Some ways to accomplish this include: Be transparent with research aims and goals. Participants were upset at the nature of the questions in the survey, indicating that when they provided informed consent, they were not made fully aware of the researchers’ aims.  Obviously under certain circumstances it may be necessary to avoid discussing hypotheses or research aims in a consent form, but when these aims will be made clear in the body of the survey, both the researcher and participant are better served if this information is available up front. Rely on university affiliation and the support of colleagues. Ogas and Gaddam worked alone, without university affiliation and an IRB (in fact, some sources suggest that after receiving a number of complaints, Boston University requested that the researchers stop using their BU email addresses to correspond with participants).  Researchers that work alone have no sounding board or opportunities to seek advice from other colleagues with similar backgrounds.  Furthermore, there is no one to review their procedure and make sure they are conducting research in a safe and sensitive manner. Respect the community’s wishes, especially for privacy. Here, perhaps, is where the “SurveyFail” truly got its name: despite repeated requests, Ogas and Gaddam continued to disseminate their survey.   While they eventually backed down, it was not before angering an untold number of participants and community members.

None of this is to say that Ogas and Gaddam are bad people, that they should be banned from the academic community or stripped of their PhDs.  Researchers and scientists make mistakes... in fact, Institutional Review Boards allow researchers to make amendments to their original study plans and have precautions in place in the event of a breach of privacy.  I would, however, be skeptical about the findings in their new book.  And I do hope that their experiences with the fanfiction community will serve as a cautionary tale to other researchers.  Online research is a new way to examine human experience, and rules of conduct have not been firmly established the way they have in other contexts.  Online data collection is cheap and relatively easy, and the sheer amount of information available is astounding.  But no matter the environment, the well-being and dignity of participants and their community must always come first.