Friday, July 22, 2016

Corpulent knuckleheads

In his signature style, Milo Yiannopoulos noted in his July 15 podcast that "fat people are stupider than the rest of us". The study his remark was based on used detailed brain imaging to obtain its results from a modest sample of 32 people. It found fat people to be less intelligent and more impulsive than non-fat people .

The GSS offers a cruder and less precise way of looking at the same thing, though our sample is 20x as large.

The following table shows average IQ estimates (converted from wordsum scores and assuming a white average of 100 and a standard deviation of 15) for female respondents whose weights were deemed either "below average", "average", "somewhat above average", or "considerably above average". Those who conducted the interviews were the ones who assigned respondents to one of the four weight categories.  To avoid language fluency confounds only survey respondents born in the US are included (n = 616):

HeftIQ
Below average100.5
Average100.3
Somewhat above99.1
Considerably above94.8

Another stereotype validated.

Fat and stupid, yes. But famous, too!

GSS variables used: INTRWGHT, SEX(2), BORN(1), WORDSUM

7 comments:

Anonymous said...

You have to account for the fact that black women are the fattest group. Their average IQ is 85 at best. Maybe they're fatter because they're dumber, but to be able to say so with confidence you would need to break it down by both weight and race.

Audacious Epigone said...

Good point, though the difference between "considerably above average" and the rest is even larger for white women than it is for all women. Sample sizes are too small to look at any other racial group.

Anonymous said...

Why females only though? The males don't show the same results in the analysis, and the results for the overall sample (males plus females) are similar in direction but reduced in size.

Audacious Epigone said...

Anon,

I didn't think to look at men because Milo's comment gave me the idea for the post but wow, that is an interesting contrast, thanks.

My first response would be that weight/physical appearance aren't as important in society to men as they are to women. Heavier men can still lead other men and be respected by other men and get attractive women. Heavier women can't really do any of these things. So being a fat woman is a real drag while being a fat man isn't as bad. But that's just a guess.

Anonymous said...

Thinking about this earlier, and using another approach to this, on GSS you can use the self reported height (height) and weight (weight) variables to calculate BMI, by weight*703/height^2.

That gives:

BMI by Wordsum - http://imgur.com/a/W7H3f
Wordsum by BMI - http://imgur.com/a/Ds0dx

(These are the Wordsum and BMI ranges with decent sample size btw)

What it looks like is:

- Overweight women (BMI 25-29) have normal Wordsum, same as low and lean BMI women, while obese women (BMI >29) have lower Wordsum
- Among men, no association between BMI and overweight or obesity (fat neckbeards and bodybuilders aren't necessarily unintelligent, I guess)
- Different Wordsum groups don't really have different BMI. This is probably because the number of obese women, who are the only BMI group with distinctive Wordsum, is relatively small, so doesn't contribute much to the overall result.

JayMan said...

Milo is of course correct here, on average.

This is confirmed by two large longitudinal studies discussed in my post

Obesity and IQ - The Unz Review

The longitudinal nature of these studies make clear that it's not that obesity causes lower IQ. Rather the relationship stems from genetic pleiotropy.

Anonymous said...

Anon above here again:

Rounding out the analysis by BMI: http://imgur.com/a/dk1Bb

BMI by Degree, Father's Degree, Self Assessed Social Class, Inflation adjusted income, Self Ident as Liberal or Conservative.

Again, for the males, none of these relate to BMI in the sample.

Trends in women are mild, but negative correlation such that higher class, education, income, paternal education is associated to lower BMI. Particularly paternal education. It seems like politically extreme women at both ends of the spectrum may tend to be heavier, though small numbers of them so take that how you will.

Unless there's a massive, systematic error in self assessed height and weight, these traits seem not very relevant to high BMI in men. Some relevance in women, but average differences between most and least intelligent / successful only equivalent to least overweight vs overweight at the obesity cutoff. All groups are overweight. That reflects what you would expect for the modern US - every group is overweight (on average) across class and intellectual barriers, compared to the past, when every group across class and intellectual barriers was lean (on average).

Also includes

- self assessed health and happiness - predictably stronger relationship with BMI with happiness for women than men.
- region and urban / rural, non-hispanic Whites only. urban lower BMI than rural, monotonically.