Great article and statistically significant study. It controls for income/socio-economic classes and the U.S.' second and third quartile income students do quite poorly.

This was a convenience sample. Convenience sample means "these are 105 high schools that we were able to include in the study conveniently, and so those are the ones we used". Convenience samples are, by definition, statistically unrepresentative. So what can we conclude about the US as a whole? Nothing.

Also, "statistically significant" may not mean what you think it means.

Anonymous wrote:http://www.centredaily.com/2013/04/04/3565862/thomas-l-friedman-global-schools.html

Great article and statistically significant study. It controls for income/socio-economic classes and the U.S.' second and third quartile income students do quite poorly. The test is comprehensive as well, unlike state tests where memorizing format and "teach to the test" is considered learning.

"'U.S. students in the second quarter of socioeconomic advantage — mostly higher middle class — were significantly outperformed by 24 countries in math and by 15 countries in science, the study found. In the third quarter of socioeconomic advantage — mostly lower middle class — U.S. students were significantly outperformed by peers in 31 countries or regions in math and 25 in science.',