Chapters 3 to 5 have shown that different social and economic groups have different average learning outcomes, even from an early age. It follows that schools taking markedly different types of students will tend to get different average results. Despite this, too many commentators, policy-makers and even some academics mistake rawscore educational outcomes for evidence of differential school and teacher effectiveness. A simple example, or what should be a simple example, is selective grammar schools in England (or similar schools across the world). Where schools select the most able or highestattaining students at age 11, and these young people then obtain good qualifications at age 16 or high levels of competitive entry to university, this shows that the schools have selected well. It does not, in itself, mean that the schools have been particularly effective in enhancing learning. These high-ability students might have done as well or better at non-grammar schools.

Since the 1990s I have been investigating and writing about the supposed differential effectiveness of schools, school types and teachers, in terms of attainment outcomes. One theme has been responding to claims by others that schools of a specific type are especially effective in producing attainment outcomes. These claims have included the supposed superiority of Welsh-medium schools in Wales (Gorard, 1998c), specialist schools in England (Gorard and Taylor, 2001a), academies (Gorard, 2005b, 2009b), and grammar schools and private schools in both Wales and Pakistan (Siddiqui and Gorard, 2017).

In general, these claims for superiority do not stand up to even superficial scrutiny because the nature of the student intake to each type of school differs by so much. It is very like the comparisons between countries and regions in Chapter 5. Once differences in pupil prior attainment or in characteristics such as chromic poverty are taken into account, the surface differences in attainment are usually adequately explained. Put another way, no type of school within the national school system has been found to be more effective than any other with equivalent students. This chapter explains why.

Progress and contextualised scores

There is a number of valid possible reasons for wanting to judge the performance of schools. In most developed countries, the majority of schools are publicly funded, and so the custodians of public money want to assess how well that money is being used.