Here is the secret for getting high scores on tests like PISA, TIMMS and so on: systematically exclude any student likely to produce low scores.
In Singapore, the children of the local indentured servant class and the children of migrant workers who cross from Malaysia every day simply are not counted because they are not permitted to attend schools in Singapore at all.
In China, even though students in a number of provinces are tested and measured on PISA by the OECD, the Chinese government only permits scores to be published from the city of Shanghai — where half of the school-age children simply are not allowed to attend school or receive any services at all, since they theoretically and legally belong to their home town out in the rural provinces somewhere.
I strongly recommend reading this article, on Diane Ravitch’s blog, as well as my wonderfully edifying comment.
and here is the original article by Tom Loveless of the Brookings Institution, and here are a few paragraphs from it:
The only reasonable conclusion is this: officials in Shanghai are only counting children with Shanghai hukous as its population of 15 year-olds, about 108,000. And the OECD is accepting those numbers. It is as if the other children, numbering 120,000 or more, do not exist. This is not a sampling problem. PISA can sample all it wants from the official population. Migrant children have been filtered out. Professor Chan of Washington agrees with this hypothesis, saying in an email to me: “By the time PISA is given at age 15, almost all migrant children have been purged from the public schools. The data are clear.”
As a researcher who studies student achievement, I use PISA data. That requires trust and confidence in the integrity of the assessment. I can be confident, for example, that the scores from Portugal are from a representative sample of all 15 year-olds in Portuguese schools. I have no such faith in PISA scores from China. PISA-OECD has been silent about its special arrangement with China. All of the data from 2009 still have not been released. The data from Shanghai apparently only represent the privileged subset of 15 year-olds who hold Shanghai hukous. I don’t know for sure. In the four volumes of data on PISA 2012, neither hukous nor the migrant children of China are discussed. Not a word. Not a peep.
PISA officials are not shy about offering policy advice to countries, especially policies that the OECD believes will promote equity. Delaying tracking and ability grouping, reforming policies governing immigration, distributing resources so that schools with less get more, and expanding early childhood education—all have been promoted as equity-based policies. But not a word about reforming hukou. Not a word on a discriminatory policy affecting the education of millions of Chinese children. Not a word on the human rights story of migrant families in China and the human suffering that they must endure.