[SciPy-User] [OT] Bayesian vs. frequentist
Fri Feb 17 21:21:06 CST 2012
On Tue, Feb 14, 2012 at 3:39 PM, Sturla Molden <email@example.com> wrote:
> On 14.02.2012 21:24, firstname.lastname@example.org wrote:
>> Do you expect an argument? sounds a bit like http://andrewgelman.com/
> No I don't. I am just tired of explaining that "significant p-value"
> does not imply "very important effect". Sorry for spamming the list with
> my rant.
Since I just came across a paper on the topic
looking for something else.
I have seen now several papers (mainly the abstracts) that test
"equivalence" instead of "equality", so we can specify in advance what
difference is "important" and what would be "equivalent".
> SciPy-User mailing list
More information about the SciPy-User