At major academic survey centers, standard practice is to place surveys in
an archive that is widely available to other scholars. These places usually
have a policy that covers privacy and timing of the release of information.
This permits expensive data collection to be used by other scholars,
possibly for entirely new questions and analyses.
Some longitudinal data from the Tecumseh study at Michigan has been placed
in the archive of ICPSR. This was a 30 year study of cardiovascular disease
in Tecumseh, Michigan, similar to the Framingham study. I don't know how
they dealt with the issues of privacy.
Making data more widely available is surely possible. It may not be exactly
the same data the original researchers have, but many privacy issues can be
overcome, by such methods as suppressing exact dates of event.
In biological sciences, replication of results is an important issue. Tests
of replication may require more data than the original authors used in a
particular analysis, if the issue goes beyond mere numerical accuracy into
issues of whether useful covariates were excluded.
David Smith
David W. Smith, Ph.D., M.P.H.
(518) 439-6421
45 The Crosway
Delmar, NY 12054
[EMAIL PROTECTED]
----- Original Message -----
From: "Robert J. MacG. Dawson" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Friday, March 08, 2002 9:52 AM
Subject: Re: NIH draft on publishing data
>
>
> Rich Ulrich wrote:
> >
> > NIH published this draft regulation concerning research.
> > Comments are being solicited.
> >
> > For the research that I know of, this seems more-than-a-little
> > bit dumb. Even for the biology which is not too complicated,
> > I think of serious problems arising (Needleman's lead research)
> > after outsiders were allowed controlled access -- despite there
> > being strong oversight while those issues were debated. Freer
> > access seems scary.
>
> Can you expand on this? Are we talking about privacy issues,
> intellectual property issues, limitation of participation to an approved
> in-group, or what?
>
> I have to admit that my own first reaction is that the idea of
> important decisions being based on conclusions drawn from data that is
> not available to other researchers ("but trust me, the data really does
> exist and I analyzed it perfectly, and the statistical technique I used
> is never going to be supplanted by a better alternative") is much more
> scary.
>
> -Robert Dawson
> .
> .
> =================================================================
> Instructions for joining and leaving this list, remarks about the
> problem of INAPPROPRIATE MESSAGES, and archives are available at:
> . http://jse.stat.ncsu.edu/ .
> =================================================================
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
. http://jse.stat.ncsu.edu/ .
=================================================================
