Student Peer Assessment: a research study in a level III core course of the bachelor chemical engineering program

Abstract

A study of anonymous student peer marking in a level III (third year) core course of the bachelor chemical engineering has shown that there was no significant difference (p > 0.05) in the average marks awarded by student assessors who had idealized solutions of the lecturer compared with those who did not; although more students with solutions (84%) completed the peer task than those without (69%). Students in a cohort of 64 (21 females, 43 males) were randomly assigned as a student-pair and tasked to mark each other's solutions to three numeric-type problems out of a possible 50, but only one student had idealized solutions. In 49 valid responses, 27 with and 22 without solutions, the maximum mark awarded by any assessor was 49 and the maximum awarded by the tutor was 50. The overall mean grade over the three problems was 14.3 for those with solutions and 14.5 without. The overall mean grade of the experienced tutor was 14.6. Despite this agreement in mean marking there were notable differences between student assessors and tutor marks in particular cases. The problems required stage calculations with reflux and column efficiencies; each had the marks indicated to be awarded for all sub-sections. Granularity in grades for both student assessors and tutor was a 1/2. There was no evidence of student collusion in marking. Independent survey results showed more than 3/4 of all student assessors highly valued this learning experience and that it stimulated interest in the course material. Students without solutions however were marginally less likely to want to have peer assessment in other courses (p = 0.095).