Abstract
A partially-automated method of assessment is proposed, in which automated question setting is used to generate individualized versions of a coursework assignment, which is completed by students and marked by hand. This is designed to be (a) comparable to a traditional written coursework assignment in validity, in that complex and open-ended tasks can be set with diverse submission formats that would not be suitable for written examination or automated marking; and, (b) comparable to e-assessment in terms of reduction of academic misconduct, with individualization acting as a barrier to copying and collusion. This method of assessment is implemented in practice. Evaluation focuses on expert second-marking, student feedback and analysis of marks, and aims to establish that the partially-automated method can be useful in practice. The partially-automated method proposed appears to be capable of adapting a coursework assignment to make it less sensitive to copying and collusion (and therefore more reliable) while maintaining its validity, though leading to reduced efficiency for the marker. This paper therefore contributes the introduction of a novel approach to assessment which offers a way to bring automated individualization to the assessment of higher order skills in higher education mathematics.
Acknowledgements
An early version of this paper was published in the proceedings of the 8th British Congress of Mathematics Education. I am grateful to Christian Lawson-Perfect (School of Mathematics and Statistics, Newcastle University) for adapting the Numbas e-assessment system for my experiment. The ‘worksheet’ theme he created to enable my partially-automated approach remains part of the free system at numbas.mathcentre.ac.uk. I am also grateful for the anonymous contributions to this work by the students who gave feedback, the second-markers and the other lecturer who administered the survey with group B. I am also grateful to an anonymous reviewer of this article who provided thought-provoking and constructive feedback which has improved the final article.
Disclosure statement
No potential conflict of interest was reported by the author.