Processing load during listening: The influence of task characteristics on the pupil response

Abstract

This study examined the magnitude of the pupillary response evoked by a number of tasks varying in the nature and complexity of the auditory and linguistic information provided. The tasks comprised passive listening, anticipation to verbally responding to a prompt signal, auditory detection, and the identification of meaningful words. Performance in the auditory detection and identification tasks was matched at 79% correct. In all, 42 normally hearing adults (aged 18–44 years, mean age 25.5 years) from three different sites (Amsterdam, Cologne, and Warsaw) participated. During each condition, the pupil diameter was measured. A Repeated Measures ANOVA was conducted to examine within and between subject (site) differences in the pupil response over 8 time intervals during the four conditions. The maximum mean pupil dilation was largest in the words-in-noise identification task (0.13 mm) and differed significantly from the maximum mean dilation in the noise-in-noise-detection task. The latter did not differ significantly from the pupil response during passive listening to noise and an answer prompt. No significant differences between sites were observed. Task evoked pupillary responses to theory-based measures of linguistic processing are robust, reliable, and sensitive to differences in task demands. Word-in-noise identification requires more processing load than nonspeech detection. To obtain information about within-subject differences in auditory processing, examination of both processing load and behavioural performance is recommended. Methodological implications are discussed.

Keywords:

• Processing load
• Pupil dilation
• Task demands
• Linguistic
• Words
• Background noise
• Detection

Acknowledgments

Part of the research reported herein was supported by Marie Curie Host Fellowships for Transfer of Knowledge; Remediation of Hearing Loss; No. 042387. We greatly acknowledge the assistance of Hans van Beek in the development of the data analysis software and testing equipment.