[extropy-chat] ESP controls

Ben Goertzel ben at goertzel.org
Sun Feb 11 19:07:31 UTC 2007


Damien is correct, there have been many psi experiments with adequate 
controls.

I have not studied the topic carefully, but based on the little reading 
I've done, my  main worry with psi experiments is whether there's a 
powerful "dataset selection" effect. 

I.e., if someone does a psi experiment and doesn't get positive results, 
they won't tell anyone and won't publish the results -- they'll just 
think their experiment wasn't good enough for some reason.  So what 
we're seeing in publications is just a subset of the POSITIVE results 
obtained by various researchers, and we don't know how large a subset 
this is over the overall set of results obtained.

I assume that statisticians studying psi experiments have attempted to 
account for this phenomenon, but I don't know exactly how they have done 
so....  It does seem difficult to address.

I really doubt that Jahn and all the other psi researchers showing 
positive psi results have fabricated their data.  There are just too 
many researchers with otherwise trustworthy appearance, showing positive 
results.  It is possible of course that every positive psi result is an 
intentional fabrication, but this seems not that likely to me.  If psi 
does not exist, the more likely explanation for the various positive 
results obtained, I would say, is the dataset-selection effect I 
mentioned above, or some other peculiarity of statistics and scientific 
protocol.

The level of BS in the psi literature is far higher than in the CF 
literature, making it much harder to penetrate to even a superficial 
level without expending a large amount of effort.

-- Ben


Damien Broderick wrote:
> At 01:37 AM 2/11/2007 -0500,  John K Clark wrote:
>
>   
>>  ESP "research" is even worse; the idea that a
>> control experiment is a luxury and not a necessity.
>>     
>
> Could you elaborate on that, John? Which psi experiments do you mean 
> that have no controls?
>
> Presumably not the bulk of the PEAR REG trials, where the 
> long-established protocol is specifically designed with no optional 
> stopping, and with ternary target conditions (higher than chance, 
> lower than chance, baseline or no effort), and those mutually 
> controlling conditions augmented by calibration runs of the REG while 
> nobody is making any psi attempts.
>
> I could list other controls for different protocols, but I'd prefer 
> to learn precisely which experiments you're talking about. Or will 
> you default to your fallback position of "Jahn and the others are 
> probably just inventing their data"? Or Dennett's dynamite rebuttal, 
> which I cited the other day: "You don't think I waste my time 
> actually reading this stuff do you?"
>
> Damien Broderick
>
>
>
>
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
>
>   




More information about the extropy-chat mailing list