This flies in the face of Western government policies which often go to significant lengths to increase ethnic diversity in schools
"Study: Ethnic diversity at school has a negative effect on learning
Fifteen-year-old pupils from schools with high ethnic diversity perform worse than comparable pupils from schools with homogenous student populations. This applies not just to the immigrant children, but also to the pupils from the country in question. For the latter group, the negative effect is strongest in school systems with a hierarchy of school types, such as the Dutch and German systems.
What’s more, the number and origin of the immigrant pupils also plays an important role. Having a higher proportion of pupils from Islamic countries at a given school negatively influences the performance of all pupils at that school. But in contrast, a higher share of pupils from South and East Asia has a positive effect. This is just one of Jaap Dronkers’s conclusions from his empirical research using international PISA data, which he discusses in his inaugural lecture.
The educational performance of immigrant pupils gets a boost from having a higher proportion of such pupils at a particular school – but only when the fellow immigrant pupils come from the same region. This applies particularly to immigrant children from Islamic countries and from South and East Asia. This positive effect of ethnic homogeneity in schools might explain the attraction of Islamic schools, for example, or white or Jewish schools."
Full Article
Reading the article more closely suggests a more complex finding. First, this took place in Europe where most of the diversity is due to immigration rather than other factors, such as busing. European countries also pose the challenge of a different educational system when trying to compare. The study also suffers from a correlational analysis so we do not know exactly what causes the findings -- only that they appear to be related. When immigrant diversity promoted by Islamic background goes up, then performance goes down. I can think of a number of reasons for this - none of which can be tested with a correlational analysis.
ReplyDelete