silly refutations of kanazawa’s blog post

there’ve been a lot — a LOT — of refutations of kanazawa’s post on the attractiveness of black women. i haven’t read them all — in fact i’ve read hardly any of them ’cause most of them just scream and yell WAAAAAAYCIST!! BURN HIM, BURN HIM!!


some of them, however, appear to be more scientific refutations. bering in mind links to another psychology today blogger (scott barry kaufman) who (along with someone named jelte wicherts in the netherlands) has supposedly (according to bering in mind) “failed to replicate” kanazawa’s findings.

i took at look at kaufman’s post — and the technical summary of their analysis — ’cause i thought, well gee — not able to replicate the findings — that’d be interesting.


here’s the evidence that kaufman and wicherts present (in the blog post) to show that kanazawa’s analysis was incorrect:

looks like black women were rated nearly as attractive as white women, right? and they were. in wave iv.

here’s how they fudged the data.

the add health thingie (from whence all the data comes) involved four waves of surveys over the course of several years (a couple of decades?). kaufman and wicherts decided that the only wave that should be included in any analysis on the attractiveness of women is wave iv, because in that wave the females were of legal age and, therefore, women.

no, i’m not kidding. they really said that!

now, i would agree with them if the subjects in the previous waves had been pre-pubescent children. but the ages in the waves were:

wave i = mean age 15.9 years
wave ii = average age 16.5 years
wave iii = mean age 22.1 years
wave iv = mean age 28.9 years

now come on! ok, so in waves i-iii most or all of the subjects were not of legal age, but probably the vast majority (except maybe for some late bloomers in wave i) were “reproductively of age” — meaning they could make babies. which is what the whole discussion is about! attractiveness, after all, is about attracting a mate.

*cough, cough* cherry-picking *cough, cough*

kanazawa didn’t include wave iv in his analysis, which is the wave when the attractiveness of whites and blacks was rated the most similar. don’t know why he didn’t use the data from that wave. kaufman mentions that the data from that wave has been available “for over a month.” well, maybe kanazawa didn’t realize the newest data had been published when he ran his analysis. i really dunno, but it’s definitely possible.

in any case, waves i-iii show that black women were rated as less attractive, altho i think the numbers in wave iii are not statistically significant. in wave iv, as i’ve said above, the rating for whites and blacks were pretty similar:

i still wanna know who the interviewers|evaluators were. were they all white folks? all asians? all illegal mexican workers picked up outside home-depot? their characteristics might’ve influenced the results.

actually, now that i mention it, the fieldwork for waves iii and iv was contracted out to a different company than waves i and ii. wonder if that made any difference in the evaluations?

btw, some real word data from okcupid maybe lends more support to kanazawa’s findings:

“Men don’t write black women back. Or rather, they write them back far less often than they should. Black women reply the most, yet get by far the fewest replies. Essentially every race—including other blacks—singles them out for the cold shoulder.”

that’s too bad.

previously: the offensive mr. kanazawa and african-american porn stars

update 06/07: see also on the add health interviewers

(note: comments do not require an email. or an subscription.)


  1. The problem with the Add Health data is that the inter-rater reliability of the attractiveness ratings is very low, which is something Wicherts and Kaufman point out as well. Moreover, there were only a few raters.

    More interesting was Kanazawa’s finding that blacks themselves think they are very attractive, much more so than other races. This does not lend credence to the narrative that normative white standards of beauty make blacks feel bad about themselves.


  2. @jl – “The problem with the Add Health data is that the inter-rater reliability of the attractiveness ratings is very low, which is something Wicherts and Kaufman point out as well.”

    yes. that’s why i’d like to know who the evaluators were (see post above and previous post). methodology would be nice to know, too, of course. for instance, were they given instructions on what to look for re. attractiveness?

    @jl – “Moreover, there were only a few raters.”

    there probably weren’t a very few raters. i addressed this question in a comment on another blog. here’s what i had to say:

    there were definitely more than 3 raters in the add health surveys.

    what kanazawa said was: “The physical attractiveness of each Add Health respondent is measured three times by three different interviewers over seven years.”

    the “three times” refers to the first three waves in the add health survey.

    in the first wave, 12,105 adolescents were interviewed for 1-2 hours during a 9 month period.

    if all the interviews were 1.5 hours, interviewing 12,105 subjects would amount to over 2270 eight-hour days.

    there’s only ca. 21.6 working days per month on average x 9 months x 3 interviewers = 583 eight hour days. but you’d need 12 interviewers to cover 2270 eight-hour days.

    and that’s just in the first wave. wave ii saw 15,000 students interviewed; wave iii, 15,170.

    also, the fieldworkers used typically work only a 20-25 hour week, so there’s at least twice the number of interviewers required than i’ve guessed at here.

    in any case, n≠3.

    there is still, however, the problem of the race and gender and age of the interviewers. that would be essential to know (along with the number!).


  3. It probably makes sense to leave out wave iv since women are most fertile between 16 – 24 years of age. From 25 onwards, their fertility is in decline which should mean that their attractiveness will also be in decline.


  4. @joe w. – good point. altho, i don’t think that’s why kanazawa didn’t include it. i think he wouldn’t have been afraid to have said as much! (~_^)


  5. Looking at wave 1, about 60% of white women were rated attractive or very attractive vs 45% of black women. So, big deal. Its not like none of the black women were rated very attractive. 45% with 14.6% very attractive is a hell of a lot of attractive black women.

    There may be an average difference but when there’s so many attractive black women, I wouldn’t get worked up about it. (did the raters rate differently according to their race?…do people have a bias towards their own race in perception of attractiveness?)

    The average difference also is not enough to justify the dearth of beautiful black female movie stars (there are hardly any) vs the number of male stars (lots).


  6. Here is something else you missed from data.

    In contrast, when the participants were asked to judge their own individual physical attractiveness, both Black men and Black women rated themselves as far more physically attractive than men and women of other races. The self-reported average physical attractiveness score for Blacks were about double those of other races

    Well black guys gave themselves a nice bump in data by voting 10s for themselves.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s