Self and Peer Evaluation in Undergraduate Education: Structuring Conditions That Maximize Its Promises and Minimize the Perils

This study canvasses reliability of students' self and peer evaluation, a method of assessment of university students that has recently gained renewed pedagogical interest and broad recognition. Two experiments, imbedded in classroom curriculum, examined the effects of the instrument of evaluation (with criteria vs. no criteria for evaluation provided), the format of evaluation procedure (anonymous vs. non-anonymous), and motivation of students (strong vs. weak) on the reliability of students' self and peer ratings. The results of the experiments revealed both a considerable unreliability of peer ratings in some cases as well as a notable consistency of peer evaluations in others. The instrument of evaluation with criteria provided had significant positive effect on the reliability of peer evaluations. Students strongly motivated to apply criteria for evaluation also produced more reliable peer evaluations compared to their peers that were weakly motivated to apply them. The results of the impact of the condition of anonymity were mixed. The study found that students exhibit significant “self-enhancement” bias by rating their academic performance above that of their peers. The positive self-bias does not disappear when students are offered criteria for evaluation of their work.