Exploring self-assessments in university undergraduate students: how accurate are they?

Abstract

Self-assessment, whereby students are actively engaged in assessing the quality of their work, has been shown to benefit them. It is not routinely carried out in all institutions. This pilot study aimed to explore the extent to which students chose to engage with self-assessment when invited to do so, and how accurate they were when they did. A short pilot tool including qualitative and quantitative elements, was circulated to students within a school of the largest faculty of Kingston University. Students completed the self-assessment and submitted it with their completed assignments. Actual grades achieved were compared with self-assessments. Qualitative data were analysed using basic thematic analysis. The highest average marks achieved were in the group who correctly self-assessed their work. More students incorrectly self-assessed than correctly assessed their work, and almost a third of students did not engage with the activity. Those who incorrectly over-assessed their work had average marks similar to those that did not engage with the activity, significantly lower than the average marks achieved by the incorrect under-assessors and the correct self-assessment groups. Correct self-assessing students were more specific about the skills they demonstrated and the support they used for their assignments.