Robin Alexander of the U.K. is a reader of this blog. He is director of the Cambridge Primary Review, a major independent evaluation of primary education in the U.K. The Reviews has published 31 interim reports between 2007 and 2009. Its final report is Children, their World, their Education, which appeared in October 2009.
He has shared some of the Review’s conclusions with us. We will, I hope, hear more from him about the lessons learned from his thorough study of British education and its implications for those of us in the U.S.
Here are some extracts that he forwarded to me:
TESTING, INSPECTION, STANDARDS AND ACCOUNTABILITY IN ENGLISH PRIMARY EDUCATION
(NB: in England the primary phase covers the education of children aged 4/5-11)
Extracts from the final chapter (ʻConclusions and recommendationsʼ) of Alexander, R.J. (ed) (2010) Children, their World, their Education:
final report and recommendations of the Cambridge Primary Review, Routledge.
67. Test results are not the best source of data for the multiple functions they are currently expected to perform – measuring pupils’ attainment, school and teacher accountability and national monitoring. Despite government claims to the contrary, the use of aggregated test results as a basis for evaluating schools does not provide a fair picture, even when the disputed ‘contextual value-added’ scores are used. This high-stakes use of test results leads to practices that not only have negative impact on pupils but fail to provide valid information, being based on what can be assessed in time-limited written tests in at most three subjects. The use of the same data for national monitoring also means that we have extremely limited information, collected under stressful conditions, which provides little useful data about national levels of performance and even less about how to improve them. The aggregation of SAT results for monitoring national levels of performance fails to reflect achievements in the full range of the curriculum.
68. There is an urgent need for a thorough reform of the assessment system in England, going well beyond the May 2009 report of the Government’s DCSF ‘expert group’, to provide a coherent set of practices and procedures suiting the aims of education in the 21st century and to meet the needs for information about the performance of individual pupils, schools, local authorities and the system as a whole. At the heart of this should be the use of assessment to help learning, leading to the development of lifelong learners … Separate systems are also required for the external evaluation of schools and for monitoring national standards of performance.
69. No single assessment procedure, including statutory assessment, should be expected to perform both formative and summative functions.
73. The practice of publishing primary school performance tables (now known as primary school achievement and attainment tables) based on the results of statutory assessment in English and mathematics at the end of key stage 2 should be abandoned.
75. The official evidence on whether standards in primary education have improved or worsened is unsafe. At its heart are two areas of difficulty: the validity and reliability of the chosen measures and procedures; and the historical tendency to treat test scores in limited aspects of literacy and numeracy as proxies for standards in education as a whole.
76. At the national level, the assumption that aggregating individual pupils’ test results in only three subjects enables trends in attained standards to be identified is problematic. Although the statistics can be computed, their meaning in terms of changes in attainment are brought into question by the limited range of what is tested, by limitations in test technology and by the impact of using the results for high-stakes judgements. We are left with little sound information about whether pupils’ attained standards have changed.
77. Subject to these substantial caveats, analysis of national test scores and international achievement surveys appears to show that standards of tested attainment in primary education have been fairly stable over the short period that usuable data have been available, with some changes up or down. Pupils’ attitudes to their learning in the tested areas are generally positive (though, as is generally found internationally, it appears to decline as pupils approach the end of primary education). There have been modest improvements in primary mathematics standards, especially since 1995, though different datasets tell different stories. The international data from 2001 show high standards in reading among English pupils by comparison with those from other countries, though the more recent data (from 2006 onwards) suggest that the 2001 results may have been misleading. England appears to be above the international average but not exceptionally so. The international data also show considerable improvements in primary science by comparison with other countries, though there have been methodological reservations about the studies in question.
78. However, gains in reading skills may have been at the expense of pupils’ enjoyment of reading. Similarly, there is some evidence of an increase in test-induced stress among primary pupils, especially at key stage 2, and much firmer evidence of pressure on their teachers. The primary curriculum has narrowed in direct response to the perceived demands of the testing regime and the national strategies, to the extent that in many schools children’s statutory entitlement to a broad and balanced curriculum has been seriously compromised; yet the national strategies have had, of themselves, a less pronounced impact on reading standards than might have been expected from the level of investment. The historically wide gap between high and low attaining pupils in reading, mathematics and science has persisted: it is already evident at a very young age and widens as children move through the primary phase. There is no reliable evidence on national standards in areas of children’s learning outside those aspects of literacy, numeracy and science which have been tested, other than that in many schools such learning appears to have been compromised by the standards drive itself.
79. Schools acknowledge the importance of being held accountable for their work and accept the need for periodic inspection. Ofsted [England’s national schools inspectorate] produces useful annual reports on the condition of the system as a whole and surveys on particular issues, on many of which the Review has drawn to its considerable benefit. The collation of evidence from inspections can be used to provide a reasonably valid, if partial, assessment of the quality of English primary education nationally at a particular time, assuming that the Ofsted criteria and procedures are accepted. However, Ofsted’s school inspection procedures attract a good deal of criticism in relation to their validity, reliability and impact; and because of frequent changes to inspection criteria and procedures, allied to the subjective nature of the process, it is much more difficult to say with confidence whether the overall quality of primary education has improved, deteriorated or remained the same over time. The same difficulty attends Ofsted inspections of individual schools. Such judgements are compromised by the successive changes Ofsted has instituted in inspection criteria and methodology and by its employment of different teams from one inspection to the next of the same school. Temporal comparisons and claims about long-term trends based on Ofsted data are thus highly problematic.
80. Teachers and schools can and should have a greater role in the assessment of their pupils and in the evaluation of their provision for learning. In the case of pupil assessment, there is an overwhelming case for extending the range of aspects of attainment that are included in reporting attained standards and in identifying the standards to aim for. At present the pupil attainment data reflect only a small part of the curriculum and within that only aspects which are easily measured by written tests. Greater use of information that teachers can collect as part of their teaching can help learning and, suitably moderated, can provide information which is a better reflection of performance acoss the full range of the curriculum. Similarly there is a strong case for moderated school self-evaluation across the full range of provision. Such evaluation should help the school’s own improvement agenda and not simply be instituted to meet Ofsted requirements.
81. Current notions of ‘standards’ and ‘quality’ should be replaced by a more comprehensive framework which relates to the entirety of what a school does and how it performs. The Review’s proposed statement of aims for primary education might provide the overall criteria for progress and success, combined with appropriate indicators for each of the proposed new aims and curriculum domains. However, we warn against moving from indicators of what can fairly be observed and judged to so-called measures of what cannot in fact be measured.
85. We take it as axiomatic that in a public system of education teachers and schools should be fully accountable to parents, children, government and the electorate for what they do. We reject any suggestion that our proposals for the reform of assessment and inspection imply otherwise. For us, the issue is not whether schools should be accountable, but for what and by what means, and the evidence shows that current approaches are in certain critical respects unsatisfactory. By insisting on a concept of standards which extends across the full curriculum rather than part of it, we are strengthening rather than weakening school accountability. It is no less important that others involved in primary education, including central and local government, are fully accountable for their part in the process. When responsibilities are shared, accountability should be shared too in order that the precise cause of problems can be speedily and accurately diagnosed and appropriate remedial action can be taken. Governments and policy advisers have been too inclined to blame teachers and a mythical ‘educational establishment’ for problems which are as likely to have their roots in policy.