Re: a ESP test (statistical problem)
- From: Timothy Little <tim-usenet@xxxxxxxxxxxxxxxxxx>
- Date: Tue, 2 Aug 2005 21:55:52 +0000 (UTC)
riparian wrote:
> A ESP test (extra sensory perception) is carried like this: From a
> deck of cards were chosen 5 cards. These cards were seen by both
> the conductor of the experiment and a person. The cards were laid
> on the table with the back side upwards in such a way that only the
> conductor of the experiment knew the sequence of the cards.
If I interpret this correctly, there are 5! = 120 possible sequences
for each experiment.
> The conductor chosed a card and the person should then say which
> card it was. In a sequence of 100 experiments the person answered
> correctly 37 times.
Were the cards perfectly shuffled and redrawn after each answer? If
so, then the null hypothesis gives a binomial distribution with mean
20 and standard deviation sqrt(100*0.2*0.8) = 4.
> Is there reason to believe that this person has special
> supernatural powers?
There is strong evidence that the null hypothesis does not hold. That
is, it is unlikely that the result was due to chance alone.
However, you've described absolutely nothing about any controls
designed to rule out non-supernatural means of obtaining information.
As the description stands, I can think of at least a dozen perfectly
ordinary channels of information you may not have closed.
- Tim
.
- References:
- a ESP test (statistical problem)
- From: riparian
- a ESP test (statistical problem)
- Prev by Date: Re: What does Gödel's Incompleteness mean for the Working Mathematician?
- Next by Date: Re: modeling courtship as a sequential game
- Previous by thread: Re: a ESP test (statistical problem)
- Next by thread: Re: a ESP test (statistical problem)
- Index(es):
Relevant Pages
|
