Does Beer Make Bud Wiser?

Here from Razib Khan at Gene Expression is a chart of percentage alcohol drinkers against score on the Wordsum test (0-10 score; a quick and dirty correlate .71 of adult IQ.).  I will let you debate causality.



Alcohol kills brain cells, but only the weakest ones. Thus the _average_ quality of your brain cells increases.

I'm a bit surprised no one in razib's comments mentioned that drinking is often a social event with lots of chattiness, and I'd think chattiness would be associated with higher wordsum scores (enter debate on whether the wordsum test is g loaded...), since a person who talks more will have the opportunity to hear and speak more and different words.

it seems odd that he doesn't see wordsum as a proxy for education, which he references as seperate. since Graduate degrees drink more than GEDs by a suprising margin, is it any suprise that more wordsum drink more than less?

At first I misinterpreted the chart as X axis being "when tested after that many beers" and Y axis being "score on % scale".

That would be a lot more interesting causality ;-)

The real one is probably something as trivial as dumber people being more religious.

Were you drinking when you wrote that headline?

Not sure how WordSum works. Do curse words count? I know I somehow know a lot more of those when drinking.

Was thinking a bit about this and trying to compare to my own personal experience and social circles.

I realized that other than pregnant women, I know literally no one who doesn't drink.

This suggests to me a big class overlay which is probably just coincidentally working in correlation with a class-skewed measure of IQ (whether or not IQ itself is class skewed, what is cause and what is effect, etc, are probably best left to a less fun thread).

Alternately, this could indicate that pregnant women suffer temporary declines in IQ.

I think Ed might be right sadly. If you correlate quality of education, the influence of the Souther Baptist, Pentecostal and evangelical churches, and dry counties. You'll find a lot of rural people receiving poor educations with high access to religion and low access to alcohol.

Marapeese rule!

Razib, thanks for the encouragement. Here are my favorite charts so far.
ROW = drink
COL = comprend

ROW = wrkgovt
COL = wordsum

ROW = tax
COL = wordsum

That GSS browser is great fun. My experiments indicate that both the religious effect and the income effect is important.

Row = drink
Column = wordsum
Selection Filter = relig(4), income(12)

(I.e. just look at atheists earning over $25k.)

You'll then see that the remaining abstention percentages aren't statistically different from the mean even at 90%. Just looking at atheists, or just looking at people earning over $25k leaves the effect still statistically significant.

Of course the sample size left after making these selections is pretty tiny, so maybe in a larger sample the effect would still persist.

This is exactly right and also explains why limiting our economic potential is not going to improve global egalitarianism.

you can find more such article in

I hope this allowed, I have never used this website before so I wasn't really sure what this was going to do. So this is just a test post. I really like this forum, it has some excellent discussions that take place.

Comments for this post are closed