Aspects of experimental design for the perceptual evaluation of violin qualities

A method for the perceptual evaluation of violin qualities is presented. To obtain reliable results from quality assessments, it is important to consider the statistical validity of the experimental procedure. The number of players can be maximized to better estimate the extent of inter-subject differences. For the purposes of our experiment, twenty violinists with at least fifteen years of performance experience were selected. Maximizing task repetitions as much as possible is desirable, but there are logistical constraints such as the total duration of the experimental session or physical limitations such as fatigue that must also be considered. Although having many repetitions can help reduce the experimental noise in the data, fatigue may have the opposite effect. Participants were asked to play the different violins and order them by preference. Results indicate that violin players are relatively self-consistent when evaluating different instruments in terms of preference.