I recently did a poll on the SAT scores of readers of pumpkinperson.com and after converting SAT scores to IQ equivalents, here are the IQ equivalents, in order from highest to lowest, of the poll respondents, as of Sept 24, mid afternoon, who were 1) American, 2) old enough to take the SAT, and 3) did not take the ACT instead (note: all IQs are expressed on a scale where white Americans average 100 with a standard deviation of 15 and thus the average IQ for all Americans is below 100):

165 (obtained on old SAT)
149 (obtained on old SAT)
146 (obtained on new SAT)
144 (obtained on old SAT)
144 (obtained on old SAT)
141 (obtained on new SAT)
141 (obtained on new SAT)
141 (obtained on new SAT)
141 (obtained on new SAT)
141 (obtained on new SAT)
138 (obtained on new SAT)
136 (obtained on old SAT)
136 (obtained on old SAT)
136 (obtained on old SAT)
136 (obtained on old SAT)
134 (obtained on new SAT)
134 (obtained on new SAT)
134 (obtained on new SAT)
134 (obtained on new SAT)
134 (obtained on new SAT)
131 (obtained on new SAT)
131 (obtained on new SAT)
128 (obtained on new SAT)
126 (obtained on old SAT)
124 (obtained on new SAT)
121 (obtained on new SAT)
96 (obtained on new SAT)
93 (didn’t write SAT/ACT, IQ 93 assumed)
93 (didn’t write SAT/ACT, IQ 93 assumed)
93 (didn’t write SAT/ACT, IQ 93 assumed)
87 (obtained on old SAT)
86 (obtained on new SAT)
64 (obtained on new SAT)

So on a scale where the average white American averages 100 with a standard deviation of 15, and where Americans as a whole now average 96-97 with an SD of 15-16, American pumpkinperson readers average 127 with a standard deviation of 23!

If I eliminate the three people who didn’t take the SAT or ACT and thus could only be assigned an extremely approximate IQ, the average IQ rises to 130, and the standard deviation falls to 21. However it’s known that largely the dullest Americans don’t write such tests, so excluding these people will bias the data upward. On the other hand, assigning IQs to people who were not tested is also highly questionable.

Let’s split the difference and assume my (American) readers have an average IQ of 129 with a standard deviation of 22. An average IQ of 129 does not surprise me, but a standard deviation of 22 is a bit surprising, because usually the assumption is that the higher the average IQ of a group, the lower the variability, but my readers are extremely variable (with one reporting an SAT score in the mentally retarded range).

Of course some of this variability might just be error built into any online poll, however I suspect that the main reason why my readers are so variable is that HBD blogs attract both brilliant people who are fascinated by IQ and genetics, and low IQ people who are obsessed with being racist, as well as people who are interested in IQ because they scored abnormally high or abnormally low.