PDA

View Full Version : sloppy data and peer review



William L. Siler, Ph.d.
09-23-1994, 12:35 AM
Dear Colleagues:

After talking with people on other lists and considering the
ideas shared by others on this list, I have more questions.
Start by considering the following thoughts from others on the
list:


From: IN%"dluedtke@admin.stkate.edu" "Dave Luedtke"

Concerning "sloppy data"; it would seem one would need to define
the minimum acceptable information to be collected when studying
each topic. If a person collected data using a limited amount of
equipment and were not able to generate all possible information
on that topic, some might consider their results as sloppy
because they did not use the most elaborate equipment available
and could not due to budget restrictions.

************************************************** ***************

From: IN%"keith@cvs.rochester.edu"

Thanks Bill for your personal note on scientific misconduct. I
agree that there is bad work done. But the peer review process
in grant competition and in publication seems to take care of a
lot of it. I think we also all know which journals we respect
more than others (based on the stringency of the review process
and the quality of what is published). We also are skeptical -
as all scientist should be - when reading sloppy work, and
therefore don't cite it. All in all, I think sloppiness is a
difficult thing to judge. I actually think it is good to have
journals and researchers to publish in those journals at
different levels of "sloppiness". By that I mean applied work
such as that published in "Applied Ergonomics" or clinical case
studies such as that frequently published in medical journals
cannot have the same scientific rigor as more basic research
published less "sloppy" journals by less "Sloppy" scientists.
But non the less, I think this work at all levels have a use and
a place.

On the other hand, I agree that we should prosecute people who
intentionally attempt to deceive their readers.

Keith Karn
keith@cvs.rochester.edu

************************************************** ***************

From: IN%"clark@SALUS.MED.UVM.EDU" "Toni Clark x63851"

I believe the term "sloppy data" usually means sloppily collected
data. Thus, unless one were privy to the data collection
procedures, one wouldn't really know whether the data were
"sloppy."

************************************************** ***************

I submit the following for your consideration:

Sloppy data is data collected in a fashion not consistent with
commonly accepted standards.

Because sloppy data is a product of methodology, sloppy data is
unlikely to be identified unless the data collection process is
observed. The definition revolves around the premise that
commonly accepted standards of data collection exist.

I am unaware of the existence of any published commonly accepted
standards of data collection in the field of biomechanics. The
text by Dainty and Norman represents a start but I do not believe
a compendium of accepted standards has been endorsed or
published. It is unfair to hold persons to unpublished
standards. The existence of such standards would appear to be
increasingly important as individuals not necessarily trained as
biomechanists begin to pursue research which is biomechanical in
nature. I propose that the profession needs to identify and
publish standards of biomechanical data collection. How do you
feel about giving ISB a mandate to produce such a document? How
would the executive council of ISB respond to such a mandate?

I do not believe that peer review can succeed as a mechanism for
identifying science fraud and misconduct. Similarly, I am
willing to bet that the rare cases of questionable data found in
the peer review process are the result of ignorance on the part
of the investigator rather than an intent to "cut corners." If I
am correct, what does the peer review process really do? How do
you feel about Kuhn's contention that the peer review process is
nothing more than a mechanism for reinforcing the existing
paradigm (normal science)?

Once again, I encourage you to post your responses to the list at
large. I miss the immediacy of the discussion format and I am
anxious to benefit from exchanges by those of you wiser and more
philosophically evolved than myself.

Thank you for your time, patience, and thoughts.

William (Bill) Siler
silerwl@sluvca.slu.edu