Reply #25: This is why professors have tenure. I looked up his "study", and he actually makes a point.

[Xithras]

It's not a study at all, but rather is a blog post discussing the dataset taken from the latest wave of the Add Health study. For those unfamiliar with it, Add Health is a recurring study ordered by the U.S. Congress and administered by various agencies, and now has a recurring dataset running back to 1994. The purpose of Add Health is to study the mental and physical health of American adolescents, plot changes to that health over time, and has now become the largest survey of adolescent health ever performed. Kanazawa was NOT involved in it.

One of the aspects of Add Health is a personal image/beauty analysis. Participants are asked to rate their own attractiveness and intelligence, and then the interviewer uses a number of race-neutral measurements to assign an actual beauty/attractiveness rating to the individual. The attractiveness score of the respondent is assessed three times by three different interviewers over a multi-year period, and their "attractiveness score" is the average of the three ratings. This can be useful for depression studies and all sorts of other research purposes.

When Kanazawa ran the numbers for the latest wave of the study, he found something interesting. Black respondents consistently rated their own attractiveness higher than all other racial groups. Blacks are far more likely to think they're beautiful. At the same time, the actual beauty score for black respondents (based on the profile developed by the NICHD and not Kanazawa) was substantially LOWER for black women than for all other racial groups. The typical black woman, according to the Add Health data, is uglier than the average woman from other racial groups. This difference does not carry across to black MEN, who had beauty/attractiveness scores roughly equal to the other racial groupings.

Kanazawa isn't calling black people ugly, but is instead pointing out that the Add Health data says they are uglier. He doesn't really address why, though he does raise and then shoot down weight as a theory (the scores are lower even with BMI factored out), and posits that testosterone levels (which are higher, on average, in black women) may have something to do with it.

After actually READING the blog post (I admittedly skimmed some of it), the only objection I had was that he didn't raise the possibility that the attractiveness rating process used by Add Health was itself flawed. He simply talked about the numbers presented by the study, without ever questioning the validity of the data itself.