Application of fuzzy ordinal peer assessment in formative evaluation

Peer assessment has been used for many years as a tool to improve learning outcomes but, only recently, it is becoming an increasingly used support also in students evaluation. Many approaches have been proposed so far to make peer assessment as reliable as possible even in case of incorrect or inaccurate evaluations proposed by students. Among these approaches, Fuzzy Ordinal Peer Assessment (FOPA) relies on ordinal evaluations (rather than cardinal ones) and on the application of models coming from Fuzzy Set Theory and Group Decision Making. FOPA has already demonstrated good results in in-silico experiments. To complement these results, in the work presented in this paper, we experiment the same model in a University context to support formative evaluation. Obtained results show better performance of FOPA with respect to competitor models and a general attitude of peer assessment models to approximate instructor ratings.