Playing the Indian Card

Friday, October 08, 2010

The Only Olympics That Matter

Which country has the best public schools?

This is an obviously useful question. If we know which practices are best, we can emulate them.

There are two international surveys that try to find out: the Trends in International Mathematics and Science Study (TIMSS), and the OECD's Programme for International Student Assessment (PISA).

Both study the same three core subjects, reading, science, and math. Both rely on standardized tests, which are of course limited in what they can measure. The PISA includes all and only countries belonging to the OECD, meaning only developed countries. The TIMSS surveys a somewhat larger and broader group of countries.

I have mashed their rankings together, on the assumption that this produces a more reliable “meta- study,” somewhat corrected for any flaws in survey or in test design, and with a larger statistical sample. Here's what I get, simply by taking their country and subject rankings, adding them up for each country tested, and dividing by the number of tests in which each country participated. Low scores therefore suggest the best system. Here is my cumulative top twelve, with scores:

Finland – 1
Taiwan – 1.5
Netherlands – 3.67
tie – South Korea and Japan – 3.75
Canada – 4.33
Latvia – 5
England – 6
Singapore – 6.33
New Zealand – 8.67
Switzerland – 9
Belgium – 9.5
Lithuania – 9.67

I suspect these rankings end up being pretty accurate; I am reassured that this is so by seeing them tend to bunch into culturally related groups of countries, as these countries are likely to have related educational traditions. There is the Nordic, Finno-Ugric group (Finland, Latvia, Lithuania). There is the Confucian East Asian group (Taiwan, South Korea, Japan, Singapore), there is the Anglo Commonwealth group (Canada, England, New Zealand), there is the Flemish/Dutch group (Netherlands, Belgium). The presence of each group reinforces the reliability of the meta-survey, and these supranational groups seem to stand out more clearly in this meta-study than in any of the individual studies that made it up.

Add two nations that tied for thirteenth place,and you add one member each to two of these leading groups: Australia to the Anglo Commonwealth, Hong Kong to the Confucian.

The precise rankings of individual countries are less reliable, I should think; since not all nations submitted to all tests, there is a variability in the quality of information available. It would also be possible for a shrewd national government to pick and choose the studies to submit to based on their own perception of strengths. The solidest data is for South Korea, Japan, and England. Achieving reading literacy can also be tougher for countries without a well-established and widespread national language, and intrinsically easier for countries that are ethnically and culturally homogenous. Some writing systems are also intrinsically easier to learn than others.

The next step is to try to identify what, if anything, these supranational groupings have in common.

The most obvious answer: these are nations that, barring bad behaviour by near and bigger neighbours, have been most notably successful overall economically, scientifically, perhaps militarily, and in terms of civil order--if not necessarily in the arts.

Obviously, there is a connection.

But is it better education, or better discipline of the sort that suits one to write standardized tests?

No comments: