Key points are not available for this paper at this time.
In its crudest terms the question before us now is whether this move towards comprehensive schooling has worked. Can we say that the reform has been a success? Three studies have recently appeared which address this question-Gray, McPherson & Raffe's (1983) Reconstructions of Secondary Education, Steedman's (1983) Examination Results in Selective and Nonselective Schools, and Marks, Cox & Pomian-Srzednicki's (1983) Standards in English Schools. They reach different conclusions. In this paper we shall argue that the answer to the question 'has it worked?' can be 'yes' or 'no' depending on the selection of data and of statistical techniques. It is in this sense that selection undeniably makes a difference. The three studies use different sources of material and employ different statistical strategies. Even if they had used the same data, their selection of statistical methods might well have led to different answers nonetheless. This can happen not because of statistical incompetence but because there are inherent technical problems in the question being asked. We do not yet understand the sources of variation in schools' examination results (exams being the main criterion of success used in these studies). Until we do so, we are likely to confuse the effects of selective education with those of other social and educational processes. It would be foolish to make firm policy decisions based on our present state of knowledge. We argue, therefore, that the primary task at the moment should be to understand what makes one school 'better' than another. If we wish to improve educational standards, there is more mileage to be made from discovering how to improve the worst schools, be they grammar, comprehensive or secondary modern, than by discovering whether the move towards comprehensive schooling 'has worked'. This is because, on our reading of the evidence on examination results, the difference between the comprehensive and selective systems is not very great whereas that between the 'best' and 'worst' schools within each system is in comparison quite substantial.
Clifford et al. (Sun,) studied this question.