On the design and validation of a rubric for the evaluation of performance in a musical contest

Evaluation of music performance in competitive contexts often produces discrepancies between the expert judges. These discrepancies can be reduced by using appropriate rubrics that minimise the differences between judges. The objective of this study was the design and validation of an analytical evaluation rubric, which would allow the most objective evaluation possible of a musical solo performance in a regulated official competition. A panel of three experts created an analytical rubric made up of five review criteria and three scoring levels, together with their respective indicators. To validate the rubric, two independent panels of judges used it to score a sample of recordings. An examination was made of the dimensionality, sources of error, inter-rater reliability and internal consistency of the scores coming from the experts. The essential unidimensionality of the rubric was confirmed. No differential effects between raters were found, nor were significant differences seen in each rater's internal consistency. The use of a rubric as tool for evaluating music performance in a competitive context has positive effects, improving reliability and objectivity of the results, both in terms of intra-rater consistency and agreement between raters.