A Measure to Cultivate Engaged Peer Assessors: A Validation Study on its Efficacy

Abstract

Despite the generally positive learning effects of peer assessment, undesirable behaviors exhibited during the process have been reported (e.g., peer assessors engaging at a superficial level, or giving biased judgements). With reference to related literature and based on non-participant observation of student assessors’ behavior in classrooms and document analysis of past student assessors’ work, a measure consisting of four variables was devised to serve two purposes: on the passive side, to alleviate such reported hindrances; on the active side, to engage peer assessors in sensible, prudent ratings. The devised measure quantifies the performance of student assessors based on the scores they give to their peers’ performance/work as compared to that of the teacher/expert on each assessment criterion and two other noteworthy variables (i.e., fine discrimination ability, completion rate). A validation study which involved two classes of university sophomores was conducted. The students presented individual projects while participating in assessing their peers’ performance via an online system, which only differ in whether the devised measure was in use (the experimental group, N = 53) or not (the comparison group, N = 47) for the two respective class. The statistically different results in peer assessors’ performance between the two treatment groups, t (98) = 8.97 < .001, attested the efficacy of the devised measure in encouraging higher quality peer assessments.