Inter-subject standards: an insoluble problem?
PDF | 259.37 KB
Whatever their constitution or the type of assessments they administer, it is a prime responsibility of all awarding bodies to engender public confidence in the standards of the qualifications they endorse, so that they have not only usefulness but credibility. Although guaranteeing comparability of standards between consecutive years is relatively straightforward, doing so between different subjects within the same qualification and with the same grading scheme is a far more complex issue.
Whether standards are established judgementally or statistically or, as in most contexts, a mixture of the two - satisfying public and practitioner opinion about equivalence is not easy. Common grade scales signify common achievement in diverse subjects, yet questions arise as to the meaning of that equivalence and how, if at all, it can it be demonstrated.
With the increase in qualification and credit frameworks, diplomas and so forth, such questions become formalised through the equating of different subjects and qualifications, sometimes through a system of weightings.
This paper is based on two collaborative presentations made to the International Association for Educational Assessment (IAEA) conferences in 2003 and 2004. It summarises some recent concern about inter-subject standards in the English public examination system, and proceeds to describe three systems. use of similar statistical approaches to inform comparability of inter-subjects standards.
The methods are variants on the subject pairs technique, a critique of which is provided in the form of a review of some of the relevant literature. It then describes New Zealand.s new .standards-based. National Qualifications Framework, in which statistical approaches to standard setting, in particular its pairs analysis method, have been disregarded in favour of a strict criterion-referenced approach.
The paper concludes with a consideration of the implicit assumptions underpinning the definitions of inter-subject comparability based on these approaches.
Jones, B., Philips, D., and van Krieken R. (2005). Inter-subject standards: An insoluble problem? Manchester: AQA Centre for Education Research and Policy