11th Speech in Noise Workshop, 10-11 January 2019, Ghent, BE

Do you understand the teacher? Case study: the effect of background noise and vocal effort on speech intelligibility in a classroom of a primary school

Berdien De Herdt(a), Michel Schellens
Department Environmental Research, Provincial Institute of Hygiene, Antwerp, Belgium

(a) Presenting

To obtain a good understanding in classrooms, a suitable reverberation time (RT) and a low background noise level (BGN) are required, especially for children with communication disorders. An objective method to evaluate speech intelligibility is by determining the Speech Transfer Index (STI).

The present study investigated the effect of changing the BGN due to different classroom conditions and the effect of raising the level of vocal effort on the STI for two listening positions and one source position. The STI was determined and evaluated in accordance with the international standards IEC 60268-16. The RT was measured according to the European Standard ISO 3382-2:2008 and evaluated by the Belgian standard NBN S 01-400-2.

A MLS signal was played through the artificial mouth-directional sound source that was placed where the teacher was located (at a standing height of 1,60 m). There were two measuring positions: (1) located in the center of the classroom at a distance of 2 m from the source and (2) located at the back of the classroom at a distance of 4 m from the source. Both microphone positions were set on a sitting height of 1,40 m. Separate measurements of the impulse response (N= 5) and the BGN were conducted. These were combined when calculating the speech intelligibility in Dirac for different scenarios with different signal-to-noise ratios (SNRs).

The STI was evaluated as ‘good’ when there was a low level of BGN (position 1 = 0,67; position 2 = 0,60). During a regular lesson, the ambient noise level was induced by babbling between the children, moving of chairs, the use of electrical sharpeners and noise coming from the adjacent rooms. The STI was considered ‘poor’ for position 1 (STI = 0,32) and ‘bad’ for position 2 (STI = 0,29). When the teacher raised the voice, the STI became ‘fair’ for position 1 (STI = 0,48), but stayed ‘poor’ for position 2 (STI=0,45). The RT was evaluated 'suitable', thus it was indicated that the level of BGN was causing the low STI values. In fact, the child with the communication disorder was sitting at an inadequate position.
This study showed that although the teacher was raising the voice, it had only little effect on the STI. We can assume that the teacher is at risk of developing vocal problems. All children will benefit of lowering the BGN and optimizing the room in terms of acoustic insulation and absorption.

Last modified 2018-12-08 00:23:30