This project aimed to evaluate the use and educational effectiveness of adaptive comparative judgement (ACJ) for peer assessment of solutions to exam-style problems. ACJ is based on ‘Thurstone’s Law of Comparative Judgment’, in which pairwise comparisons are used to estimate the relative quality of a set of solutions.
We used an online system which automates the operation of the underlying algorithm and presents student judges with pairs of pieces of work for comparison over iterative cycles.
Contact:
Project dates:
October 2014 to July 2015
Funding:
Higher Education Academy (HEA)
Direct link: