Yet England was still reluctant to embrace the concept of state education. The public (i.e. private) schools of the Victorian period offered a curriculum that 'paid little or no attention to the scientific and the technical', but they enjoyed 'enormous popularity and prestige with industrialists, entrepreneurs and technological innovators'. As long as these schools held on to their dominant role in the nation's life, 'there was little incentive for the new bourgeoisie to create state schools with different values' (Chitty 2007:11).

However, the campaign for mass education continued to grow and was given a boost by the passing of the 1867 Representation of the People Act, which doubled the number of people who were allowed to vote. When it finally became clear that the country needed an education system which catered for all, those who were horrified at the prospect of children of different classes being educated together sought to ensure that the school system itself was clearly divided on class lines. Thus three commissions produced reports, each relating to educational provision for a particular social class:

• the Clarendon Report (1864) and the 1868 Public Schools Act focused on the nine 'great' public schools for the upper class;

• the Taunton Report (1868) and the Act which followed it in 1869 dealt with separate institutions for the middle classes; and

• the Newcastle Report (1861) and the 1870 Elementary Education Act (the Forster Act) sought to provide a basic education for the working class.

Standards

Streaming and setting were unknown at this time. From 1862 children in elementary schools were allocated to classes on the basis of 'standards'. The standard was the level the average child was expected to have reached at the end of a year's work and most children were expected to progress at one standard a year. Schools were anxious to get children into Standard I as soon as possible after the age of six because they were financed under the system of 'payment by results' up to - but not beyond - Standard VI.

Payment by results ended in 1898, but the notion of standards regulating promotion through the school survived because the 1902 Education Act (which abolished the School Boards and established a rigid distinction between elementary and secondary education) introduced a system of junior scholarships under which 'free places' in grammar schools were offered to the most able children. The vast majority of children remained in elementary schools until they reached the statutory school leaving age, which had been set at 12 in 1899. They were encouraged to progress quickly through the standards so as to have a better chance in the scholarship examination at the age of ten or eleven. Unfortunately, 'accelerated promotion for some was counterbalanced by delayed promotion for others who repeated the same work year after year. Some left from Standard II or III. Practice had not made perfect' (Plowden 1967:789). The school leaving age was raised to 14 in 1921, though many working class children were forced to leave school at 12 to seek employment.

The Committee's final report (1933) dealt with Infant and nursery schools. It noted that classification in these schools was usually by age, a tradition which could be justified because 'differences in intelligence are not so wide as they become in later stages', but that many schools had experimented with vertical classification. 'Here each class contains children of all ages from five to seven or eight, each occupied with work appropriate to his powers, and they remain throughout the infant stage in the care of the same teacher.' There were advantages in this system for both teachers and children. It 'obviously calls for special gifts in the teacher, but it "works"' (Hadow 1933:99).

To sum up, then, the 1924 Hadow Report considered the applicability of psychological tests of intelligence in schools, decided that they could be useful in some circumstances but had reservations about their widespread use. The 1926 Report recommended that schooling for all children should be divided into two phases, primary and secondary, with the break at the age of 11. Written (and, where possible, oral) tests should be used to select children for different types of secondary education, with the use of psychological tests restricted to deciding borderline cases. The 1931 Report urged the creation of A, B and C streams in primary schools, especially the larger ones. And the 1933 Report supported the practice of most infant schools in allocating children to classes on the basis of age, but felt there was value in experiments involving vertical classification.

Many of Hadow's recommendations were implemented, though some - nationwide provision of primary schools, for example - took many years. Undoubtedly, two of the most significant outcomes were the introduction and widespread use of streaming within schools and a much greater emphasis on selection procedures at the age of eleven, both based on measurements of attainment or 'intelligence'.

There was an apparent contradiction in the committee's recommendations. On the one hand, Hadow promoted a 'progressive' child-centred and activity-based style of primary education. On the other, its recommendation that pupils should be selected for different types of secondary school resulted in primary education becoming 'a sorting, classifying, selective mechanism' which necessitated streaming as 'the basic form of internal school organisation for all primary schools large enough to form parallel classes in each age group' (Galton, Simon and Croll 1980:36).

How could the committee reconcile a 'child-centred' approach to education with streaming on the basis of intelligence? Galton, Simon and Croll have argued that child-centred approaches, particularly Froebelian, were 'based fundamentally on the notion that the child's inborn characteristics must be allowed to flower ... the school's role is to provide optimum conditions for such development'. Since, according to the dominant school of psychology in the inter-war years (psychometry), a child's 'intelligence' was fixed, inborn, and not subject to change, 'what was necessary was to provide an education appropriate to the child's inborn, and measurable, intelligence level' (Galton, Simon and Croll 1980:37).

The Spens Report (1938)

The theory which underpinned the Hadow Reports also informed the Spens Report on Secondary education with special reference to grammar schools and technical high schools. This recommended that there should be three types of secondary school - modern schools, grammar schools and technical high schools (Spens 1938:22) - for which children would be selected at the age of 11.

The committee did argue that 'many benefits might accrue if children above the age of 11 were educated together in multilateral [i.e. comprehensive] schools, since the transfer of pupils at various ages to courses of teaching appropriate for their abilities and interests would be facilitated, and children differing in background and objective would be working in close association within the same school'. But they decided they could not advocate the adoption of multilateralism as a general policy in England and Wales because it would be 'too subversive a change' (Spens 1938:1). They did, however, support the establishment of multilateral schools, 'especially in areas of new population' (Spens 1938:130). They were right to think that the notion of children from different backgrounds working together was a subversive one, but it's a pity they didn't pursue the idea: it might have changed the whole nature of English education in the post-war years.

The Norwood Report (1943)

The Norwood Report on Curriculum and examinations in secondary schools endorsed the Spens committee's view that there were 'three broad groups of pupils', i.e. the academic, the technical and the practical. 'Accordingly we would advocate that there should be three types of education, which we think of as the secondary grammar, the secondary technical, the secondary modem, that each type should have such parity as amenities and conditions can bestow' (Norwood 1943:14).

The 1944 Education Act

Regrettably, these two reports, based on the still unquestioned belief that a test at the age of eleven could accurately predict what a child might go on to achieve, underpinned the divided education system which emerged in the wake of the 1944 Education Act. Few technical schools were ever built, so the vast majority of children either 'passed' the eleven plus and gained access to the grammar schools; or 'failed' and went to one of the new secondary moderns, which many saw as 'merely the old elementary schools writ large' (Chitty 2007:20). Thus England's post-war system of education was based on old ideas. The new primary schools streamed their pupils because that made it easier to prepare the more able children to pass the eleven plus, and the new secondary schools divided children into the academic (about a fifth) and the less able (about four fifths). And even within these secondary schools, the pupils were streamed. There was now 'education for all' but it was an education based on division and segregation at every level.

Facilitating the selection of pupils for secondary education had always been part of the role of elementary schools. Before the 1944 Act roughly ten per cent of pupils were selected for secondary schools, while the rest remained in the elementary system either in the same school (if it was 'all-age') or in the senior school. After the 1944 Act, the increased importance of the 11 plus examination put greater pressure on the new primary schools to stream their pupils, since a school's success was now measured largely by the number of its pupils who gained grammar school places. As Galton, Simon and Croll put it:

'It is difficult now to reconstruct the intense pressure on schools and teachers that built up in the 1940s and 1950s relating to the selection examination; the league tables that parents drew up for local schools, the telephoning round to find out who had done well and the sense of failure that some teachers experienced when their pupils won fewer places than others, or than expected; not to speak of the effects on the children.' (Galton, Simon and Croll 1980:37-8)

Unfortunately, the Certificate of Secondary Education (CSE), introduced in England and Wales in 1965, created even greater pressure on schools to divide students between 'academic' and 'non-academic' streams. 'Within comprehensives, GCE students were placed in different teaching groups from CSE students, while in the secondary modern school ... students who were deemed capable of CSE entry were separated from those who were not' (Jones 2003:84-85).

Peter Mauger, then head teacher of Nightingale County Secondary School in the London Borough of Redbridge, argued that non-streaming, though clearly important, was not an end in itself. What was also needed was 'a fundamental examination of the content of the secondary school curriculum' (Mauger 1966, quoted in Benn and Chitty 1996:259).

Not everyone agreed, of course. In a speech attacking Labour MPs for their support of comprehensivisation in January 1965, Conservative MP Quintin Hogg declared that secondary modern school pupils were happy 'banging metal and sawing wood'. These boys and girls were getting 'an education tailor-made to their desires, their bents and their requirements' (Hansard, House of Commons, Vol. 705, Cols 423-4, 21 January 1965, quoted in Chitty 2007:20-21).

The tensions for pupils, teachers and schools created by attempts to reconcile 'progressive' education (child-centredness, individualisation, 'discovery' methods etc) with selection and streaming were coming to a head. 'The post-war divided system, underpinned by spurious notions of "intelligence", was seen as having to be drastically modified or perhaps even abandoned altogether.' Parity of esteem between grammar and secondary modern schools was a sham and 'failure' in the eleven plus was 'invariably a cause of much distress both for a child and its parents' (Chitty 2007:91-2).

The Plowden Report (1967)

So the publication of the Plowden Report Children and their primary schools could not have come at a more appropriate moment. Across the country, local education authorities were now 'going comprehensive' and primary schools, freed from the constraints imposed by the need to 'get good results' in the eleven plus, were beginning to experiment with the curriculum, teaching methods and pupil grouping policies. Oxfordshire was one of the first counties to scrap the exam, along with Leicestershire, the West Riding of Yorkshire, Bristol and London.

'It was in these areas, also, that the system of streaming, which reinforced the methodology of class teaching, was most rapidly discarded. The swing from streaming in the junior schools in these and other areas, which started very slowly in the mid 1950s, meeting strong opposition, suddenly took off with extraordinary rapidity in the mid to late 1960s, gaining influential support from the Plowden Report of 1967.' (Galton, Simon and Croll 1980:39)

On the issue of pupil grouping, Plowden began by describing current practice.

It noted that some infant schools arranged their classes strictly by age, though it was common for children who were making progress in reading to be promoted first. In a minority of schools, children were graded by attainment at the age of six; in a few grading began even earlier (Plowden 1967:793). Vertical classification, which had been seen as early as the mid 1930s, remained 'a minority practice' but the two or three areas in which it was prevalent were among those which had 'the most lively infant school work in the country' (Plowden 1967:794).

With regard to junior schools, the National Survey conducted for Plowden found that a majority of the schools took into account attainment as well as age in allocating pupils to classes. The much larger NFER survey put the proportion at 56 per cent 'for certain' with another 14 per cent 'probables'. In two-form entry schools each year group was normally divided into an 'A' stream and a 'B' stream (Plowden 1967:795).

In making class allocations, most head teachers relied on information from the infant schools: less than half the junior schools in the NFER survey used a standardised intelligence test. Other methods used, in decreasing order of popularity, were the head's own judgement, the results of internal school examinations, a child's position within the age group, and standardised attainment tests. In the later years of the junior school, increasing reliance was placed on school examinations, teachers' judgements and standardised attainment tests. 'Most schools would claim that they group children both on attainment and on ability' (Plowden 1967:796).

And what of mixed ability grouping? (Incidentally, Plowden rarely used the term 'mixed ability', preferring 'unstreamed'.) It noted that about a fifth of the children in schools included in the NFER survey were allocated to classes which were intended to include the whole range of ability in the school. Some schools tried to ensure classes were exactly parallel in ability by relying on children's records, test results etc. Others allocated children to classes at random, according to where they lived, or alphabetically by surnames.

Having described current practice, the report went on to discuss the various criteria for classification.

Age. Plowden was against moving children out of their own age group. It noted that research in European countries where promotion from grade to grade depended on achieving a specified standard showed that failure tended to be cumulative and that pupils who repeated a grade on one occasion, far from catching up with their contemporaries, were more likely to have to do so a second or even a third time. It concluded that 'children are better with their friends in their own age group, unless there is clear evidence to the contrary' (Plowden 1967:798).

Vertical classification. The report noted some disadvantages of vertical grouping in the infant school. 'For children to spend the whole of their infant school life with one or at most two teachers may be to distribute the strengths and weaknesses of the staff unfairly. It may also limit children's contacts with adults to too narrow a circle and some children may be too sheltered or lack stimulus' (Plowden 1967:801). Younger children might be 'overshadowed by the older ones, may imitate them too closely, and have insufficient experience of the kind of play that is an investigation rather than a use of materials' (Plowden 1967:802). However, the committee agreed that vertical classification could have advantages and they were 'impressed by the liveliness and good quality of the work in infant schools where classes extend over two or three age groups' (Plowden 1967:804).

Classes including less than a year group. Plowden was not generally in favour of sub-dividing year groups by age, warning that teachers needed to guard against regarding a younger class as a less able class (Plowden 1967:805).

Classification by attainment or ability (streaming). Streaming was still 'by far the most common way of organising junior schools' (Plowden 1967:806), but there was evidence that the picture was changing. In just two years - between 1962 and 1964 - the proportion of primary teachers who favoured streaming had fallen from 85 to 30 per cent (Plowden 1967:807). Teachers were ahead of the public in this respect: two thirds of parents still preferred their children to be taught in streamed classes (Plowden 1967:808).

Before the second world war, streaming had been seen as 'a device for opening the grammar schools to talented working class pupils'. But since the war, selection for secondary education had been challenged on three grounds: 'the accuracy of the selective process, the contrast in the provision made for children of differing ability and the effect of segregating them on their achievement' (Plowden 1967:809).

Similar arguments, said the report, could be made against streaming.

First, streaming involved selecting, and in schools which were streamed throughout this effectively meant that children were being selected at the age of seven. 'We know of no satisfactory method of assigning seven year old children, still less those who are even younger, to classes graded by attainment or ability.' It noted that between ten and twenty per cent of the predictions made at 11 were subsequently proved wrong and it warned that the earlier such predictions were made, 'the less relevant they are bound to be to an education which will continue to 15 or later' (Plowden 1967:811).

Because it was inevitable that many children would be wrongly placed at seven, ease of transfer between classes was important. It would be reasonable to expect ten per cent of pupils to be transferred each year, but the NFER enquiry had shown that the actual figure ranged from 2.3 per cent to six per cent over three years (Plowden 1967:812). Plowden argued that this was probably because, once in a class, children adapted themselves 'to the rate of progress or work which the teacher expects' (Plowden 1967:812).

The NFER enquiry had also shown that streaming favoured girls and older children within the year group (Plowden 1967:814) and that it served as a means of social selection. 'It is not simply that middle class pupils congregate in upper streams and the children of semi-skilled and unskilled workers in lower streams. That might be expected from the association of intelligence with social class and occupation. Evidence is also available that more middle class children are to be found in upper streams and fewer in lower streams than would be expected from their results in objective tests' (Plowden 1967:815).

The fact that selection was inevitably inaccurate would not matter too much 'if the conditions for upper and lower streams were equally good'. However, the NFER research suggested that upper streams were mostly taught by older and more experienced teachers. 'Teachers may be streamed, no less than pupils. The more established the teacher the more probable it is that he will get one of the better classrooms and a generous supply of books and equipment.' The NFER survey even showed that 'a higher proportion of the lowest streams were in classrooms which faced north' (Plowden 1967:816).

Plowden reviewed all the available evidence of the effect of streaming on children's achievement and attitudes, including the 1959 research by Yates and Pidgeon, the Manchester Survey of 1964, and its own National Survey, which included the NFER cross-sectional study of attainment in matched streamed and unstreamed schools. It noted that both the Manchester Survey and the NFER study showed that 'by ten the lead of children in streamed schools had been reduced in all tests and there was no significant difference in reading'. There was other evidence which suggested that 'children who are taught by informal methods make a slower beginning and catch up towards the end of the primary school' (Plowden 1967:817).

There was some evidence that achievement 'in the limited field of measurable attainment' was higher in streamed schools, but it was 'not so marked as to be decisive', and Plowden's view was that 'forms of organisation are less critical than the underlying differences in teachers' attitudes and practice which are sometimes associated with them' (Plowden 1967:818). Organisation was, nonetheless, important because it could 'reflect and reinforce attitudes'. Where unstreaming was 'established with conviction' and 'put into effect with skill', it produced 'a happy school and an atmosphere conducive to learning'. 'We welcome unstreaming in the infant or first school and hope that it will continue to spread through the age groups of the junior or middle schools' (Plowden 1967:819).

If schools were determined to retain streaming, it was essential that staff were aware that 'any classification is bound to be faulty' and that there would be big differences between individuals in a class which would increase as the children grew older. If all the children in a class were working at the same pace and the same level, 'they are probably conforming to what their teacher expects of them' (Plowden 1967:821). The distinction between streams should not be obvious: the fourth year 'A' class should cease to be the preserve of one or two teachers. This might be one way of 'speeding the disappearance of exercises which serve little purpose save to prepare for an examination which is itself now disappearing' (Plowden 1967:822). Finally, the report warned that 'Streaming can be wounding to children. Great care ought to be taken not to suggest that trust and responsibility, or prowess in games, or the ability to look after library books are the preserves of certain classes. No more certain way could be found of alienating children from school or of creating irresponsibility' (Plowden 1967:823).

Within the class, Plowden commended individual and group work. Groups should be based 'sometimes on interest and sometimes on achievement, but they should change in accordance with the children's needs'. There were still too many stories of children streamed by the tables they sat at, of 'top tables' and 'backward reader' tables (Plowden 1967:824). Group work fostered social skills. 'Children learn to get along together, help one another and realise their own strengths and weaknesses as well as those of others'. It also fulfilled a pedagogic function: children 'make their meaning clearer to themselves by having to explain it to others, and gain some opportunity to teach as well as to learn' (Plowden 1967:757).

What Plowden proposed, then, was a focus on the needs of individual children through a combination of individual, group and class work. And this seems to have been exactly the system many primary schools chose in the years following Plowden's publication. Class teaching was mostly rejected; complete individualisation based on workcards and assignments was fairly widely implemented. But the most popular option 'comprised individualisation together with the use of grouping within the class'. This provided 'a rational means of controlling (or managing) the independent activities of some thirty plus children' (Galton, Simon and Croll 1980:55).

Research by Rosenthal and Jacobsen, published in 1968, backed up Plowden's views on streaming by revealing the extent to which performance depended on environment. 'Teachers were told that one group had scored high on IQ tests, and another low, when in fact both groups were randomly assigned. But the bright group got brighter, and the slow group slower' (Holt 1978:165).

Plowden's views were timely. Indeed, 'the period from 1965 to 1970 was one of intense interest in the whole question of pupil grouping' (Benn and Chitty 1996:250). Unstreaming was 'very much a teachers' movement, in its inception, at least' (Galton, Simon and Croll 1980:54). There were well-attended conferences and workshops on this issue, and articles in Forum and Comprehensive Education. By the beginning of the 1970s it was 'difficult to find a primary school that grouped its pupils into classes on the basis of ability - at least in any overt manner' (Kelly 1978:2).

HMI Survey of Primary Education (1978)

By 1978, the HMI Survey of Primary education in England could report that 'in very few schools were classes streamed according to ability' (DES 1978b:3.3). Within the mixed ability classes, some vertically grouped, teachers employed a range of methods. Most teachers grouped children in various ways for some of their work. These groups 'varied according to the subject being taught or the kind of activity being undertaken and were usually formed or reformed for particular purposes or according to the needs of the moment' (DES 1978b:3.5).

Almost three quarters of the classes were grouped by ability for their work in mathematics, between a half and two thirds were grouped by ability for reading and, in fewer classes, for writing. 'Children were rarely grouped by ability for their work in any other area of the curriculum.' Individual work assignments were used in 'a considerable majority of classes for the teaching of reading, writing and mathematics'. (HMI 1978b:3.7)

The Survey was unable to compare the attainments of 11 year olds in streamed and mixed ability classes because the number of the former was too small. It did note, however, that 'when the NFER scores in reading and mathematics for the streamed classes were omitted from the calculations there was virtually no change in the average scores'. (HMI 1978b:7.22)

Mixed ability work in comprehensive schools (1978)

Meanwhile, another investigation into the use of mixed ability groupings in comprehensive schools in England carried out by HMI between 1975 and 1978 found that just over a third of schools used mixed ability groupings for most subjects in the first year, almost a quarter in the second year, and more than one in ten in the first three years. Only two per cent of schools taught mixed ability groups in all five years (DES 1978a, quoted in Benn and Chitty 1996:253).

Maurice Holt commented that 'to many educationists the streamed comprehensive school is a contradiction in terms.' He advocated the use of a range of teaching strategies and lamented the fact that 'most teachers have been trained, and are still being trained, to use teaching styles which work best with streamed groups' (Holt 1978:164).

Kelly: A rationale for mixed ability teaching (1978)

With mixed ability teaching now widespread, AV Kelly sought to offer a rationale for its use. In his 1978 book Mixed ability grouping, he argued that the production of such a rationale was bedevilled by problems of definition, the variety of practice in schools and the methods used to create mixed ability classes. It should be approached from two directions - 'from a consideration of the criticisms that have been levelled at the system of streaming ... and from more positive arguments for replacing that system with mixed ability classes' (Kelly 1978:6). The arguments against streaming had been dominant in most recent discussions and had led to 'something called "unstreaming" rather than mixed ability grouping' (Kelly 1978:6).

The first negative argument was that the assumptions underlying streaming had largely been discredited. Few now accepted 'the rather simplistic notion of intelligence which was clearly fundamental to the introduction of streaming' (Kelly 1978:8), or the concept of 'general ability' which proponents of IQ tests had argued would have significance for all kinds of intellectual activity. Streaming assumed an unsophisticated model of teaching in which the teacher taught the class 'the same material in the same way at the same pace and at the same time' (Kelly 1978:9-10). When more effective teaching methods were adopted, there was 'less need to attempt the impossible task of achieving intellectual homogeneity in our teaching groups' (Kelly 1978:10).

The second negative argument was that streaming had been a contributory factor in the wastage of talent which had been revealed by research preceding the publication of the Gurney-Dixon Report on Early leaving in 1954 and the Crowther Report 15-18 in 1959. It had had damaging effects on the progress of individual pupils and on the morale of teachers. There was no evidence that able pupils achieved more in streamed classes or less in mixed ability classes, while the less able did better when when working alongside pupils of all abilities. What was clear, however, was that streaming impacted adversely on the social development of children. Pupils allocated to lower streams appeared to lose 'all sense of identification with the aims and purposes of the school'. The result was that 'a delinquescent subculture' emerged, the prime goal of which was 'to reject and even to subvert the values of the school rather than to embrace them'. Unstreaming, on the other hand, had positive effects. 'Every study undertaken has revealed the same thing' (Kelly 1978:15).

Finally, Kelly argued that the development of children's ability to think divergently was more likely to happen in 'the less formal atmosphere of non-streamed schools than in the more formal atmosphere of those that stream' (Kelly 1978:16-17).

Kelly's positive rationale for mixed ability grouping had three strands. First, it was based on a view of people as cooperative social beings and of society as egalitarian. Second, it was the result of major changes in 'the view we take of what education itself is'. And third, it was 'a consequence of some major questions that are being asked about the nature of values themselves and the basis for making judgements of value about anything, especially about different kinds of knowledge, and a resultant desire to avoid dogmatism in the planning of a curriculum and to endeavour to make it meaningful to the individual child in the light of his own needs and interests' (Kelly 1978:22-23).

Continuing technological and social change required the education system itself to be flexible. Streaming would not do because 'one of its major characteristics is its total inflexibility' (Kelly 1978:23), whereas 'the purposes of a school are multifarious and many different kinds of grouping are needed to meet them' (Kelly 1978:24). This was, in Kelly's view, the strongest case for mixed ability grouping - that it could and should be 'flexible enough to allow for the creation of different groupings for different purposes and to facilitate continuing development of all kinds' (Kelly 1978:24-25).

Attacks on progressive education

While the period from 1968 to 1979 may be regarded as the heyday of mixed ability teaching - and of 'progressive education' in general - it also saw a series of increasingly vitriolic attacks by traditionalists. An economic recession resulted in cutbacks in education spending and a 'general disenchantment with education as a palliative of society's ills' took hold (Galton, Simon and Croll 1980:41); a series of 'Black Papers' argued for a return to traditional teaching methods; the 'William Tyndale affair' gave the traditionalists ammunition; Prime Minister Jim Callaghan gave his Ruskin College speech, initiating a 'Great Debate' about the nature and purposes of education; the Assessment of Performance Unit was established; and local authorities began the mass testing of pupils.

On top of all this, eugenicism was still alive and well in some quarters. The second of the Black Papers, The crisis in education (1969), for example, contained articles by Cyril Burt, Hans Eysenck and Richard Lynn.

In his piece The mental differences between children Burt reiterated Francis Galton's view that such differences were wholly or largely inherited and even argued that the Wilson government's 'progressive' reforms had resulted in standards being lower than they had been before the First World War (Burt 1969:23, quoted in Chitty 2007:94).

Eysenck, Professor of Psychology at London University, contributed The rise of the mediocracy, in which he argued that the abolition of eleven plus selection would deny working-class children the possibility of social advancement. It would result in the promotion in society of 'a large number of people of mediocre ability, while keeping submerged many people of superior ability' (Eysenck 1969:40, quoted in Chitty 2007:95).

The most outrageous of the three articles was that by Richard Lynn, Research Professor of Psychology at Dublin University. In Comprehensives and equality he talked of the 'lies' perpetrated by 'radical progressives' (whom he referred to as 'young red guards'). One of these lies was that it was the fault of society that slum dwellers were impoverished and their children did badly in school. 'They do not realise', he declared, 'that slum dwellers are caused principally by low innate intelligence and poor family upbringing' (Lynn 1969:30, quoted in Chitty 2007:95).

The Primary Review had been established in 2006 to investigate 'the condition and future of primary education in England'. Supported by the Esmée Fairbairn Foundation and independent of government, the Review was based at the University of Cambridge and led by Professor Robin Alexander.

In May 2008 it published Research Survey 9/2 Classes, groups and transitions: structures for teaching and learning. This collated the findings of dozens of research projects 'to explore different school and class grouping arrangements, the factors influencing them, and their impact on pupil learning and adjustment in the primary phase'. It distinguished between grouping at the class and within-class levels, and argued that 'the latter is likely to be more important for pupils' educational attainments, behaviour and attitudes to schooling' (Blatchford et al 2008:1).

The authors began by noting that in the years following the 1988 Education Reform Act educational reform in primary schools was primarily concerned with curriculum and assessment arrangements. They argued that more emphasis needed to be given to the social contexts - classrooms and groups - because of the effect they had on teaching and learning.

The survey was extremely comprehensive. Among the research reports considered were those by Bealing (1972), DES (1978), Lee and Croll (1995), Hallam et al (2003), Kutnick et al (2006) and Ofsted (1998). Evidence from primary school inspections was taken into account, as were the descriptive studies of primary schools and classrooms by Galton, Simon and Croll (1980), Bennett et al (1984), Tizard et al (1988), Mortimore et al (1988), Alexander (1995, 1997 and 2001) and Galton and Hargreaves et al (1999). The survey also examined some longitudinal studies of ability grouping - Barker Lunn (1970), Goldberg et al (1966), Borg (1965), Tizard et al (1988) and Mortimore et al (1988); studies involving classroom mapping - Alexander (1995 and 2001), Mortimore, Sammons, Stoll and Ecob (1988), Alexander (1997) and Kutnick, Blatchford and Baines (2002); and research based on classroom observation - Galton, Simon and Croll (1980) and Alexander (1997).

The survey faced three methodological challenges. First, there were many school variables: the number on roll, intake, staffing, the number of classes, and grouping patterns such as streaming, setting, vertical, mixed ability and within-class, all of which might change for different curriculum areas. Second, there were a number of educational outcomes to be considered - some relating to children's personal and social development, some to academic achievement. And third, the effects of grouping practices were 'not consistent in size, over time, in different subject domains or between teachers' (Blatchford et al 2008:5-6).

The authors first considered between-class grouping policies: streaming, setting and mixed-ability teaching. They noted that the 'heated debates' on this issue over many years had sometimes been 'unhelpfully polemical' and argued that 'the reality is more complex and less clear'. They aimed to achieve a 'balanced account of what we know about grouping at this level', seeking to explain why schools adopted particular forms of grouping, and the impact of these on teaching and on pupils' learning and attainment (Blatchford et al 2008:1). They concluded that grouping classes by ability had 'no positive effects on attainment but has detrimental affects on the social and personal outcomes for some children'. The allocation of pupils to such groups was 'a somewhat arbitrary affair' often dependent on 'factors not related to attainment'. And, while movement between groups was theoretically possible, in practice it was 'frequently restricted', limiting the opportunities for some children (Blatchford et al 2008:28).

On grouping within classes, the authors began by noting that 'one of the few features of educational life that can be stated with certainty is that all pupils are grouped within classrooms'. Such grouping included the whole class working together, small groups, larger groups, pairs of children, and individuals working alone. They recommended grouping pupils in different ways for different activities. This 'offers more flexibility, facilitates movement between groups structured by ability, and avoids limiting the opportunities for some children'. Teachers needed to tailor work more specifically to pupil needs and to be aware that pupils' attainment levels 'do not follow a stable trajectory': groupings should be 'constantly reviewed to take account of this' and should vary according to the nature of the task. This would avoid children 'labelling themselves as being in one specific group'. There should be more research on how such flexibility could best be developed and how it impacted on pupils and teachers (Blatchford et al 2008:28).

In their conclusion, the authors argued that 'Over the long history of research into school structure and classroom grouping, there has been little transfer between research findings and widespread classroom application'. Concerns about underachievement, poor attitudes and exclusion had often been met with calls for more differentiation by ability or attainment. The review made clear that 'such moves are not supported in the research literature'. In fact, differentiation by ability or attainment had resulted in 'limited access to knowledge by some pupils, domination of pedagogic practices by teachers, preferred teachers for "elite" pupils and enforcement of social divisions among pupils' (Blatchford et al 2008:30). The development of 'classroom-based social pedagogy' (including the effective use of pupil groupings) must now become a priority. When teachers were committed to developing 'relational and other social pedagogic practices' within their classrooms, 'pupils respond with improved attainment, classroom behaviours and pro-learning attitudes' (Blatchford et al 2008:31).

Social mobility (2008)

November 2008 saw the publication of two reports on social mobility. Getting on, getting ahead, published by the Cabinet Office's Strategy Unit, claimed that government initiatives in the past ten years had at last begun to break down social class barriers in education but it acknowledged that bright children from poorer homes were still likely to be overtaken by less able children from middle-class families and that attainment gaps which opened up early in a child's life were much harder to close later on.

A week later, Manchester University's School of Education published Successful leadership for promoting the achievement of white working class pupils. Denis Mongon and Chris Chapman focused their research on pupils from a white British ethnic background who qualified for free school meals. They found that family income and status were 'by far the most significant correlates of success in the school system' (Crace 2008).

In other words, the relationship between underachievement and social class was virtually unchanged since the late 1800s. Mongon said it was important for teachers to understand that this was not a causal relationship. 'We have to get rid of the idea there is a defining underachieving stereotype', he said.

Which is, I think, where we began.

Postscript

I was intrigued to see the headline 'Complex instruction teaching technique puts 'streaming' to the test' in The Times. (Alexandra Frean, 27 September 2008).

Apparently, 'complex instruction' was devised by Californian academics and is being pioneered in Britain by Professor Jo Boaler of Sussex University. She followed the progress of 700 American students for four years, at the end of which 41 per cent of the students who had been taught by the system in mixed ability groups were in advanced classes for calculus compared with 27 per cent of pupils in the schools which had used setting and traditional teaching methods. They were also better behaved and enjoyed maths more.

'Complex instruction' involves dividing pupils into groups of four with each member having a distinct role. The groups are given tasks to solve together and at the end of the lesson one member of each group has to present their findings to the class. As they do not know in advance which member of the group it will be, they have to ensure that all are capable of doing so.

I was intrigued because 'complex instruction' sounds exactly the same as 'collaborative learning', a scheme which was being developed by ILEA in the mid 1980s (before Thatcher consigned that progressive local authority to the dustbin of history). There's nothing new under the sun ...