Should you group students according to academic ability? It is a question that can split schools down the middle in more ways than one.

Politicians of all parties consistently call for an increase in setting and streaming. Ofsted appears to support it, parents are supposed to be impressed by it and teachers often find it much easier than taking mixed-ability classes.

But a long line of academic studies have cast doubt on its effectiveness as a tactic for raising achievement among all students.

A serious weight of research now suggests that the effect of ability groups is at best limited and at worst damages the prospects of the very children most in need of an educational boost.

Yet the separation of classes along the lines of academic ability has been making a comeback in England’s state schools.

Some secondaries have gone as far as reintroducing that most controversial and rigid form of the policy: streaming. And some school leaders have even made a virtue of it and created “grammar school streams” within their comprehensives, complete with their own special uniforms and dedicated staff.

In and out of favour

The internal organisation of schools according to student ability is nothing new and was commonplace from the early 20th century onwards. It declined in popularity during the 1960s when mixed-ability teaching became fashionable but made a strong comeback under the Conservative governments of the 1980s and early 1990s.

According to Ed Baines, a senior lecturer in the psychology of education at the University of London’s Institute of Education, by the early 1990s more than 80 per cent of secondaries used setting for at least two subjects at the end of key stage 3, and 63 per cent for at least four subjects.

When Labour took over towards the end of that decade it wanted the trend to continue. The new administration’s Excellence in Schools White Paper, published in 1997, stated that “setting should be the norm in secondary schools” and added that “in some cases it is worth setting in primary schools”.

The growth in setting continued. But it was not fast enough for the Tories. By 2007, the party, still in opposition, issued a Green Paper that argued that setting by ability was the “only” way to “ensure that each pupil should be given the opportunity to learn in accordance with their particular aptitude and ability, so that the brightest students continue to be stretched at the same time as students who might be struggling are given extra support”.

The paper complained that despite Labour’s public commitment to the policy, the proportion of all academic lessons set by ability had increased only 3 per cent between 1997 and 2006, from 37 to 40 per cent.

So much for the growth in setting. But what of its more extreme cousin - streaming? Here, schools do not focus on grouping students according to their abilities in particular subjects. Instead, they make a general assessment of their ability and then place them in a single stream that operates for all or most of the subjects they study.

This method has always been more controversial because of its rigidity. It allows no scope for the possibility that students might have differing ability levels in different subjects. It is also likely to be harder for students to move up a stream than it is for them to move up a set. Critics argue that streaming can be particularly damaging to students’ self-esteem.

Unsurprisingly it has been less popular than setting. According to Baines, estimates in the 1990s suggested that streaming was being used in just 11 per cent of schools. Now, though, there are signs that it is making a comeback.

Roger Leighton first decided to introduce streaming, or at least a version of it, to the Sydney Russell School in Dagenham in about 2003, roughly six years after he became head of the East London comprehensive.

He says it was a trip to Switzerland - where he saw secondaries dedicated to students with particular aptitudes andor ability levels - that inspired him to group classes by ability.

Leighton felt he had to change things because results at the Sydney Russell School - situated in the middle of one of Europe’s largest council estates - were at rock bottom. In 1996, just 10 per cent of students were achieving the GCSE benchmark and 30 per cent were leaving without any exam passes at any grade.

“We needed to do something radically different,” he says. “In a school with multiple disadvantages and low aspirations, what we needed was a clarity of structure that could serve the needs of all the students.”

The result was what Leighton describes as “banding” - essentially a more nuanced version of streaming.

In schools that operate pure streaming, students are likely to be put in the same level of class - decided according to student ability - for all lessons. But Leighton’s version gives the teachers scope to place students in different sets for different subjects within three broad ability bands or streams, known in the school as “pathways”.

Following ‘pathways’

A decade on, results have improved dramatically. Leighton regards banding as such a success that he is introducing it to Riverside School in Barking, East London, a temporary foundation trust school he opened in September 2012 (in consultation to become the Riverview Free School).

Over in the south of the capital, Platanos College went down a similar route last year when head Judette Tapper introduced a new Grammar School Pathway to the comprehensive.

Designed for those who “show academic ability and are developing skills as independent learners”, this top stream aims to prepare promising young people for progression to university. Students at the school, which until September was known as Stockwell Park High, have to sit a test in English, maths and science to determine their eligibility for the pathway.

Their progress is then monitored throughout KS3 using teacher assessment, to decide whether students should continue in the stream or drop out of it.

“The children love it and become more competitive because of it,” Tapper says.

In a different corner of South London, another comprehensive, competing with real grammar schools, decided to take streaming to its ultimate conclusion. Crown Woods College in Eltham does not simply group students by ability into different “pathways”; it puts them into different “schools” within the school.

Students sit cognitive ability tests at the start of Year 7 to determine which of these “small schools” they will be placed in. Each comes complete with a distinctive uniform and colour-coordinated buildings.

Those finishing in the top third go to Delamere and are instantly recognisable by their purple ties. Students of a lower academic ability go to either Ashdown, and wear red, or Sherwood, where they dress in blue. Students in each small school play in their own fenced-off outdoor areas and even eat their lunch at different times.

The school says that in its first four terms of operation this extreme form of streaming has improved exam results and behaviour. But it has caused a great deal of controversy - so much so that headteacher Michael Murphy declined to discuss it publicly with TES last year.

In January he did speak to the BBC but said he preferred that the word “streaming” not be used, because, he said, it was “a pejorative term that has all sorts of connotations”.

Anyone wondering why need only look at a chapter in a book published late last year in which Baines sums up the huge body of research into grouping by ability (see panel, page 7).

He says the evidence shows that, in general, grouping by ability lowers performance and exacerbates existing socio-economic inequalities.

Streaming can have positive effects for some gifted students but only if they are taught a specially enriched curriculum. And lower-ability students are likely to have poorer outcomes under the system, according to the research he discussed.

In September, the point was emphasised when the government-funded Education Endowment Fund reported on various options available to schools wanting to raise attainment among disadvantaged students. It concluded that flexible within-class grouping was preferable to streaming. “While there may be some benefits for higher-attaining students in particular subjects, these benefits are largely outweighed by the direct and indirect negative effects for mid-range and lower-performing learners,” the report says.

Inflexible ability grouping can damage the confidence of lower-ability students and is “a clear example of what not to do if you want low-income students to benefit, as they are more likely to be assigned to lower-attaining groups”.

A pragmatic choice

Why are schools opting for the system when so much evidence seems to point against it? One reason is that they are marketing themselves to parents who are unlikely to be aware of such evidence, and may well see grouping by ability as a good thing.

Secondaries, particularly those in London, operate in an increasingly competitive environment and if they do not attract enough students they can struggle to survive.

Crown Woods College is a case in point. As a comprehensive competing with selective grammars in nearby Bexley it found itself losing out in the fight for the brightest students.

Murphy says that introducing streaming helped him to turn that around. “I felt if we made explicit the provision for high-ability children, we would be able to attract those children and their parents who would rather not put them in to take the Bexley 11-plus, but would feel comfortable with the type of provision we’d make for them - and that’s entirely what’s happened,” he says. “I think it was Margaret Thatcher who said you can’t ignore the market, you have to respond to it... If you have a really hard-nosed view and want your school to succeed, this is what you have to do. We wouldn’t have attracted the students otherwise.”

Tapper, at Platanos College, admits that similar reasons lay behind the introduction of the Grammar School Pathway at her comprehensive and has been unapologetic about her belief in the grammar school brand. “It’s traditional,” she says. “It’s what many parents would think of when you say traditional. It is something that is rigorous, something that is stable.”

But it would be wrong to portray branding as the only factor contributing to a more favourable climate for streaming. Schools cannot afford to ignore the views of Ofsted, and last September chief inspector Sir Michael Wilshaw made some outspoken comments on the subject.

He warned that some students were actually being held back by “the curse of mixed-ability classes without mixed-ability teaching”.

“Where there are mixed-ability classes unless there is differentiated teaching... it doesn’t work,” he said, adding that effective differentiated teaching is “hugely difficult” to achieve.

“We will be very critical when we inspect schools, particularly in the secondary sector, if we see mixed ability without mixed-ability teaching,” Wilshaw argued. “It mustn’t just be an article of faith.”

In other words, the approach has to work in practice as well as theory. It was that kind of pragmatic thinking that lay behind Leighton’s switch to streaming.

When he first suggested it at the Sydney Russell School, he encountered “vociferous opposition” from some members of staff who said he was proposing “social engineering” that would prevent students of different abilities from mixing.

“My answer was to ask, ‘How many of your own personal friends are markedly different to you in their intellectual capabilities?’” he says. “The truth is we naturally tend to gravitate towards people of a similar intelligence. Secondly (I said), ‘Just look out in the playground, when students are no longer forced to be with certain people, and see who they choose to congregate with.’”

Leighton emphasises that for streaming or banding to work, lower-ability students must not be left with the weakest teachers. But he disputes the notion that, in practice, streaming - seen as controversial - differs in any profound way from setting by ability, which is commonplace.

“You look at schools that set and you will find that most kids tend to be, broadly speaking, in the same sets for everything,” he says. “This idea that you can be a genius at English and a complete no-no in science is actually not really true.

“Broadly speaking, if you come in above-average ability you are probably going to be in the top sets for everything. To me it is just a question of facing reality rather than pretending it doesn’t exist.”

Introducing changes

- Define the key decision-makers within your school and try and get them on board as soon as possible.

- Take them to visit schools where the system is working successfully so they can see the reality in operation.

- Be clear why you are doing it. This is not a quick fix to solve your A*-C grade issues. It is a method and a structure designed to benefit all students.

- Ensure you are clear on the practicalities: timetabling, ongoing assessment of student progress and organising moves from one band to another.

- Make sure your communication strategy is sound. Ensure that the people you need to convince - parents, governors, students, teachers and other staff - are clearly informed so that you can dispel some of the myths right away.

- Avoid any suggestion that anyone in the lower bands will get a raw deal with the weakest teachers and poorest resources. Guard against that both in reality and perception.

- Avoid any labelling or a culture of low self-esteem and maintain that as you go.

What the academic evidence says

Ed Baines, a lecturer at the University of London’s Institute of Education, summed up the research on grouping by ability and streaming in chapter of a book called Bad Education: debunking myths in education and found the following:

General grouping by ability:

The more schools group by ability, the lower student performance is overall.

Gains made by more able students are offset by the less able.

Grouping by ability exacerbates existing socio-economic inequalities between students.

Lower-ability groups can end up with “conceptually weak and fragmented” teaching, resulting in “slow and repetitive” and “tedious” lessons.

Activities in lower-ability groups rarely involve use of analytical skills or foster creativity or independence.

The faster pace in higher-ability groups can be “too fast”, making it difficult for students to develop a detailed, long-lasting understanding.

The most knowledgeable and experienced teachers tend to end up in higher-ability groups, with the least knowledgeable and experienced staff teaching lower-ability students most in need of good teaching.

Ability grouping can put too much pressure on high-ability students and create low expectations in others, causing them to switch off.

It can polarise student friendships into separate academic abilities and create an anti-learning culture in lower-ability groups, exacerbating classroom disruption.

Streaming

At primary and secondary level it makes no consistent difference to student outcomes.

The overall average effect is negligible.

But there can be positive effects for some gifted students if they are taught an enriched curriculum.

The effect is more marked when teaching is varied by ability - higher-ability students benefit but their lower-ability counterparts do “much worse”.