Children of average ability make better progress in grammar schools than in comprehensives, but the wider impact of selection is overwhelmingly negative, according to a new report from the National Foundation for Educational Research.

The report, which looks at the impact of selection on pupils in Slough, backs the findings of a national study published by the foundation last October.

Both studies found that the most able pupils perform just as well, if not better, in comprehensives as in grammar schools. But borderline grammar-school pupils - who may have passed their 11-plus by as little as one mark - do significantly better than children of similar ability in comprehensives.

This contradicts the widely-held view that borderline pupils are better off in non-selective schools. “It accords rather with the alternative view that such young people need the challenge and encouragement ...(of) mixing with high-ability students,” say Ian and Sandie Schagen, the reports’ authors.

Their national study examined how much value schools were adding in three types of education authority: fully comprehensive; those with up to 20 per cent of secondary pupils in grammar schools (“low selection”); and those with more than 20 per cent in grammar schools (“high selection”).

Studying pupils’ progress between key stage 3 in 1998 and GCSE in 2000, the researchers found that “high” and “low” selection authorities had a slight overall advantage over others. They also discovered that selection had the greatest impact on pupils of middle ability. Those at grammars in low-selection areas obtained better GCSE results than pupils of the same ability elsewhere (ability being measured by prior attainment).

A similar value-added analysis looked at progress during KS3. Using KS2 results for 1997 as a baseline, it showed that both high and low selection authorities get better results overall than areas with fully comprehensive systems.

Again grammars had the biggest impact on their least able pupils.In maths, English and science, results for pupils with an average KS2 level of 4.5 were similar in secondary modern schools and comprehensives. But in grammar schools they were around half a level higher in each subject. If a level is equivalent to two years’ progress, this means that, at KS3, borderline grammar-school pupils are around a year ahead of similar pupils in other schools.

“While it appears that selective systems obtain somewhat better overall results, the widely differing achievements of equally able pupils in different types of school give cause for concern,” say the researchers.

The challenge for comprehensives, they add, is to create a climate of high achievement and aspiration, so pupils of average ability perform as well as they would in grammars.

While the NFER’s national study focused on the academic impact of selection, the Slough research looked at other effects.

Interviews with the heads of 32 secondary and primary schools in the Berkshire town suggest that the 11-plus has a negative impact on primary schools, which are often under pressure to prepare children for the test.

Primary heads said that failing the 11-plus demoralised children, and the heads of non-selective secondaries said new pupils arrived feeling second-rate.

Borderline students at grammar school expect much better GCSE results and are three times as likely to plan to go to university as their peers in other schools.

“The impact of the structure of secondary education in Slough” and “The impact of selection on pupil performance” are at www.nfer.ac.uk

KEY FINDINGS

Brightest pupils perform as well in comprehensives as in grammar schools

Selection affects middle ability pupils most

At KS3, grammar pupils of average ability are around a year head of comparable pupils in comprehensives

Grammar school pupils are up to 20 times more likely to be entered for higher tiers at KS3 than similar pupils in comprehensives

11-plus has a negative impact on primary schools

Children who fail the 11-plus feel second rate