IQ
measurement is not exact science when trying to measure beyond 145 (SD=15).
Tests, such as used by Mensa, do not discriminate at
this level.
As a mathematician,
I've noticed that we can still make good guesses based on few data available.
Read more...
I'm not going to completely reveal my norming methods,
but here is a general explanation.
I need at
least 20-30 testees to make the first norm report and 120 testees to make the
final norm report.
Additionally, I need some statistical information. Here
is an example.
Scores: 22, 15, 15, 14.5, 13, 12, 11.5, 11, 10, 10, 9.5, 9, 7, 6.5, 6, 5, 5, 5, 2.5, 1
Frequencies:
90%, 90%, 85%, 72.5%, 67.5%, 65%, 52.5%, 45%, 40%, 35%, 27.5%, 25%, 25%, 25%,
21.25%,
20%, 20%, 20%, 17.5%, 17.5%, 15%, 12.5%, 12.5%, 10%, 10%, 10%, 5%, 5%, 0%, 0%
This is statistical
information for the first 20 testees on Numerus IQ test presented here.
By score, I mean only the better of two possible attempts
of the single testee.
By frequency, I mean the solvability of a single item.
For instance, 21.25% is the frequency of an item solved
by 4 testees, unsolved by 15 testees,
and one testee got 0.5 point in one attempt and 0 point
in the second.
Totally, 4.25 points out of 20 possible.
Grand
IQ Society is open to any author and any kind of IQ test.
I will accept any IQ test of sufficient difficulty,
for which statistical data is provided as exampled here.
Detailed personal information about the top five testees
will be helpful.