This symposium presents three studies of moderation practice to achieve valid and reliable assessor judgements. The first, conducted by the Qualifications and Curriculum Authority, England, analyses models of moderation within assessing pupils' progress. Moderation proved a highly effective way of learning how to improve the quality of teachers’ own assessments as well as how to confirm the assessments of others. Findings revealed a positive impact on the nature and range of evidence used to support assessment, the accuracy of assessment, and the understanding of what characterises performance at a particular national curriculum level.
The second study, analyses Queensland’s senior secondary system of externally moderated school-based assessment. Students who take the same Queensland Studies Authority subject in different schools and who achieve the same standard through assessment programmes based on a common syllabus should be awarded the same level of achievement. The underpinning principles of reliability and validity, essential to any high stakes assessment system, will be highlighted in the analysis of how comparability of students’ results with respect to levels of achievement is realised.
The final study, funded by ARC, QSA and NCCA, is situated in the middle years of schooling where teacher use of standards to judge the quality of student work has been introduced. The focus is on moderation and the use of standards to inform teacher judgment. Findings include: that the ‘meticulous specification’ of the procedure for making judgments results in an analytic approach while the use of representations of standards as continua helps promote a more holistic approach. Secondly, different disciplines deploy standards in different ways. Thirdly, annotated exemplars and accompanying commentary to illustrate standards prove beneficial to teacher judgment, however greater support, in the form of guidelines about the approaches and models of moderation, is required to achieve reliability and consistency.