A Comparison between the Efficacy of Task-Based Vs. Data-Based sEMG Sonification Designs

Abstract

Historically, many sonification designs that have been used for data analysis purposes have been based on data characteristics and have not been explicitly based on the listener’s task. These sonification designs have often been described as annoying, confusing, or fatiguing. In the absence of a generally accepted theoretical framework for sonification design, there is a need for improvements in sonification design as well as a need for empirical evaluation of taskbased sonification designs. This research focuses on surface electromyography (sEMG) sonification and two sEMG data analysis tasks: determining which of two muscles contracts first and which of two muscles exhibits a higher exertion level. Both of these tasks were analyzed using a task analysis technique known as GOMS (Goals, Operators, Methods, Selection Rules) and two sonification designs were created based on the results of these task analyses. Two Data-based sEMG sonification designs were then taken from the sEMG sonification literature, and the four designs (2 Task-based and 2 Data-based) were empirically compared. Significant effects of sonification design on listener performance were found, with listeners scoring more accurately using the Task-based sonification designs. Based on these results, we argue for wider application of task analysis methods to sonification design and for the inclusion of task analysis methods into a generally accepted theoretical framework for sonification design.