The TAVIS has now been taken at least 358 times (excluding people who clicked on the link but didn’t submit, or who submitted their answers without attempting any questions, or who answered only one question (a hard one) because they were likely trying to confirm if a guess was right).

Below are the 358 scores that are likely legit (though it’s possible the highest scores are from people involved in the above mentioned monkey business).

2,3,4,4,5,5,5,5,6,6,6,6,6,6,6,6,6,6,6,6,6,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,7,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,10,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,11,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,12,13,13,13,13,13,13,13,13,13,13,13,13,13,13,13,14,14,14,14,14,14,14,14,14,14,14,14,15,15,15,15,15,15,15,16,16,16,16,16,17,17,17,17,18,18,18,18,19

Since the first norming, the mean and SD have both increased slightly (they’re now 9.98 and 2.62 respectively) and the distribution has become more bell shaped:

There are more people with extreme scores than a perfectly normal distribution would predict, especially at the high end. It’s unclear why this is but several possibilities come to mind:

  • The test items did not increase in difficulty in a sufficiently linear way.
  • A bias in the sampling
  • Some of the people tested are not part of the biologically normal population
  • Monkey business

To normalize the distribution, each raw score was converted into a percentile rank which in turn was converted to the expected IQ on a perfectly normal curve. But because research suggests my average reader is two standard deviations smarter than the average American, 30 IQ points were then added.

These norms look plausible at the high end but seem too generous at the very low end. Future normings will use equipercentile equating with self-reported SATs, ACTs and Wechsler IQs to see if that gives different results.