[ExI] mersenne primes again

[spike]

> ...On Behalf Of Mike Dougherty
> Subject: Re: [ExI] mersenne primes again
> 
> On Thu, Apr 1, 2010 at 1:44 AM, Damien Broderick 
> >...Wow, *symmetry*, eh? That certainly saves us 
> from all those bad experiments and a slap in the face.
> 
> ...Seems to me that 
> supposing something is correct based on the 'demands' of 
> symmetry is a form of religious belief in the inherent value 
> of symmetry as a ruler...

Ja that is pointed up by Chris Calwell's page, the Wagstaff conjecture: the
number of Mersenne primes less than or equal to x is about (e^gamma/log 2) *
log log x.  (Here gamma is Euler's constant).  It is such a beautiful
theory, gorgeous it is.  But I think the relationship is coincidental, for
reasons I will cover at some later date, perhaps with a subgroup of ExI math
fans.  

For now I want to stay on the aspect that is relevant to extropy. 

> I think what spike was suggesting is that using statistics 
> alone, we may experience real psi effects but still not "see" 
> it because the 95% confidence criteria is not met...

Ja.  Take the textbook example, or one that is in my textbook.  Twenty
babies are born to smoking mothers, and 14 of those are below average
weight.  Is this data statistically significant?  I get 93.4% confidence,
which doesn't quite make the cut, but the cut is arbitrary.  There is plenty
of reason to be suspicious that smoking somehow causes lowered birth rates.

I would suggest we change the way statistics textbooks present that topic.

> There are three kinds of lies: lies, damned lies, and statistics.

Even if the statistics are done correctly, they are often, if not usually,
misleading.  That discipline is so filled with shortcuts, the results are
more often useless than otherwise.

spike