Development of an adaptive test of musical scene analysis abilities for normal-hearing and hearing-impaired listeners

Resumen

Auditory scene analysis (ASA) is the process by which the auditory system makes sense of complex acoustic environments through organizing sound mixtures into meaningful events and streams. Although music psychology has acknowledged the fundamental role of ASA in shaping music perception, no efficient test to quantify listeners’ ASA abilities in realistic musical scenarios has been published yet. This study presents a new tool for testing ASA abilities in the context of music, suitable for both normal-hearing (NH) as well as hearing-impaired (HI) individuals: The adaptive Musical Scene Analysis (MSA) test. The test uses a simple ‘yes-no’ task paradigm to determine if the sound from a single target instrument is heard in a mixture of popular music. During the online calibration phase, 525 NH and 131 HI listeners were recruited. The level-ratio between the target instrument and the mixture, choice of target instrument, and number of instruments in the mixture were found to be important factors affecting item difficulty, whereas the influence of the stereo width only had a minor effect. Based on Bayesian logistic mixed effects model, an adaptive version of the MSA test was developed. In a subsequent validation experiment with 74 listeners (20 HI), MSA scores showed acceptable test-retest reliability and moderate correlations with other music-related tests, pure-tone-average audiograms, age, musical sophistication, and working memory capacities. The MSA test is a user-friendly and efficient open-source tool for evaluating musical ASA abilities and is suitable for profiling the effects of hearing impairment on music perception.

Áreas y palabras clave

Cómo citar (APA 7)

Hake, R., Bürgel, M., Nguyen, N. K., Greasley, A., Müllensiefen, D., & Siedenburg, K. (2023). Development of an adaptive test of musical scene analysis abilities for normal-hearing and hearing-impaired listeners. https://doi.org/10.31234/osf.io/udq6p