Supporting criteria-based marking

School:  Informatics

Team Members: Paul Anderson, Susan Rhind, Anna Wood

Abstract

Criteria-based marking is capable of providing more reliable assessment and more meaningful feedback to the students. Using finer-grained criteria provides more specific feedback and helps to mitigate the variation which inevitably occurs with multiple markers and large classes.  However composing these criteria in a meaningful way is not always straightforward, especially when a fine grained numerical result is required – as for the Common Marking Scheme. For a number of years, we have been experimenting in Informatics with software to compose larger numbers of fine-grained criteria. This has the potential to produce a fine-grained, interpolated result and to identify specific strengths and weaknesses as feedback to the student. It also allows the criteria to be easily adjusted to ensure that the result represents a meaningful assessment of the appropriate objectives.

We have recently been working to create a version of this experimental software which is usable by other courses. The aim of this project is to evaluate this software and the criteria-based approach in general, in a different context – specifically vet medicine. This work has the potential for impact across multiple contexts and disciplines.

Final project report

Download the final project report (PDF)