Improving assessment of learning and for learning

Abstract

The focus of this article is recent work by the Assessment Reform Group (ARG) on the role of teachers' judgements in the summative use of assessment. A brief overview of the early work of the ARG is followed by discussion of the desirable properties of assessment for summative uses. The work of the ARG's Assessment Systems for the Future project provided evidence and arguments concerning the validity, reliability, impact and cost of tests and of summative assessment by teachers. Whilst there is ample evidence that the teachers' judgements are more valid than, and equally reliable as, tests, there is a danger of unwanted impact on teaching as long as results are used for ‘high stakes’ evaluation of teachers and schools. Implications for policy include an end to the practice of using the results of pupils' summative assessment, however they are derived, as the sole basis for target setting and school accountability.