Some Preliminary Data on the Use of Self-Evaluations and Peer Ratings in Assigning University Course Grades

Abstract

AbstractThis paper presents the results of three investigations of attempts to increase university student participation in the determination of their final course grades. Increased participation took the form of 1) self-evaluations and 2) peer-ratings. The data from two experiments indicated that students were unable to assign their own grades realistically. This was shown by low agreement between self-evaluations and peer-ratings, and self-evaluations and instructor ratings. Self-ratings were always more favorable than peer and instructor ratings when disagreements occurred. Although many students felt that peers were performing at C and D levels, no students believed that he deserved less than a B grade. The third experiment showed that peer-ratings yielded: (1) a more typical grade distribution, (2) a high degree of internal consistency, and (3) good agreement with instructor ratings.