Pupils who scrape into grammar schools end up with poorer GCSE results than their primary classmates who just miss out and attend secondary moderns, a new report shows.

Researchers say their findings, released exclusively to TES, illustrate problems with the 11-plus grammar school entrance test, and show that key stage 2 national curriculum tests are better predictors of GCSE success.

They conclude that 11-plus results are likely to have been skewed by private coaching. As a result, poorer children do not win grammar school places even though they have more academic potential than many classmates who do get in.

Rebecca Allen director of Education Datalab, the new research organisation that produced the report, told TES: “I don’t think we can fix the selection system at 11 so we have to get rid of it or we have to delay it further [to a later age]. Either of those things would mitigate the horrible consequences of the 11-plus.

“Maybe grammar schools will say ‘Fine - out of a class of 30, 25 will deserve to be there and that is good enough for us’. But I don’t think that is good enough for our children. If selection hasn’t been fair then that is a good enough reason not to have the selection.”

The study used government data from more 500 primaries whose children routinely go on to grammar schools. In each primary, the researchers compared the pupil with highest KS2 national test score who did not go on to a grammar, with the pupil with the lowest KS2 score who pass the 11-plus.

The secondary modern children were twice as likely to be eligible for free school meals. But they also ended up with slightly better GCSE results in terms of their overall points score than their former classmates who went to grammar schools. Their individual results in English, maths and their best science GCSEs were also better.

The National Grammar Schools Association said that without full details of exactly which subjects the GCSEs were taken in, it was impossible to know whether the study was comparing like with like, as grammar schools could be offering a more academic curriculum.

When all the pupils’ results were compared, the analysis showed that in many cases the KS2 test scores seemed to bare little relation to whether pupils got into grammar schools.

“It seems hard to believe, but there are children at grammar schools who only achieved a Level 3 [one national curriculum level below what is expected for 11-year-olds] at English or maths,” the report says.

“Equally, some children at non-selective state schools in grammar school areas have KS2 scores equal to the smartest pupils at grammar schools.”

The study also found that many of the weakest performing pupils who got into grammars were then labeled as having special educational needs. 

Speculating as to why their national test scores were so low, the researchers note that they “still did pretty badly at GCSE”.

“So we don’t think they just ‘threw’ the KS2 tests by writing silly answers or didn’t make an effort,” they say. “We can only guess they had some pretty intensive coaching for non-verbal and verbal reasoning tests.”

The report says its findings demonstrate “the shortcomings of the 11-plus exam across every highly selective local authority, regardless of whether they use verbal reasoning, non-verbal reasoning, numeracy, English or any other test paper”.

Dr Allen acknowledged that the pupils they looked at had taken the 11-plus in 2008 and that some areas had introduced changes since then, designed to make the test less susceptible to private tutoring. But she said: “I don’t believe anybody can devise a test where preparing for it doesn’t help.”

Robert McCartney, chairman of the National Grammar Schools Association, said: “No one has ever suggested that the 11-plus is a foolproof test; everyone accepts there an element of coaching.

“Something like computer adaptive testing could make it better. But the best thing that could happen would be to make more grammar school places available.”

But Ian Widdows, founder of the National Association of Secondary Moderns, said the research “punches a hole through the argument that grammar schools promote social mobility”.

“It also is almost unique in clearly demonstrating the excellent work done by secondary moderns, schools which are far too often judged unfairly as ‘second best’,” he added.

The report is one of seven being published this evening by Edcuation Datalab, a division of the FFT education charity set up to analyse the large amounts of education data now collected by government.

Related stories: 

More grammar schools ‘will increase social mobility’, Conservative campaigners say - December 5, 2014

Grammars not doing enough to attract poorer pupils, report finds - 8 November, 2013