Acoustics'08 Paris

[ Lay Language Paper Index | Press Room ]

Deaf children can hear well, but not always in classrooms due to acoustics.

Frank Iglehart, Ph.D.
Clarke School for the Deaf, Northampton, Massachusetts, USA

Popular version of paper 5pNSa3, "Speech perception in classroom noise and reverberation by children with typical and impaired hearing - Preliminary results"
Presented on Friday July 4, 2008, in Room 252A

Tens of thousands of children worldwide are profoundly deaf yet can learn to hear and speak clearly thanks to new technology called cochlear implants and well-designed habilitation. These tiny processors, worn comfortably behind the ear, convert sound into rapid digital pulses that transmit to a small receiver implanted under the child’s scalp, and from there to the brain by means of the auditory nerve. After surgical implant, the sound signals take on meaning with several years of gentle and consistent auditory habilitation. A deaf child can speak clearly and understand spoken language speech with great accuracy when in near-ideal listening conditions. These conditions can usually be found in a quiet living room or small office, with little noise and reverberation. (Reverberation is the reflection of sound off hard surfaces in a room. These reflections can “smear” speech and make it difficult to understand.)

Forty to sixty percent of communication in a classroom is estimated to be auditory. The results of this study suggest, however, that these children’s remarkable ability to understand spoken language may often be lost in classrooms. Studies show that many, if not most, classrooms around the world have too much noise and reverberation, even for students with typical (normal) hearing. Classrooms for young children, a place for critical language acquisition, tend to be the worst. As a result, not only is academic progress threatened for a child with hearing loss but many verbal interchanges so important to her social development may also be confused and misunderstood.

An eight-year-old boy, deaf and wearing a cochlear implant, recently participated in a study of classroom acoustics. When he came into the office, he appeared bright, verbal and communicated easily through hearing alone with his implant. Later, in a quiet classroom that had no acoustic treatment for control of reverberation, the child was unable to understand what was said to him. He understood only every other word or so. Test scores indicate he is more than a year behind in vocabulary acquisition, and his father reports that he comes home most days from school exhausted. As acoustic panels were brought into the test classroom and reverberation decreased, his ability to understand speech began to return.

In order to determine the acoustics a child with hearing loss may need in a classroom, a research project at Clarke School for the Deaf [] in the US evaluates the effect of noise and reverberation on the perception of speech by students with cochlear implants or hearing aids in a classroom setting. Children in the study listen in a classroom with a “teacher” speaking in the presence of background noise. The teacher is a loudspeaker at the front of the classroom at a loudness level that approximates a teacher’s voice. The child sits 10 feet away, about the second row, and listens to and repeats simple sentences coming from the loudspeaker. (The sentences are the BKB-SIN test produced by Etymotic Research - Noise (four-talker babble) comes from four speakers throughout the room at a loudness level soft enough for the child to hear her best, and then rises with each sentence until the child can perceive few, if any, words correctly. The noise levels encompass what are known to exist in classrooms in many countries. This process repeats six times in each of three different reverberant situations: high, moderate and low durations of reverberation. Reverberation is controlled by means of addition or removal of commercially-available acoustic pads. The resulting durations of reverberation represent everyday classrooms.

Children, ages five to 16 years, with hearing loss and those with typical (sometimes called “normal”) hearing participate. These tests are providing a broad range of information, and some patterns are emerging. Noise disrupts spoken communication to widely different degrees depending on inherent hearing ability. In one test condition, the “teacher’s” voice is about six decibels (dB) louder than the background noise, or a +6 dB speech-to-noise ratio (S/N). This is labeled +6 dB S/N [LISTEN: 0.0s RT +6 dB SNR]. Researchers from several countries report that classrooms are often not this quiet. At this speech-to-noise ratio, participants with typical hearing in the study correctly perceive between 96% and 98% of the words presented, depending on the amount of reverberation. Students with cochlear implants in the study, however, understand only 20% to 35% of the words, functionally too little to follow a casual conversation much less a class lesson with ever-evolving vocabulary and content. When noise reduces to +21 dB S/N, scores for students with hearing loss improve dramatically. If reverberation time (RT, or the amount of time a sound takes to decay by 60 dB) exists to the degree reported in some classrooms (0.9 seconds RT; LISTEN: 0.9s RT +21 dB SNR), children with typical hearing find speech easy to understand; the reverberation is barely noticeable. Students with severe-to-profound hearing loss, however, cannot follow spoken sentences in this degree of reverberation, and score only 64% words correct on average. This score indicates an overall inability to follow by hearing alone anything other than simple, familiar conversation. Reduce the reverberation (0.3s RT) with acoustic pads, and these children perceive 79% words correctly. At this score level, communication becomes less challenging. With no reverberation [LISTEN: 0.0s RT +21 dB SNR], these children can understand 90% of the words, and have much greater auditory access to class lessons and the learning process. The reduction in reverberation time provides a modest improvement in scores for the students with typical hearing. Students with mild to moderate degrees of hearing loss demonstrate trends between those for students with typical hearing and those with severe-to-profound hearing loss.

These data help reveal the minimal classroom acoustic conditions in which students with cochlear implants can optimally hear. The next step is to have classrooms in which they, like the boy with the cochlear implant, can optimally hear and learn.

This research is funded by the Gustavus and Louise Pfeiffer Research Foundation and the National Institute on Disability and Rehabilitation Research.

[ Lay Language Paper Index | Press Room ]