I think my institution's response comes from looking at cases like the
David Baltimore case. The name and reputation that people remember is the
PI. The more removed the ownership is from the person collecting the data,
the more, I would argue, the more the ability to delete or alter primary
data should be removed from the user.

I certainly understand not archiving every piece of garbage ever
collected, no more than I would suggest that every piece of agarose ever
poured should end up hermetically sealed for archival purposes. But I have
been stung too many times by what I thought was "junk" that turned into
publications down the road. So my policy has been to keep every file that
is actually saved. In fact, we make two copies of the CD's in case
something happens to one of them (theft, loss, breakage, natural disaster,
acts of God, etc). One stays available to the user and one is boxed up and
moved off site.

________________________________________________________________________________


Paul Goodwin
Instrumentation Laboratory
FHCRC, Seattle, WA

On Tue, 29 Jul 1997, Dr M Cannell wrote:

> Dear Paul,
>
> I'm not suggesting deleting important long-term archive data. I'm
> talking about deleting intermediate data that one does not have time
> too look at during the expt. It seems that we take about 100-200 Mb
> per imaging session and after about a week of this our primary hard
> disk server would be nearly full. We therefore move the data off to MO
> where it can reside for as long as one wants...
>
> Also, I would have thought that the primary data belongs to the people
> who paid for it, not the PI. This is either the institution (who
> pays the PI salary) or the grant body depending on the terms of the
> grant. I agree that it does not belong to the post-doc but it also
> probably does not belong to the PI either...
>
> Perhaps another line for discussion...
>
> Regards
>
> Mark cannell
>
> On Tue, 29 Jul 1997 09:34:33 -0700 Paul Goodwin wrote:
>
> > From: Paul Goodwin <[log in to unmask]>
> > Date: Tue, 29 Jul 1997 09:34:33 -0700
> > Subject: Re: Archiving of files - media prices
> > To: [log in to unmask]
> >
> > Our General Counsel has looked at the issue from the Patent
> perspective.
> > His conclusion is 20 years for primary data. This is the time that
> he
> > thinks would be necessary for retention of data needed to defend an
> > intellectual property suit and prior art issues. Since imaging data,
> like
> > ratio imaging, is the only primary data, our Counsel would require
> that we
> > save it for 20 years. Since we are a Shared Resource providing a
> service
> > to other labs, we feel that it is our responsibility to do whatever
> we can
> > to retain the data for at least 20 years. That means that not only
> must
> > the media be readable for that time, but that readers/drivers must
> exist
> > to permit us to access the data 20 years from now. Right now, I am
> in the
> > process of migrating 50GB of Optical Disk data to CD. To me, the
> cost of
> > replicating the experiment far exceeds the cost of any digital
> media, so
> > why would you ever consider deleting files from an archive
> intentionally?
> >
> > The other issue that we are trying to address is ownership of data
> and
> > data integrity. We are in the process of changing our archive system
> so
> > that users do not have write permission for their original data. The
> data
> > belongs to the PI and the Center where I work takes science ethics
> very
> > seriously. The data ultimately belongs to the PI, not the user. To
> permit
> > the user to have write access to the files places in jeopardy the
> > financial investment and integrity of the PI and the Center. I would
> much
> > rather drop another $0.08/MB on archiving than have to explain how
> and why
> > my resource could have been used to generate fraud.
> >
> > It is a sticky issue and one that I have spent considerable time the
> past
> > three years trying to solve.
> >
> >
> ______________________________________________________________________
> __________
> >
> >
> > Paul Goodwin
> > Instrumentation Laboratory
> > FHCRC, Seattle, WA
> >
> > On Tue, 29 Jul 1997, Phil Allen wrote:
> >
> > > Hi All;
> > >
> > > Mark Cannell raises an interesting question.  When are images
> "Junk" and
> > > can be discarded?  We're actively doing ratio imaging and the
> question of
> > > when we can discard "Junk" data constantly comes up.  What do
> people on the
> > > list think?  TIA
> > >
> > > Phil Allen
> > >
> > > ________________
> > > Philip G. Allen; Ph.D.
> > > Instructor in Medicine
> > > Brigham and Women's Hospital
> > > Division of Experimental Medicine
> > > LMRC 301
> > > 221 Longwood Ave.
> > > Boston, MA 02115
> > >
> > > 617-278-0321 (lab)
> > >        -734-2248 (fax)
> > >
>