[R] acceptable p-level for scientific studies

Michal Bojanowski bojaniss at poczta.onet.pl
Thu Dec 19 15:36:02 CET 2002


Wednesday, December 18, 2002, 3:12:51 PM, you wrote:

FEHJ> Want to open up the floodgates?  Some personal opinions:

FEHJ> - Alpha=0.05 is arbitrary, silly, and boring
FEHJ> - Reporting P and letting the reader decide is a bit better
FEHJ> - Bayesian posterior probabilities are still better although
FEHJ>   more thinking is required
FEHJ> - Confidence limits can be good compromise solutions (some journals are
FEHJ>   almost disallowing P-values in favor of CLs)
FEHJ> - P-values are dangerous, especially large, small, and in-between ones.
FEHJ>   See http://hesweb1.med.virginia.edu/biostat/teaching/bayes.short.course.pdf for a full sermon.

Hello,

I'm very interested in differences between bayesian and frequentist
decision making. The only source of knowledge about Bayesian approach
I have is:

James O. Berger "Statistical Decision Theory and Bayesian Analysis"
Springer-Verlag 1985

Could you point me to some books/articles that will explain
differences between those approaches. The explanation in the book
above is not satisfying (for me).

I'm especially very interested in the `philosophical' grounds, that
constitute Bayesian approach.

Thank you in advance.


Michal


~,~`~,~`~,~`~,~`~,~`~,~
Michał Bojanowski
bojaniss at poczta.onet.pl



***************r-e-k-l-a-m-a**************

Masz dość płacenia prowizji bankowi ?
mBank - załóż konto
http://epieniadze.onet.pl/mbank




More information about the R-help mailing list