reproducible research in the neurosciences

Lester Ingber ingber at ingber.com
Thu Apr 11 08:10:09 EST 1996


In article <4khmhe$qsd at eis.wfunet.wfu.edu>,
Mark Laubach <laubach at biogfx.neuro.wfu.edu> wrote:
:bill at nsma.arizona.edu (Bill Skaggs) wrote:
:
:>Even if the organizer tries to write a detailed
:>description of every possible problem with the data -- often a very
:>tedious task -- the downloader won't necessarily read it or understand
:>it.  The result is likely to be at least a few papers drawing
:>earthshaking theoretical conclusions from sophisticated analysis of
:>experimental artefacts.  This is what needs to be avoided.  I'm not
:>saying it's impossible, but it isn't all that easy.
:
:If something like this were to happen then it seems to me that the
:person making the claim would have more to loose than the person who
:collected the data.  Also, I would hope that anyone using someone
:else's data for the purpose of publication or re-analysis would make
:the data "owner" aware of their intention and would collaborate with
:them on the work.  But this may just be optimistic thinking on my
:part.
:
:Mark Laubach

Why all this unreasonable concern with data!?  That is, look at any
journal in any science, and you will see people using techniques from
other researchers, across  other disciplines and fields of expertise
than their own, e.g., from other mathematical, physics, biological,
experimental, computational, etc., fields.  Yes, they often misuse
these techniques, but that is what peer review is about, even given its
defects especially when applied to interdisciplinary research.  In
fact, I think a good argument can be made that this is more the problem
these days with sophisticated computational techniques than it is with
data.

I find appalling the arguments that raw data can be used incorrectly or
that the data often is tenuous and requires the tender loving care of
the original investigator to properly massage into correct results.
These arguments are not based on any scientific concerns; raw data is
raw data, and that often is its virtue.

No one should be considered a coauthor unless they contribute to a
paper.  Certainly, no one should should be a coauthor on a paper,
solely for the use of data that has been used in previous
publications.

I agree with the sentiments expressed here, that published work should
be reproducible.  This includes the use of data, computational
algorithms, codes, etc.  Otherwise, the choice is simple:  Don't
publish, get a patent, open a business, and let the marketplace be your
source of critical review.  If you want to add a paper to your list of
publications, then you should admit the journal (usually) owns the
copyright, and the world (always) has the right to reproduce your results.

Lester
-- 
/*             RESEARCH                                 ingber at ingber.com *
 *       INGBER                                      ftp://ftp.ingber.com *
 * LESTER                                          http://www.ingber.com/ *
 * Prof. Lester Ingber _ P.O. Box 857 _ McLean, VA 22101 _ 1.800.L.INGBER */




More information about the Neur-sci mailing list