It’s a saddening comment on our methods of financing research that a very important book about education should be essentially an after-thought to a public health survey. In his introduction to Dr J. W. B. Douglas’s book, The Home and the School,footnote1 Professor Glass says, ‘The investigation . . . was originally undertaken to examine the availability and effectiveness of the ante-natal and maternity services in Britain’. For this purpose over 5,000 families were selected, in fact every family which had a child born in one week in 1946. ‘It had not originally been intended to continue the research beyond the 1946 study,’ but as Professor Glass rightly observes, ‘a superb opportunity would have been missed’.

Such are our preoccupations now that the third book about these children deals not with their respiratory tracts but rather with their ‘abilities’ and ‘achievements’ at school and their selection for secondary education. The book is as Professor Glass states, frankly ‘obsessed with the underprivileged’, i.e. that majority of children whom ‘selection’ selects out. Alongside this obsession runs the simple equation that not only do able children need education but the country needs them. Unfortunately the equation doesn’t necessarily balance—the education children need isn’t in a superficial analysis the same as the country’s ‘scientific manpower reiquirements’ on which the provision of higher education is based. Still as Dr Douglas states, it is very important to understand what is happening when we talk of the ‘pool of ability’ at 18, and ‘to know how large the pool would be if there were no leakage of talent’ earlier.

To find this out ‘the teachers assessed the level of interest the parents showed in their children’s school progress’. ‘The headmasters described the location and amenities of the schools and the type of children who came as pupils. . .’ ‘Health visitors and school nurses . . . were the chief sources of information about the children’s families and home circumstances.’ While at school the children were assessed by tests of ability and achievement administered at 8 years 3 months and again just before the 11th birthday. This is the first thing to note about this book: it is a very thorough study of a limited area of the education system, primary schooling, but in that system’s own value terms and measurements. Such an attack is both a strength and a weakness. The results are a revelation of what can be achieved by persevering, inspired sorting of punched cards and the delicate weighing of statistical evidence. Anyone who doesn’t ‘believe in’ statistics must surely be convinced by those deployed here—though it’s significant that to get figures which really tell him something, a opposed to merely supporting hunches, Dr Douglas had to abandon x² and go on to ‘relatively sophisticated statistical techniques’ which require specialist analysis and a computer.

Straight away a truly damning fact emerges. The measurements of ‘ability’ and ‘achievement’ could not be separated: ‘with a few excepttions the circumstances which are associated with a deterioration in the scores for the school achievement tests are equally associated with a deterioration in the mental ability scores.’ So much for selection ‘by ability’. Measured ‘ability’ itself is demonstrated to be dependent on social class, on housing conditions, family size, teaching, parental encouragement; in other words on environment in its widest sense. ‘Able’ working-class children (those who scored highly in the eight year tests) suffered a deterioration of score up to the age of eleven if their environment was unfavourable in any of the above respects; so that by the time they faced selection, tests would indicate that there were fewer suitable for grammar school education. Middle-class children suffered not at all, or less, from these disadvantages so that their scores improved relatively or absolutely. Further, even on the strength of this measured ability at eleven the working-class children didn’t get the proportion of places to which ability alone would entitle them. The very few late transfers are also largely middle-class children.

The analysis is broken down further to isolate individual causes. Working-class children do worst when provision of grammar school places is low because they are the first to suffer. Yet ‘regional differences in the provision of grammar school places in no way reflect real differences in the abilities of the child en or in the availability of private schools’. Such regional differences account for the able child’s better chances of selection if he lives on a council estate. But overall, ‘when housing conditions are unsatisfactory, children make relatively low scores in the tests. . . For (working-class children) over-crowding and other deficiencies at home have a progressive and depressive influence on their test performance’. Streaming emerges as a major cause of waste of ability. It makes ‘bright’ (relatively more often middle-class) A stream children grow brighter, and ‘dull’ (more often working-class) B stream children relatively ‘duller’. The B streams contain ‘those who during the first six years of their lives were said to be dirty, badly clothed and shod’, and children from large families. Dr Douglas notes that teachers would have many more girls than boys go to grammar schools, and suggests a failure of sympathy on the part of a predominantly female teaching staff. He can state with confidence that the better success of early-maturing girls is not a result of this physical factor so much as the constellation of advantages which a large proportion of these children enjoy as only children. So refined are his statistical techniques that they can lead him towards oversubtlety: I pondered for a long time over the discovery that ‘parents influence the test performance of their children between the ages of eight and eleven only to the extent that they succeed or fail to encourage them to work hard’.