Investigating the Judges Performance in a National Competition of Sport Dance

Many sports, such as gymnastics, diving, figure skating, etc. use judges’ scores to generate a rank for determining the winner of a competition. These judges use some type of rating scale when assessing performances. Human ratings are subject to various forms of error and bias. The overall outcomes may largely depend upon the set of chosen raters. The aim of this paper is to illustrate how results from the Many-Facet Rasch Measurement framework can be used to highlight feedback to judges about their scoring patterns. The purpose is to analytically detect anomalous rater behaviours. We consider the field of Sport Dance, a discipline which enjoys increasing public interest and passion in recent years. We analyze data relating to two national competitions held in Italy in 2018 and 2019.