Auditory and Visual Evaluation of Fixed-frequency Events in Time-varying Signals
Alexander, Robert L.
Gilbert, Jason A.
Zurbuchen, Thomas H.
MetadataShow full item record
This study directly compares the auditory and visual analysis capabilities of participants in a structured data analysis task. This task involved the identification of transient fixed-frequency sinusoid events that were embedded within white noise and noise derived from solar wind time series. It was hypothesized that participants would be able to identify the number of embedded events more quickly and accurately through auditory data analysis than through visual analysis. While visual analysis outperformed auditory analysis overall, additional investigation revealed that auditory analysis outperformed vision in instances where these events were embedded in solar wind data. This task - involving the detection of transient periodic activity occurring within background turbulence - closely mirrors a type of spectral analysis conducted by heliospheric scientists. Additionally, several data examples contained embedded events that were correctly identified through audition while being consistently overlooked through visual inspection. The largest disparity between visual and auditory performance was found in the analysis of white noise spectra that contained no embedded events. In these instances, auditory analysis regularly resulted in the identification of events when none were present; a potential reasoning for these false positives is discussed. The results of this study suggest that the analysis capabilities of each modality may vary based largely on the complexity of the masking stimuli that are present.