In his fascinating book The Folly of Fools, Robert Trivers theorizes that people are very prone to self-deception in order to better deceive others to one's advantage. You lie more effectively if you believe the lie yourself. Since having status is particularly important for men, they are more likely than women to inflate their abilities. To test this, I looked at Add Health data. I examined the relationship between gender and self-described intelligence, controlling for objective intelligence (based on a vocabulary test). I also added race/ethnicity to the model to see if groups differ in the tendency to exaggerate one's intelligence. All minority groups are compared to whites (who serve as the reference category):
Standardized OLS regression coefficients (sample size = 6,169)
Objective IQ .32*
Male (vs. female) -.02
American Indian -.01
First, people do have a tendency to rate their intelligence correctly--self-described intelligence is positively correlated with measured IQ--but the tendency is only moderate. Next, males are not more likely than females to inflate their smarts. By contrast, blacks are significantly more likely to--compared to whites. The other racial/ethnics groups do not differ from whites.
Regarding blacks, I'm not sure if this is a racial trait or something that is induced by living in a society with a high average IQ. Hispanics and American Indians have lower than average IQs, but do not inflate their intelligence more than whites, which makes me suspect that above-average boastfulness is a black thing.
One other item--the results do not support the claim that Asians underestimate their IQs (out of cultural modesty). The tendency for Hispanics to do so falls just short of statistical significance.
A common claim by IQ skeptics is that the test simply measures social class. If an IQ score is simply a proxy of social status, then the tw...
In the comments in the last post , some readers contended that Jews are not ethnocentric. Using the same question I used in the comments se...
Which factor reduces family size the most? Below are the standardized OLS regression coefficients for a sample of whites ages 40-59: Stand...
Via a reader at iSteve, it looks like this might be the vocabulary test used by the General Social Survey. (Someone please tell me if I'...