James Kirkley wrote:
> I am getting in on the data discussion quite late but I think the data
> access problems are perhaps overblown. Some US agencies have incurred
> substantial cost and time to make data available from a webb site. You can
> now download a tremendous amount of commercial and recreational data from
> NMFS in a minimal amount of time. Data are also available from the National
> Weather Service and the U.S. Dept. of Agriculture as well as several other
> federal agencies. FAO also has quite a bit of data available. It sounds as
> though the data problems are specific to certain geographic regions of the
> world.
In many instances, you can either electronically access much of the
data. SEAMAP is a good example in the Southeast US; collected by a
variety of researchers throughout the southeast US, these data are
computerized in a standard format and archived in a single location so
that the compilation can be accessed with ease.
On the other hand, although some long-ter databases are available, the
format may vary over time or other variable, or the material is in a
form that the receiver must spend considerable time trying to make
"heads or tails" out of the format (usually requiring actual visits to
the lending institution to meet with those who computerized the data).
In this latter instance, the receiver of these data may make some
"unavoidable" assessment errors simply because of their unfamiliarity
with the database (as was pointed out yesterday).
I don't have an answer, but these kinds of problems have only recently
needed address because of the new advances in computerizing data and
making it generally available (vs. working from paper copies).
Currently, standardized sampling formats are being developed for new
broad-scope research programs, and even smaller scale projects are often
requested to follow a designated format so that the data will be
compatible.
>
> Regarding researcher collected data. This is a fuzzy area.
> [snip] > Data access really comes down to common courtesy.
I totally agree, and the [snipped] caveats to sharing data are usual for
academics (coauthor, student use, etc.). I have found several instances
in time however where a researcher used their database to make specific
claims, but then refused to allow others to see more than the final
results. Of course, idealistically, when one makes a claim (based on
their research) I suppose we should accept that these claims are in fact
valid, but alas, I fear that in our current world, this is not often the
case. Thus, it is becoming more common for dissenters of those claims
to request a more detailed look at the data itself; it is unfortunate
that a researcher need to "back-up" his claim, but it is becoming more
of a necessity. On the other hand, if the claim is valid, then what
does the researcher have to fear by releasing the data that generated
the claim?
Lastly, yes, public funds do support most research, and most granting
agencies do now require that the raw data be provided for archival (or
use). However, there needs to be some mechanism to guarantee the
researcher's proprietary rights to that data for a given period (except
perhaps for agency use i.e. for management purposes, not staff
publications). This might be one year following the close of the
overall program (say a 3-yr grant - if the researcher is planning a
20-yr program that is another thing, but I think they will need to use
some the data for publication before 20 years is completed); this allows
the researcher time to "crunch" all the numbers, and work up a
manuscript for submission for publication. At that point, public acess
should be allowed - a receiver of such data couldn't "scoop" the author
and get a _recognized_ publication out earlier. If the original
researcher doesn't meet the 1 yr deadline, then (s)he is more at fault.
--
=======================
Steve Branstetter, Ph.D., Program Director
Gulf and South Atlantic Fisheries Development Foundation
Ste. 997, Lincoln Center
5401 W. Kennedy Blvd., Tampa, FL 33609
Phone: 813-286-8390 FAX: 813-286-8261
email: [log in to unmask]
|
