From: Damien Broderick (damien@ariel.ucs.unimelb.edu.au)
Date: Thu May 07 1998 - 05:15:13 MDT
At 08:50 PM 5/5/98 -0700, Hal wrote:
>I would feel better if a single [psi] study, done with iron-clad protocols,
>replicable, could be done at a large enough scale to demonstrate the
>phenomenon at a strongly significant level.
The Princeton Engineeering Anomalies Research corpus of data does exactly
this, by brute force during 20 years of dogged work with stable protocols
and many Ss.
On the other hand, phenomena with small effect size frequent *don't* yield
significant results in modest experiments, and cannot be expected to - a
point that seems to elude non-statisticians. Prof. Jessica Utts is good on
this.
>Seeing around the world isn't any harder than seeing into the room
>next door. Foretelling events a few seconds away is no easier than
>foretelling things which will happen hours or days later.
It's not altogether clear that this is the case. Take a look at the
reports in Radin's book on `presentiment', where they found electrodermal
orienting responses immediately in advance of Ss. seeing shocking or
distressing or erotically arousing images, but not with control neutral
imagery. This experimental set-up is said to have been replicated
succesfully, and is easy enough to do.
And of course the same generic objection could be made against non-local
quantum effects. The difference is that QT has powerful theory predicting
the observed effects; parapsychology or anomalies research has only the
offensive data and a batch of rough-and-ready explanatory schemata. But
the data need to be explained, and at this stage of the game fraud and
sloppy thinking are not good enough explanations.
Damien Broderick
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 14:49:04 MST