Abstract

The aim of reducing central regulation of qualifications and allowing a range of different providers to offer the qualifications they judge to be appropriate may have educational benefits, as well as fitting with current political directions. Against this, the pressure for regulation comes partly from the need to ensure comparability of standards across different specifications, subjects and qualification types. How can we reconcile these conflicting demands? Can we reduce regulation and allow a diverse range of qualifications, but still maintain comparability?

In this talk Professor Coe will propose a structure for doing this. The starting point is that there is no such thing as absolute comparability, only comparability for a specific purpose. For key purposes such as using qualifications to select applicants to higher education, or for school accountability, there are a number of statistical models that can be used to define comparability at the point of use, rather than in the process of awarding. These would allow awarding bodies to define, judge and certify outcomes as they see fit, but then to assign different values to those certified outcomes (eg grades) according to their relevance to the particular purpose being considered. Examples will be given, together with a discussion of pros and cons of different approaches.