Reader Comments

Post a new comment on this article

Undersampling undermines results. A comment to Heck, Simons, & Chabris (2018)

Posted by apinon on 12 Aug 2019 at 14:33 GMT

The results presented by Heck, Simons, & Chabris (2018) are unsurprising. Individuals with moderate, severe, and profound mental retardation are unlikely to respond to any survey. This means that a random sample of the entire population cannot be obtained. Consequently, the expected percentage of people who claim to be above average in intelligence must be higher than 50%, because the lowest group is undersampled.
Asking individuals a question about their intelligence doesn’t allow to find out how many of them overestimate their IQ. The study should have included an IQ test, in order to compare the test results of individuals with their belief to be above average.

No competing interests declared.