News for hackers. Stuff that matters.
Re:After correcting for demographics, US schools fare better... by pmorici on Tuesday December 28, 2010 @12:59PM (#2046136)
Thats not what he concludes. It's "as long as your parents are educated".
Re:After correcting for demographics, US schools fare better... by Alex3917 on Tuesday December 28, 2010 @01:34PM (#2046264)
And he's wrong again. The reason high-SES kids score better on standardized tests isn't because they learn more in school, it's because they learn more at home. In fact low-SES kids actually learn almost as much at school than high-SES kids[1], it's just that they start out 2-3 years behind in Kindergarten[2], and instead of learning over the summer they actually regress[1]. What the data actually shows is that students DON'T learn if they're white, but in a feat of epic illogic this guy has somehow managed to convince himself that the opposite is true.

[1] http://www.eric.ed.gov/PDFS/ED037322.pdf (C.f. figure 2 and 3)

[1] http://www.amazon.com/Meaningful-Differences-Everyday-Experi...

Re:After correcting for demographics, US schools fare better... by towelrod on Tuesday December 28, 2010 @02:22PM (#2046421)
He didn't control for parental education, he controlled for recent immigration status.
Re:After correcting for demographics, US schools fare better... by pmorici on Wednesday December 29, 2010 @01:41AM (#2048157)
It seems like he is using immigration status as a proxy for parental education level. As he later notes the effect is non existent and in fact opposite in Canada and Australia which he attributes to the skills based immigration system.
Re:After correcting for demographics, US schools fare better... by kenjackson on Tuesday December 28, 2010 @03:39PM (#2046676)
Which is unfortunate because parental education and income are in the dataset. I wish I had the time to sift through it.