[tgb]

Well, they're tracking the location of the university not the nationality of the student and the absolute best students go to MIT/Stanford even if they're from outside the US. I haven't read the paper but the article here says:

> But the researchers corrected for this by separating out anyone who didn't list English as their native language; the gap was unaffected.

But this only corrects for half the equation; it removes foreign students from elite US institutions but does not "add them back" to their home country's results. Tsinghua may have a gap of students that went to Stanford instead. Though I'm a little skeptical that this accounts for much of the effect. China and India have too many top students for the US to steal enough of them to effect the outcome this severely and it's unlikely the test is able to separate the one-in-a-million students from the one-in-a-thousand stduents.