🚀 go-pugleaf

In article <1d5lldINN1up@frigate.doc.ic.ac.uk>, sjv@doc.ic.ac.uk writes:
> In article <1992Oct30.235244.463@black.ox.ac.uk> microsoc@black.ox.ac.uk
> (Microsoc) writes:
>>As we seem to be on about Popper can anyone answer me this;
>>How can you falsify (even in principle) a probabilistic theory, eg QM..?
>>
>>Marc Read
>>
> 
> The whole of statistical testing is about falsifying probabilistic theories:
> 
>   the null hypothesis is a probabilistic theory
>   you reject it by falsifying it
> 

True.

> Of course you don't falsify it absolutely, but from the evidence you do
> conclude that it's most unlikely (P < epsilon) for the null hypothesis to be
> correct.
> 

Not really. The 'P' here is the probability of the data given
the null hypothesis {speaking a little loosely} rather than
the probability of the null hypothesis given the data.

I have just spent some time on bit.listserv.edstat-l trying
to convince people that it was a waste of time trying to ram
home this point in intro stats. I didn't expect to be having
to make the point in this forum!

> This is one way how, in practice, people do go about falsifying probabilistic
> theories.
> 
> Steve Vickers

Is this the right thread in which to raise the question of Rabin's
probabalistic primality test. It seems that there are some situations
in which it seems best to rely on probability arguments even when
formal proofs are available in principle.
-- 
Murray A. Jorgensen [ maj@waikato.ac.nz ]    University of Waikato
Department of Mathematics and Statistics     Hamilton, New Zealand      
__________________________________________________________________
'Tis the song of the Jubjub! the proof is complete,
if only I've stated it thrice.'