6 Comments
User's avatar
⭠ Return to thread
Steve QJ's avatar

"A ten point difference would be unmistakable and not masked well by overlap."

I admit I'm not an expert, but wouldn't that depend on where the difference occured on the scale? The diference between a 70 and 80 IQ would be very clear I think. But would that be the case for a 100 and 110?

But yeah, I guess the wider point is that definitive proof of an average racial IQ gap wouldn't change much about how society ran.

Expand full comment
Chris Fox's avatar

Ten points is a LOT no matter where it is, except maybe for the difference between 170 and 180. A 110 IQ person is qualitatively different from a 100; more intellectually engaged, more curious, more likely to be a reader.

Even between 70 and 80 you have the difference between someone who can care for himself and hold a menial job and someone who likely can't.

Expand full comment
Steve QJ's avatar

Ah, okay, fair enough, in that case, pretend I wrote "two or three" instead of "ten."

Expand full comment
Chris Fox's avatar

I was imagining something like 0.2 points. The bigots would take that and say all black people are retarded. Real people would call it statistically insignificant.

Expand full comment
Peaceful Dave's avatar

If I'm not mistaken +/- one standard deviation from nominal is IQ 85-115 which is the middle of low and high average (68% of humanity). IQ 70-130 is two +/- standard deviations (95% of humanity). Since we don't wear a red badge of IQ we don't really know the IQ of people we work with beyond levels of perceived competence. I doubt that I could guess anyone's IQ with accuracy, but then I've never cared to try.

Expand full comment
Chris Fox's avatar

You're not mistaken, those figures are right. And ± 3SDs is 99.7%.

I would swear that in the past it was 10 points that comprised a standard deviation.

I wouldn't try to guess numbers either but there is an unmistakable indication when you're talking to smart people: they grasp what you're saying before you finish the sentence.

Expand full comment