I’m very gratified by the huge response to the PATMA. This test is dear to my heart because it’s the first program I ever executed online. I had to learn both Python and HTML (“Oh Easy Peasy; like doing long division in the 4th grade” as Oprah would say).
Speaking of what Oprah would say, there’s an entire generation that still programs in Quick Basic, and they just need to die:
Now here are the results of the poll of self-reported PATMA scores.
If we remove the ten fake scores of 10 added by commenter “Mug of Pee”, then perhaps 127 real votes give the following percentages.
10 = 14%
9 = 24%
8 = 24%
7 = 23%
6 = 11%
5 = 2%
3 = 1%
From here we get the following percentiles with their respective normalized Z scores:
10 = 86 (+1.07)
9 = 61 (+0.27)
8 = 37 (-0.33)
7 = 14 (-1.07)
6 = 3 (-1.87)
5 = 1 (-2.33)
4 = 1 (-2.33)
Based on previous research on a Gestalt test normed on the general white population, my readers have the following IQ distribution:
As you can see, an IQ of 137 puts you in the top 26% of my readers (+0.66 SD) while an IQ of 106 puts you at the 10th percentile (-1.27 SD). From here we can guestimate that on a scale where whites have a mean IQ of 100 with an SD of 15, my readers have a mean IQ of 126 with an SD of 16.
Thus, to convert PATMA scores to IQ, we multiply the normalized Z scores by 16 and add 126.
PATMA 10 = IQ 143 (very brilliant)
PATMA 9 = IQ 130 (brilliant)
PATMA 8 = IQ 121 (very bright)
PATMA 7 = IQ 109 (bright)
PATMA 6 = IQ 96 (U.S. average)
PATMA 5 = IQ 89 (dull)
PATMA 4 = IQ 89 (dull)
Of course these numbers are all very tentative and more research is needed.
[update aug 20, 2020: an earlier version of this article overestimated the IQ variability of my readers by giving too much weight to low outlier scores on the Gestalt test. Helpful feedback from commenter Rahul caused me to revise the numbers]