Dear all,

I'm sure that many of you noticed the recommandation of ASA about the use of p-values (http://www.amstat.org/newsroom/pressreleases/P-ValueStatement.pdf).

It is now clear that p-value can not be used to determine if an hypothesis is true (or if results are important)...

I'm thinking about the power of tests (1-beta) which been forgot for a while in papers...

Does it mean that, now, we have/should tell both Alpha and Beta values in papers when hypotheses tests are performed ?

(However, another problem is that Beta evaluation is not well documented, not very well teach... therefore not so easy to calculate. And moreover, should not be calculated but should be a factor to determine sample size etc. ...)

Regards

More Cyril Iaconelli's questions See All
Similar questions and discussions