IN THURSDAY’S Advertiser Des Morgan writes again supporting selective education.

I only want to comment on his claim that “statistics can be used to prove streaming works.”

There is now a long history of studies on many aspects of this question and the related questions of setting and in class grouping.

Individually and in meta-analyses some conclusions are common.

The studies show no significant differences in attainment overall between streaming and mixed ability teaching.

The lower attaining students achieve higher attainment in mixed ability. This is true for both primary and secondary schools and in all subjects. The most able sometimes show marginally higher attainment at best in top sets, but even this may be due to participation in individualised programmes occurring in addition to normal lessons.

Grouping within a class shows significantly better progress than setting or streaming.

Children in lower sets have more negative attitudes than in mixed ability. Some studies have even shown worse attitudes amongst the higher attaining too.

Internationally this picture is even clearer.

Des’ ambivalence on statistics might be due to the lies the Tory press tells about education statistics.

In 2006 they misreported a paper in the British Journal of Educational Psychology. They said it showed streaming benefitted under achievers.

But the paper said nothing about achievement, under or otherwise.

It said, “Results showed that the lower-ability stream students had a more negative academic self-concept than the higher-ability stream students immediately after streaming.”

After three years it said their self-concept had improved.

PETER SMITH Woodside Avenue, Swindon