Speech intelligibility and listening effort in university classrooms for native and non-native Italian listeners

Autor: Simone Torresin, Chiara Visentin, Andrea Gasparella, Nicola Prodi, Francesca Cappelletti
Rok vydání: 2019
Předmět:
Zdroj: Building Acoustics. 26:275-291
ISSN: 2059-8025
1351-010X
DOI: 10.1177/1351010x19882314
Popis: Listening effort describes the allocation of attentional and cognitive resources for successful listening. In adverse conditions, the mental demands for listening increase, interfering with other cognitive functions. This is especially relevant in learning spaces, where complex tasks that recruit more cognitive resources (e.g. memorization of information and comprehension) are performed by the students. This study focuses on the case of university classrooms and investigates the effects of different types of masking noise on both speech intelligibility and listening effort. Speech-in-noise tests in the Italian language were presented to 25 young adults with normal hearing (13 native and 12 non-native listeners) within an existing university classroom located in Bozen-Bolzano (Italy). The tests were presented in three listening conditions (quiet, stationary noise, and fluctuating noise), grouping the listeners around two locations within the classroom. The task performance was assessed using both speech intelligibility and two proxy measures of listening effort: response time and subjective ratings of effort. Longer response times and higher subjective ratings were taken to reflect increased listening effort. Results in noisy conditions were compared to the quiet condition. A disadvantage in task accuracy performance was found for non-native compared to native listeners; concerning response time, it was found that when the target signal is masked by a fluctuating noise, additional processing time is requested to non-native listeners compared to their native peers. The interaction was not pointed out by subjective ratings, supporting the hypothesis of a different sensitivity to listening conditions of the two proxy measures of listening effort.
Databáze: OpenAIRE