The editor of the "Journal of the American Medical Association" (JAMA)
has written an editorial for researchers making submissions to JAMA and
for medical research in general which can be read (free) here:
http://jama.jamanetwork.com/article.aspx?articleID=1699471&utm_source=Silverchair%20Information%20Systems&utm_medium=email&utm_campaign=JAMA%3AOnlineFirst06%2F20%2F2013
One of the key aspects is the registration of clinical trials (experimental
studies on treatments) of key aspects of the research (e.g., how it is
to be conducted and analyzes, etc.) BEFORE the conduct of the research.
Discrepancies between what was registered and what ultimately winds up
in manuscripts for publication have to be explain (presumably, if a
researcher
had to change something, they would notify the registry of what was changed
and why -- this is to avoid ad hoc "tweaking" that might be done). Jack
Cohen
used to say that psychology researcher should write a letter to themselves
before they undertook research, identified the hypothesis/hypotheses they
were
testing, what sort of analysis they would do, would they use one-tailed or
two-tailed tests, and related things like the effect size and amount of
statistical
power. Sadly, I think few people took Jack's advice but perhaps a registry
of research studies in psychology might be a good idea though this would
probably be very hard to put into practice.
-Mike Palij
New York University
[email protected]
---
You are currently subscribed to tips as: [email protected].
To unsubscribe click here:
http://fsulist.frostburg.edu/u?id=13090.68da6e6e5325aa33287ff385b70df5d5&n=T&l=tips&o=26173
or send a blank email to
leave-26173-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu
