LLN, le 7/03/05
Hi All!
>... I would only say that behavioral papers have been
>reporting effect sizes for tests that did not pass significance...
Matthew, that some authors do use not-significant results to support
their claim does not make it right or even acceptable.
>... without causing chaos.
I'm afraid I disagree on this point. Reviewers shouldn't let the
above mentioned authors do, and they all - these reviewers and these
authors - should be condemned to be burnt to the stake (hopefully,
this will free some space in our labs!), or at least to hell, where
they will eternally pay for their sin.
Seriously, there's probably nothing really bad to speculate a bit on
trends as long as the speculative aspect is made clear for the
readers. But there are already to many papers/people out there
spreading unsupported claims and it's not because we don't see the
chaos that it's not there...
>we should set ourselves the rule never to comment on not-significant
>results...
Maybe too strong. Going back again to behavioral data, none of us - I
hope! - would accept discussing a result at p>.2, but we would
certainly be indulgent with a "reasonable claim" at p<.07 or even
p<.1. So, perhaps the way out is more in searching what would be a
"reasonable claim" with near-significant results than finding the
universally accepted statistic - which, as can be seen from this
discussion, we are far from having found. Therefore,
- if we all stop deriving unsupported claims (i.e., across regions)
from thresholded maps,
- if some of us use unthresholded maps along with clear caveats,
- if the reviewers/editors AND readers do their job,
we'll (nearly) all be happy, won't we? On top of that, if people
using unthresholded maps do a nice job, they will progressively
convince other people, and Matthew will be happy too.
Nice program, isn't it? Yours,
Mauro.
--
_____________________________________
Help fighting hunger: http://www.hungersite.com
Just click your mouse and sponsors of The Hunger Site donate a
serving of food to a person in need - at no cost to you.
______________________________________
Mauro PESENTI
Research Associate, National Fund for Scientific Research (Belgium)
Unite de Neurosciences Cognitives
Departement de Psychologie
Universite Catholique de Louvain
Place Cardinal Mercier, 10
B-1348 Louvain-la-Neuve
tel.: +32 (0)10 47 88 22
fax: +32 (0)10 47 37 74
E-mail: [log in to unmask]
http://www.nesc.ucl.ac.be
http://www.nesc.ucl.ac.be/mp/pesentiHomepage.htm
|