3 ms·
"Salzman says that reports citing low U.S. international rankings often misinterpret the data. Review of the international rankings, which he says are all based
by Xichekolas 18y ago
"Salzman says that reports citing low U.S. international rankings often misinterpret the data. Review of the international rankings, which he says are all based on one of two tests, [...] show the U.S. is in a second-ranked group, not trailing the leading economies of the world as is commonly reported."
If I am understanding that correctly, that is not so much misinterpreting the data as not being able to frickin read.
Misinterpreting data, to me, signals an improper application of statistics to some raw data to assert something.
Not being able to look at a grouped list of rankings and realize that just because a certain group as a whole is below another group as a whole does not imply that every country in the second group is below every country in the first group is, well, sheer idiocy, not merely misinterpretation.
If the basis of the 'US is behind in math' idea is really that policy types and journalists can't properly interpret the set of rankings, then it really scares me that they are in charge of our public discourse/policy.
- ig1 18y agoThe OECD rankings look pretty clear to me: http://www.infoplease.com/ipa/A0923110.html http://www.infoplease.com/ipa/A0923110.html http://news.bbc.co.uk/1/hi/education/7126388.stm http://news.bbc.co.uk/1/hi/education/7126388.stm
- kragen 18y agoFor those who don't follow the link, the first link places the US, with 95% confidence intervals: - 25th to 28th among the 40 countries on math; - 12th to 23rd among the 40 countries on reading; - 20th to 27th among the 40 countries on science. The second link just divides the countries into above average, average, and below average. The US is "below average" on math, and no reading rank is reported.
- tokenadult 18y agoIt's often said that the countries that score better than the United States in internationally standardized achievement tests are only testing a high-ability sample of their population, while the United States is testing a representative sample. But this is baloney. One example of a webpage about this issue is http://www.wgquirk.com/timss.html http://www.wgquirk.com/timss.html More generally, I know the assertion is wrong because I can read the languages of other countries and I have lived overseas and I have observed their school systems and how the common people of other countries live in daily life. Several of the newly industrialized countries of east Asia and southeast Asia have a much higher level of numeracy, such that algebra is a standard curriculum subject for all students in seventh grade, including the below-average students, and such that people from those countries think it is bizarre that American store clerks count out change from a purchase for their customers, as if the customers can't calculate their own correct change mentally. Simply put, the policy analysts and journalists who say the United States is behind other countries and thus behind where it should be (given the very large expenditure on K-12 education in the United States) are correct. A good point made by a previous reply here is that sci-tech skill sets are relevant in more than just sci-tech jobs. It would indeed be good if more lawyers, physicians, business managers with business degrees, and technicians and industrial workers without college degrees had strong knowledge of math and science. That is what makes manufacturing successful in many developing, low-wage countries: the workers on the factory floor and their foremen have a basic level of technical skill that allows efficient manufacturing that can produce products with a finished cost (including shipping to distant markets) and level of quality that is competitive in international markets.