(no subject)

[rivka]

It's time for me to take my annual "refresher course" in rights and protections for human research subjects. This is more or less something I can do in my sleep. Nothing changes about the Belmont Report or the informed consent requirement or the special federal protections afforded to prisoners from year to year.

But this year there was something new in the section on Internet research (emphasis mine):

One of the most controversial issues regarding Internet research involves the observation of online communications. There is currently no consensus in the research community about whether online communications in open forums constitute private or public behavior. Conclusions about whether they are public or private behavior will affect if and how the regulations are applied. If the behavior is public, then this research could be considered exempt. If, however, there is a "reasonable expectation of privacy" on the part of the subjects, then IRB review may be required.

Another issue in observing online communications is whether the individuals engaged in this communication are identifiable. Although the subjects' actual identities are not "readily accessible", for many individuals, their online identities are as important to them as their real identities. Again, whether the subjects are identifiable affects how the regulations are applied.

Researchers should consult their IRB for guidance on how these issues are applied at their institution.

I'm so glad they noticed both of these things.

I know the conventional wisdom is that you're an idiot if you think the things you say on the Internet are private. Yes, on the one hand, the public Internet is public. On the other hand, I think it is legitimate to say that in some Internet contexts, people are speaking to a particular assumed audience and may reasonably feel that their privacy is violated if the assumed audience is bent too far. No, I don't think that you have the right to rant about someone on your public blog or LJ and then complain if that person comes along and responds. But on the other hand, I think that, for example, people posting to a bereavement-focused message board expect that they are speaking to other bereaved people and would have reason to feel violated if a new poster later turned out to just be there collecting research data. It's complicated. Research guidelines acknowledging that it's complicated are an improvement.

And of course people's online identities are important. Given the number of public discussions of online activity that equate "accountability" with "using your legal name," it's good to see an acknowledgment that online identities are often stable, personally valuable, and backed up by history and reputation such that it would be damaging to discard them.

Comments

[jenett]

Very cool that they're putting that out there. (Though I wish people would stop using 'real' as the opposite of 'online', because they're really not.)

One of the Pagan boards I read gets a fair number of people coming through interested in research: they have a fairly reasonable policy about it which basically wants to make sure that if people are quoted, they're aware it's going to happen, and that people don't do 'gotcha' things (asking a deliberately inflammatory question in order to see what happens.) It's struck me as a good balance between wanting to be of use to people doing research, but recognising that the forum does not exist in order to make researchers happy or make their research easier.

[jonquil.livejournal.com]

I'm really glad this is being debated in the scholarly community; the whole Ogas debacle is fresh in my mind. (Although of course the Ogas thing would never have passed an IRB in the first place.)

Another complication, of course, is Google. If you go to a bereavement convention, what you say vanishes into the air. If you type it into a newsgroup, it can wind up Googled ten years later, when you've moved on with your life. (Ask me about my very few postings to alt.sex.bondage.)

[jonquil.livejournal.com]

One last comment -- many researchers (and indeed many people!) are unaware of how easy it is, in many cases, to discover the real identity corresponding to Internet identities. Some good computer science research has demonstrated that a quite small subset of somebody's normal interactions can identify him or her. See the AOL scandal when they released search records that they claimed were anonymized.

[antonia-tiger.livejournal.com]

I've an account on Second Life.

It's not dreadfully hard to connect it to me. For one thing, I've paid Linden Lab some money. But they make a point of saying that, legal process excepted, nobody but me has the right to reveal the RL/SL connection.

On the other hand, they seem to have a tenuous grasp of the personal significance of the names we choose to use there, if the handling of the Display Name proposal is any guide.

Oh, and I've encountered researchers. Some of them seem pretty naive about the emotional hazards of pushing at limits.

[eeyorerin]

Oh how sensible. I wish those had been in place when I was working on my dissertation, as I had some wrangles over the public/private thing with my school's IRB over what constituted public and what constituted private publication.