The best way to describe this argument is "not even wrong."
Judging the quality of our schools solely by using standardized tests to compare our students with students in other countries is moronic, and excluding minorities is even more so. Yes, of course parenting and culture have an enormous effect on student outcomes, but this data doesn't actually illuminate anything of use. Don't get me wrong, it makes an interesting talking point, but there's nothing here that you can actually use to make schools better.
The claims made by this article are easily falsifiable: do a more accurate ethnicity-based slicing of the data and show that US schools fare poorly as measured by standardized tests.
You may not find the results useful for your chosen aims ("make schools better"), but that's irrelevant. Is evolution also "not even wrong" because it can't be used to "make schools better"?
I don't understand how that follows from the data? Even comparing apples to apples, so to speak, the US is at the 25th %ile. With most countries above (and below) being likely countries, once you exclude the poorest 35% from our country.
To put it another way -- stacking the cards in our favor should really put us a lot higher up the stack, considering he didn't control from spending or income.
Judging the quality of our schools solely by using standardized tests to compare our students with students in other countries is moronic, and excluding minorities is even more so. Yes, of course parenting and culture have an enormous effect on student outcomes, but this data doesn't actually illuminate anything of use. Don't get me wrong, it makes an interesting talking point, but there's nothing here that you can actually use to make schools better.