[extropy-chat] Re: Damien grants psi evidence

[Damien Broderick]

At 07:06 AM 12/18/2004 +1100, Brett wrote:

>Yet the size of the effects reported (small), and the nature of the
>way in which they were reported, (with apparently good scientific
>method and a seemingly very capable understanding of statistics)
>didn't offer me reason enough to need to push further immediately.

Utts observes that the small but fairly robust effect sizes are slightly 
larger than those relating daily doses of aspirin to mitigation of heart 
attack risk; this latter finding is regarded as well-established by the 
medical profession, and a double-blind test of the aspirin effect was shut 
down early because it would have been unethical to deny patients the 
benefit of the drug.

Of course it's also obvious that psi effects *must* usually be small, since 
we don't see paranormal effects most of time. Experiments are designed to 
concentrate or elicit these effects, making them visible to an extent not 
found in ordinary daily life.

>I would still
>suspect, as Damien seems to, that most that are in that area are
>cranks and fools.

That's not quite what I said. The many, many `psychic' pests in the 
marketplace who prey on the gullible and vulnerable are probably liars and 
frauds (although if psi is indeed a human capacity, even they might 
sometimes gain some extra prowess via that route). Many of the lab 
parapsychologists do hold beliefs I regard as suspect or ludicrous; they 
include a greater than average proportion (among scientists) of believers 
in religion or spiritualist doctrines, and many are metaphysical dualists. 
That makes me more cautious in judging their findings, since they might be 
more prone to find `evidence' supporting their beliefs. But contemporary 
psi work is very well-designed. Even if `screening' runs are conducted less 
rigorously (self-testing and reporting, for example, to locate 
`high-scorers', some of whom will have scored well sheerly by chance, 
others by trickery), the subsequent tests in the lab are done with extreme 
care and many safeguards against fraud or, say, accidental leakage of 
target information.

As for statistical competence, it's worth noting that Utts is a professor 
of statistics at UC, Davis, with a PhD in Statistics. This is not an 
argument from authority, but it does suggest that she's not reaching these 
conclusions out of ignorance.

Damien Broderick