I’m glad to announce that the first in the series of previously announced Psi Experiments to test the abilities of the human mind has been launched. This is a most basic parapsychology experiment with more interesting ones to follow.
Tell all your friends and encourage them to participate.
Go to Psi Experiment #1 and do your best.
-Hatter
Were you to generate a different random target for each guess, then the trials would be independent, and you could aggregate them with standard statistical distributions. With the current experiment you cannot do that.
If you want to show real evidence for a paranormal effect, you'll need better experimental design. You should also do the tests double-blind. You should state exactly what data you record and what are the statistics of record (and of course you should report all that you record, except where you have promised privacy). You should put your code under version control and make it available for review.
I've only seen just one on-line PSI test that does it really well: The RetroPsychoKinesis Project
http://www.fourmilab.ch/rpkp/
They document their random number generators. They make the code available, so the conditions are known and repeatable. They fully report the results.
Box 4.