PDA

View Full Version : Re: What criteria should we use for interpreting epidemiology?



idavis
02-20-2008, 01:46 AM
Frank,

I also vividly remember those words of wisdom from Dr. Alexander and
couldn't agree with you more.

all the best,

Irene

Irene Davis, PhD, PT, FACSM
Professor, Dept. of Physical Therapy
305 McKinly Lab
University of Delaware
Newark, DE 19716

(302)831-4263 (ph)
(302)831-4234 (fax)

mcclay@udel.edu
http://www.udel.edu/PT/davis/index.htm

On Feb 20, 2008, at 9:47 AM, Buczek, Frank (CDC/NIOSH/HELD) wrote:

> Chris and others,
>
> I can think of nothing that would more quickly dilute the truth, than
> research by consensus. Even small scientific teams, composed of a
> senior scientist and a handful of post docs, working on a limited
> budget, can find the truth if they use good experimental
> techniques. If
> their sample size is considered by some to be small, then others can
> replicate the study with a larger sample to confirm or refute the
> findings. This is, after all, the way science corrects itself over
> time.
>
> I can recall an exchange between two men I very much respect, at the
> 2003 ASB conference in Toledo, Ohio, USA:
>
> Dr. McNeil Alexander had just finished giving a wonderful Borelli
> Award
> Lecture. After a few questions from the audience, Dr. Peter Cavanagh
> asked Dr. Alexander if he had any advice for the young people in the
> audience who were just beginning their careers. After thinking for a
> moment, Dr. Alexander replied, "Don't follow the herd." He went on to
> say that important discoveries in science seem to happen more often
> when
> people allow themselves to think independently.
>
> It is difficult for me to imagine this happening as easily within the
> consortium described by Dr. Ioannidis, as quoted in an earlier
> BIOMCH-L
> submission.
>
> [These are my opinions, and not necessarily those of NIOSH.]
>
> Best regards,
> Frank
>
> Frank L Buczek Jr, PhD
> Branch Chief, HELD/ECTB
> National Institute for Occupational
> Safety and Health (NIOSH)
> 1095 Willowdale Road MS 2027
> Morgantown, WV 26505
> 304-285-5966 voice, 304-285-6265 fax
> fbuczek@cdc.gov
>
>
> -----Original Message-----
> From: * Biomechanics and Movement Science listserver
> [mailto:BIOMCH-L@NIC.SURFNET.NL] On Behalf Of Chris Kirtley
> Sent: Wednesday, February 20, 2008 7:11 AM
> To: BIOMCH-L@NIC.SURFNET.NL
> Subject: Re: [BIOMCH-L] What criteria should we use for interpreting
> epidemiology?
>
> Very interesting, Adrian!
>
> For those interested, the original article is at:
> http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1182327
>
> I have often thought that research papers have following weaknesses:
>
> 1. In the introduction or background (which should build the case for
> the
> question being answered), reference is often made to statements by
> others in
> a similar section of their paper, or speculative comments in the
> discussion.
> In my view, reference should only be made to the conclusions previous
> papers.
>
> 2. Particularly in biomechanics (and even more sports biomechanics) I
> suspect, scientists often rely on their intuitive feelings about
> what is
> true, refusing to believe results which have been confirmed over and
> over
> again.
>
> 3. The numbers of subjects and tests are usually so small as to
> make the
> application of statistics absurd.
>
> Great food for discussion, Adrian & David!
>
> Chris
>
> ---------------------------------------------------------------
> Information about BIOMCH-L: http://www.Biomch-L.org
> Archives: http://listserv.surfnet.nl/archives/Biomch-L.html
> ---------------------------------------------------------------