Effects of environmental acoustic factors, individual differences and musical training on speech perception in simulated classrooms

Abstract: Formal learning takes place primarily through speech perception in classroom environments and is therefore dependent on the listener’s ability to cope with a number of acoustic factors that interfere with the speech signal. As classrooms are inclusive spaces accommodating learners with a broad range of abilities and backgrounds, this research investigated some of the ways in which individual differences in supporting cognitive skills are related to speech perception outcomes in various challenging acoustic environments. A potential avenue for the remediation of speech perception problems was also investigated. Studies comparing trained musicians’ with non-musicians’ speech perception have consistently shown a “musicians’ advantage” for speech-in-noise tasks, therefore this research longitudinally investigated whether one year of musical training was capable of producing improved speech perception thresholds in children. Therefore, the first three studies investigated relationships between the cognitive abilities of the listener and speech perception under various challenging environmental conditions, and the final study reported whether a year of musical training was able to produce learning effects generalizable to speech perception in the same challenging auditory environments tested in the two preceding experiments. Tests involved attending to a target talker underexperimental conditions in which the target signal was increasingly difficult todiscern either by lengthening reverberation time or by adding noise (i.e., competing sounds or talkers). In the second, third and fourth studies, the configuration of target and noise sources in the simulated room were manipulated to be either collocated or spatially separated from one-another. In order to additionally explore the relationship between speech perception and underlying cognitive processes, a number of measures were taken including phonological working memory capacity (number updating, digit span) and language assessments (vocabulary, expressive language) and analysed in relation to speech perception outcomes under the various experimental conditions. In all four studies, speech perception was tested in virtual classroom environments that were simulated based on actual classroom acoustic measurements taken in participating Swedish schools. The cumulative findings of this body of work linked differences in language ability, background and performance on various cognitive tests to speech perception thresholds, suggesting that not all learners are on equal footing in the classroom environment. However, musical training of the intensity and duration provided was unable to improve group performanceon speech perception or cognitive measures.