Task Analysis for Sonification Applications Usability Evaluation
Ag Asri Ag Ibrahim · 2006 · Proceedings of the 8th International ACM SIGACCESS Conference on Computers and Accessibility (Assets '06) · doi:10.1145/1168987.1169060
Summary
This poster paper from University Malaysia Sabah and the University of York proposes a task analysis framework for evaluating the usability of sonification applications — software that transforms data into sound representations. The author argues that while usability evaluation has become essential for sonification applications, existing usability inspection methods are designed for graphical user interfaces and are not well suited to auditory interfaces. The framework is built on a unified HCI Sonification Application (SA) Model that describes three data transformation stages: raw data (RD) is converted into soundable data (SD), then into acoustic ready data (ARD) with acoustic parameter attributes, and finally into sound representation (SR) that the user hears. These transformations are analyzed from three perspectives — data, acoustics parameters, and sound representation — and three viewpoints: user tasks (performed entirely by the user), interaction tasks (user interacts with the system), and application tasks (performed by the system without user intervention). This produces nine categories of task combinations that together cover the full scope of sonification application functionality.
Key findings
The framework defines a structured task description format with five components: category (one of nine task types), goal, type of task (identifying verbs/actions from the goal), task description (open explanation of how to accomplish the task), and list of actions. Input and output specifications are also detailed, categorizing data by type (nominal, ordinal, interval, ratio, discrete, categorical, continuous), organization (category, time, location, mnemonic, continuum, table), and information level (point, regional, global). Acoustic parameters are specified by physical properties (frequency, amplitude, spectrum, etc.), perceptual properties (loudness, pitch, spatial, timbre), number and style of parameters, coordination mode (simultaneous, alternate, separate), and sound category (point, discrete, regional, branching, global, continuous). The author proposes that evaluators can use these task descriptions as inspection materials, following the data flow through the transformation diagram to detect design anomalies without needing to execute a strict sequential walkthrough of every step.
Relevance
This paper addresses a methodological gap in evaluating non-visual interfaces that remains relevant today. As sonification is increasingly used in accessibility contexts — helping blind and visually impaired users interact with data visualizations, monitor system status, and navigate environments — having systematic methods to evaluate the usability of these auditory interfaces is important. The framework's insight that sonification applications involve a complex chain of data transformations (raw data to soundable data to acoustic parameters to final sound) provides evaluators with specific points to inspect for usability problems. While the paper is theoretical and does not include empirical validation, it contributes a structured vocabulary and analytical approach for thinking about how data becomes sound and where that process can fail from the user's perspective. For accessibility practitioners designing auditory interfaces, the framework offers a systematic way to ensure that the mapping from data to sound is comprehensible and useful.
Tags: sonification · task analysis · usability inspection · auditory display · accessibility evaluation · non-visual interface