Transhuman community IQs

A while back I asked people on ExI-chat and SL4 to do the Tickle classic IQ test for a bit of a fun since there were alot of discussions about IQ, personalities and child prodigies going on. There are also a large number of seemingly very intelligent people active and lurking on the lists, so I thought I’d see how they compare to the general populace.

Please note that this was all just for curiousity and fun. Thus it is not meant to be serious and I’ve included some general comments, some of which come from respondants.

The original email was:

Would anybody be interested in finding out just what the IQ distribution of ExI chat is?

If so, I suggest we use something like the tickle classic IQ test: http://web.tickle.com/tests/uiq/?test=uiqogt

To prevent this from turning into a pissing contest and decrease the likelihood of people lieing, I volunteer to collect peoples IQ scores via email and make some nice graphs of the distribution. Once done I’ll destroy the emails with your IQs and promptly forget the actual values.

Please only report the tickle IQ scores otherwise test differences may bias the distribution – and please be honest!

I got 18 replies from the extropy-chat and 11 from SL4. For extropy-chat the mean was 138.1, s.d. 3.9. For SL4 the mean was 137.9, s.d. 3.7. These values were not significantly different as judged by a two sample t-test. Below is a graph summarising these results.

iq

Notes:

  • After some discussion and rumours about what the limit of top IQ score was, I emailed Tickle and they informed me that 144 was the maximum possible score.
  • Knowing this, there was one respondant that lied about their score by claiming one above this limit.
  • There was some disagreement about just what the correct answers were. When you have to find the odd one from a group such as [ apple, peach, pear, grape ] there are multiple interpretations. All but grape grow on trees, apple is the only one beginning with vowel etc.
  • Because of the generally high scores and several people getting the maximum possible, Tickle probably wasn’t a good choice to get an accurate representation. Let alone the fact it is a product designed to sell intelligence profiles so could potentially be biased. One respondant suggested some high level tests which probably give more insightful results.
  • Lastly, I want to apoligise to list members as a large number of replies were sent on list rather than directly to me. Ah well, people never follow instructions even when they do top out IQ tests ;P



3 comments ↓

#1   Martin Striz on 11.29.05 at 1:15 am

What surprises me is the tight standard deviation for 18 respondents. About 70% of the respondents were between 134 and 142? Sounds suspicious. This test is invalid because a lot of people were probably just topping out the test but with a few mistakes.

Also there may be a selection bias because people who got “low” scores (below 130) either didn’t report them or lied.

#2   Joel on 11.29.05 at 1:43 am

Yup, and to tell the truth the distributions were not normal – ExI was in fact pretty much uniform for the frequencies of values 130-144.

I do realise that this whole “study” is flawed. As I said, just a bit of fun. Anyone who has the time to do a more enlightened survey is most welcome.

#3   Arthur on 11.29.05 at 8:10 am

Mentifex IQ scorings (so that you may judge whether Mentifex Seed AI is worth evaluating) over time rank in the 145-149 range. Graduate Record Exam (GRE) Verbal 740; Quant. 710; Law School Admission Test (LSAT) 637; U.S. Army General Technical (GT) Score 149.

Leave a Comment