Saturday, November 27, 2010

Gays and Straights: No IQ difference

Using GSS data, I found in an earlier analysis that gay men and lesbians had higher average IQs. To check this, I examined mean IQs by sexual orientation for adolescents from the Add Health Study. IQ is measured with a vocabulary test. Students were asked if they were ever sexually attracted to a female and/or a male. I limited the analysis to whites: the overall mean is 104.3 (it is high because the mean for the whole sample that includes many NAMs is set at 100).

Mean IQ score (sample size = 6,145)

Bisexual female 108.1
Straight male 105.9
Lesbian 105.9

All whites 104.3

Bisexual male 104.3
Straight female 104.0
Gay male 102.6

There are a range of means, but the only statistically significant difference is between bisexual and straight females. It's close to one-third of a standard deviation difference, which is fairly small.

Unlike the GSS analysis, lesbians and gay males do not appear to be smarter than other groups. The GSS asked about sexual behavior, not attraction. It might be the case that smarter people are more likely to act on non-conformist desires.   


  1. Hey. I am happy I've found this page, I've been trying to remember the web address. Anyway I would just like to thank you for your time and the posting, I've enjoyed it. I have saved as a favorite it and also subscribed to the RSS feeds.

  2. Possibly more intelligent people are able to move into occupational roles which then move them into communities in which they are in more of a position to express that sexual behaviour (i.e. that community is more permissive on that axis of behaviour or just more atomistic generally).

  3. It could be that the method of selecting the sample may have controlled-away any possible differences (I can't see the paper to check).

    In other words it might be a matter of range restriction.

    For example, if the samples come from a particular college - taking students at a particular stratum of educational attainment - then the range of IQ sampled may be so narrow that IQ differences between different types of students would be made insignificant.

    This would be compatible with there being significant homo- versus hetero- sexual IQ differences detectable in a random sample from the national population.

  4. "Ever?" So, in the world of this analysis, an exclusively behaviorally heterosexual person who once and only once had an episode of same-sex-attraction (and then admitted it) is "bisexual?" Not criticizing you: the survey has the questions it has.

    What %age of people are bisexual by this measure? If it's less than, say, 50%, then we should suspect that there is a lot of lying going on.

    Are high IQ people better at answering questions? Are they more honest?

  5. Anonymous12:08 AM

    The most obvious problem with the data is that it shows any category of woman with a higher IQ than any category of men.

    Given the lower mean spatial ability of women, their IQs normally work out to be 3-5 points lower on average.

  6. Maybe lesbians have higher spatial ability? For whatever reason, I knew quite a few smart bi women...

  7. Maybe lesbians have higher spatial ability? For whatever reason, I knew quite a few smart bi women...

  8. Yo creo que todos somos iguales independientemente de nuestra condiciĆ³n sexual...

  9. This Site Learn More Here replica ysl handbags find out here now


How much has mean IQ declined in the US?

 I looked at General Social Survey data to see how much IQ has declined in the US since the early 70s when the GSS was started. I set whites...