[Air-l] Science News Online: The Social Net

Jim Hudson jhudson at cc.gatech.edu
Thu Jun 6 10:27:31 PDT 2002

Since I'm not sure that Joe and I substantively disagree, I just want to
clarify my comments.  My main points about publicly archived data is
that, if we work from the position that this is human subjects research,
we must deal with issues of consent.  Being unable to reach a poster
does not excuse the researcher from the obligation to obtain consent.
If we work from the position that it is not human subjects research,
however, we must deal with a different set of issues.  Namely, crediting
the source.  As Joe points out, it is certainly fair use for a
researcher to use this information, but "fair use" also requires
crediting the source of information obtained.

To me, however, neither of these positions are really comfortable.
These hypothetical archives are certainly publicly available.  So,
there's a strong foundation for saying that this is not human subjects
work.  But, as research by Joe and others has suggested, people often
feel freer to say certain things online than they would in other media.
(Of course, the devil is in the details of that statement.  Suffice it
to say, online interaction is often different than other media.)  The
environment, in some ways, leads people to be less aware of the
consequences of their actions.  As the Deja News situation showed,
people are often shocked by how public their actions online are, even
though they shouldn't be.  I believe that since the Internet is so new
to the common user, we cannot just assume that Internet users are truly
consenting to the public archiving of various forums.  

That said, I don't know what the answer is.  I tend to think that the
answer lies somewhere between these two perspectives.  As researchers,
we have a right to use publicly available data, but we also have an
obligation to protect subjects, even sometimes protecting them from
themselves.  So, I guess it probably comes down to the potential for
harm.  There's probably little harm in most linguistic work, but much
more in certain sociological studies.  Maybe the necessity for consent
with publicly available material should increase as the potential for
harm increase?  Just speculating...


Jim Hudson              |  Ph.D. Student
College of Computing    |  jhudson at cc.gatech.edu
Georgia Tech            |  http://www.cc.gatech.edu/~jhudson
Atlanta, GA  30332-0280 |  Phone: 404-894-9761
                        |  Fax:   404-894-0673

More information about the Air-L mailing list