On 24.10.2012 20:31 meekerdb said the following:
On 10/24/2012 5:31 AM, Stephen P. King wrote:


Woo-woo.  Small effect sizes which are *statistically* significant
are indicative of bias errors.  I'd wager a proper Bayesian analysis
of the original data will show they *support* the null hypothesis
(c.f. "Testing Precise Hypotheses" Berger & Delampady, Stat Sci 1987
v2 no. 3 317-352 and "Odds Are It's Wrong" Tom Siegfried, Science
News 27 Mar 2010).  Meta-analyses are notoriously unreliable and
should only be considered suggestive at a best.

It is a general situations with a statistical treatment. When people like results based on mathematical statistics, as for example correlations in a neurosience, they say that this is a good science. And when people do not like statistical results, they can always say woo-woo.


You received this message because you are subscribed to the Google Groups 
"Everything List" group.
To post to this group, send email to everything-list@googlegroups.com.
To unsubscribe from this group, send email to 
For more options, visit this group at 

Reply via email to