PDA

View Full Version : Re: BioNet controversial topic #5: replication of numericalstudies



Paul Ostic
04-01-2002, 06:38 AM
Hello All

Please excuse the blunt way I express my opinion here. The topic
is controversial, and my true feelings are much milder than my comments
here. In general, the integrity and intelligence of my colleagues and
peers are taken for granted. But to illustrate a point, I take the
opposite approach here.

Numerical codes, simulations and models must be shared. If they are not
shared, then he who refuses to share them has something to hide. Sharing
them would promote more rapid growth in understanding and knowledge,
benefiting the entire field. Indeed, other researchers are certain to find
applications and refinements to the model that would bring additional
credit to the one who first developed the model, since we would all
acknowledge the source that we were building on.

There are assumptions implicit in all models. Some are obvious, some are
not. Sometimes when I approach a problem, I am not even aware of the
underlying assumptions implicit in my approach. Only by having others test
my model using their own data, will I be convinced that my inherent
assumptions are not unduly limiting to the model. Or at the very least, I
want to make sure that these assumptions and limitations are
documented. If I don't want to share my model, then it is because I am too
vain to admit that my understanding may be imperfect.

I am a skeptic. I know that data can be manipulated to show things in many
different lights. If I am not permitted to read the fine print, try the
tool for myself, then it is very difficult for me to judge something
favourably.

Paul Ostic
MSc student
Queen's University
Kingston Ontario
Canada

---------------------------------------------------------------
To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l
---------------------------------------------------------------