Statistics of Test For ESP
                                    June 1999
                                 Paul  Cooijmans
                                  De Wolwever 39
                                5737 AD Lieshout
                                    Netherlands

   The Test For Extrasensory Perception was published in various IQ journals
and is on  http://www.eskimo.com/~miyaguch/esp.html.  The scores achieved between early
1997 and June 1999:

   Raw       Frequency
   0         24
   .5        4

   A new estimate of each score's probability:

   Raw       Probability (1/x)
   0         1.15
   .5        8
   1         30
   1.5       1 000
   2         30 000
   2.5       3 000 000
   3         1 000 000 000
   3.5       100 000 000 000
   4         >100 000 000 000
   4.5       >>100 000 000 000
   5         infinite

   As no correct answer has been given, the test will continue for another two
years.  It is below.

Test For ESP
   Does Extrasensory Perception exist?  Let's find out once and for all.  In a
chest of drawers, standing in a corner of my office at the Paul Cooijmans In-
stitute For Advanced Study, I keep five postcards.

   1  What number is written on the first?
   2  The name of which historical figure is on the second?
   3  What English word is on the third?
   4  What geometrical figure is drawn on the fourth?  Answer by drawing it.
   5  What object or organism is drawn on the fifth?  Answer by drawing it.

   On face value, the chance of answering one of these questions right (when
attempting all five) is about 1 in 30.  The chance of two right is 1 in 30 000,
and of three right 1 in a billion.
   If all humans would answer these questions at random, the highest score
would probably be 2.  If anyone scores 3, I start doubting about the existence
of ESP, and if anyone scores 4 I consider it proven beyond reasonable doubt.  If
anyone scores 5, ESP exists.  
   If you submit answers and enclose US $3, you will receive a score report
within two weeks.  The deadline for submitting answers will initially be Jan. 1,
2001.  If no correct answers received by then, the test will continue for
another two years, and so on.  After withdrawal of the test the
correct answers will be published and available on request.  They have been le-
gally registered early 1997.  A statistical report on the project will be
published with the answers.
   The drawings should be clearly recognizable but do not have to be accurate
to the millimeter.  Repeated attempts and discussions with others are allowed.

  Return to the Uncommonly Difficult I.Q. Tests page.