Exploring Effective Relationships Between Visual-Audio Channels in Data Visualization

In recent years, there has been a growing trend towards taking advantage of audio--visual representations. Previous research has aimed at improving users’ performance and engagement with these representations. The attainment of these benefits primarily depends on the effectiveness of audio--visual relationships used to represent the data. However, the visualization field yet lacks an empirical study that guides the effective relationships. Given the compatibility effect between visual and auditory channels, this research presents the effectiveness of four audio channels (timbre, pitch, loudness, and tempo) with six visual channels (spatial position, color, position, length, angle, and area). In six experiments, one per visual channel, we observed how each audio channel, when used with a visual channel, impacted users’ ability to perform the differentiation or similarity task accurately. Each experiment provided the ranking of audio channels along a visual channel. Central to our experiments was the evaluation at two stages, and accordingly, we identified the effectiveness. Our results showed that timbre, with spatial position and color, aided in more accurate target identification than the three other audio channels. With position and length, pitch allowed a more accurate judgment of the magnitude of data than loudness and tempo but was less accurate than the other two channels along angle and area. Overall, our experiments showed that the choice of representation methods and tasks had impacted the effectiveness of audio channels. Graphical abstract