[ExI] Precog PCs

[Damien Broderick]

On 10/24/2010 1:43 PM, Stefano Vaj wrote:
>> The question you ask is very tricky.  How to distinguish peformance due to
>> >  the PC from performance due to experimenter?  Unless you can get the PC (it
>> >  would have to be an AI) to devise, execute and publish the research by
>> >  itself.
>
> The performance of the experimenter in developing a guessing program?

No, the possible influence of the experimenter directly and by 
(da-da-DA-da da-da-DA-da) the very same mysterious psi process you're 
trying to investigate.

Experimenter effects are quite frequent in psi research; famously, a 
joint protocol devised and followed exactly by both skeptic Richard 
Wiseman and parapsychologist Marilyn Schlitz got null results for 
Wiseman's cohort and positive effects for Schlitz.** There was no 
indication of cheating by Schlitz (Wiseman would have been quick to crow 
had he found any).

The implication is that any psi experiment is always subject jointly to 
at least the influence of the agent or subject (who should be double 
blind) and the experimenter (who should be blind to the target).

With a program devised to "guess" (whatever that means for an 
algorithm--do you throw in some sampling of quantum noise?), any 
significant effect could always be attributed to the desires of the 
experimenter.

**http://www.richardwiseman.com/resources/staring1.pdf

Damien Broderick