The book this "evidence" comes from is the much derided "Bell Curve." The poster is a HN racist troll. I checked his comment history and he pretty much only posts stuff similar to this.

I gave a link to stereotype threat earlier - http://isteve.blogspot.com/2012/10/john-list-on-virtual-none... And keep in mind, even if the studies were replicable, the studies did not even show that stereotype threat shows up in the real world, or that it causes the one standard deviation gap. What the original study showed was that on a non-meaningful test, if the professor suggests a stereotype, the gap in test scores grows bigger than one standard deviation.

I think the whole idea of stereotype threat is ridiculous. Just because of a stereotype I'm going to get dozens of questions wrong on the SAT? That just does not make any sense. (And now that we have a black president who was from Harvard Law, wouldn't we see average LSAT scores start to converge, surely there would be less of stereotype against black people in law).