Acoustical Quality, Student Learning, and Teacher Health
Good acoustical quality in classrooms is critical for student learning. Research has shown that noise exposure affects educational outcomes and provides evidence of mechanisms that explain the effects of noise on learning. Speech intelligibility studies indicate that students’ ability to recognize speech sounds is decreased by even modest levels of ambient noise, and this effect is magnified for younger children. This problem is frequently not appreciated by adults, who are better able to recognize speech in the presence of noise.
Most learning activities in school classrooms, especially for younger children, involve speaking and listening as the primary communication modes: Students learn by listening to the teacher and to each other (Goodland, 1983). Excessive background noise or reverberation (i.e., multiple delayed reflections of the original sound) can interfere with speech perception and thereby impair learning. Careful attention to acoustical design requirements, then, is essential for creating an effective learning environment. Nonetheless, a 1995 report of the U.S. General Accounting Office estimated that the acoustical quality in approximately 22,000 U.S. schools attended by 11 million students was unsatisfactory (GAO, 1995a).
Figure 6.1 illustrates the four typical major sources of noise in classrooms and reflected speech sounds (reverberation).
HVAC systems are perhaps the most common source of ambient noise in classrooms. However, significant levels of noise can also be transmitted through walls and windows from outdoors or from adjacent indoor spaces. Reflected sound within a classroom, if uncontrolled, degrades speech intelligibility.
One way to describe the desired acoustical quality in a classroom is to specify an acceptable maximum ambient noise level. This level is measured in terms of A-weighted sound levels or octave band sound levels that can be used to determine other measures such as noise criterion (NC), room criterion (RC), or balanced noise criterion (BNC) values. By combining the effects of sound at different frequencies in a manner similar to that which takes place in the human hearing system, these measures rate the loudness of sound to listeners. A second way to describe acceptable room acoustical quality is to specify the reverberation time, which is approximately the time it takes for a loud sound to die away to inaudibility after the source is turned off. Reverberation times increase with room volume and decrease as sound-absorbing material is added to a room. However, excessive sound-absorbing treatments will have the negative effect of reducing speech levels and degrading the intelligibility of speech in a classroom.
People’s ability to understand speech is influenced largely by how loud speech sounds are relative to ambient noise or any other competing sounds, hence the importance of an adequate signal-to-noise ratio (i.e., speech to background noise ratio) for a classroom to function well. Reverberant sound causes one word to smear into the next and can decrease the intelligibility of speech. Acoustical design should be aimed at improving the recognition of speech sounds in the classroom. The focus should be first on reducing unwanted noise and then on controlling excessive reverberation. Good acoustical design can facilitate learning by allowing for more accurate verbal interaction and less repetition among teachers and students because spoken words are clearly understood. There is also evidence that good acoustical design may have a health benefit for teachers
by reducing the potential for vocal impairment, and it may have the ancillary benefit of reducing teacher absenteeism. This issue is discussed later in this chapter.
EFFECTS OF EXCESSIVE NOISE
Excessive noise can interfere with learning by affecting memory (Hygge, 2003) and acting as a distraction that impairs a student’s ability to pay attention. The ability to pay attention is most important when students are engaged in tasks that demand higher mental processes, such as learning new concepts, or when teachers are presenting new or complex information (Hartman, 1946). (See also Anderson, 2004, for a review of the effects of noise on children and classroom acoustics issues.)
Excessive background noise in a classroom can come from outside the building (aircraft and traffic noise, lawn mowers and other equipment, or students engaging in sports activities) or from within it (heating, ventilation, air conditioning, plumbing systems, adjacent classrooms, hallways, gymnasiums, or music rooms) or from the students themselves. The level of residual noise from the students may be dominant, but is strongly related to the ambient noise in the room. That is, student chatter will increase as the general level of ambient noise increases, an example of the Lombard effect (Junqua, 1996).1 Thus, it is important to minimize all other sources of noise to ensure lower levels of student noise. It is equally important to educate teachers about the effects of noise on speech communication. As adults, teachers may not appreciate the additional problems that noise creates for younger listeners.
Although the importance of classroom acoustics to educational outcomes is well supported in the research literature, it is frequently ignored by school officials and by those designing schools. Anderson (2004, p. 118) suggests there are at least four reasons for this:
First, administrators walk into classrooms, listen briefly with adult ears and do not recognize that auditory immaturity causes young children to experience greater listening problems and less coping ability than the mature auditory systems of adults. Second, most school administrators have not been exposed to the extensive body of research that illustrates the effects of excessive background noise and/or reverberation on students’ listening, learning, and behaviour. Third, administrators often believe that good classroom acoustics are only needed for children with hearing impairment and that children with hearing loss and auditory learning difficulties comprise only a very small proportion of the children educated in inclusive classroom settings. Fourth, school administra-
tors are typically unaware of the health issues faced by teachers who instruct in noisy classrooms and the expense that this may cost the school district.
Effects of Noise and Reverberation on Speech Perception
Speech perception studies have investigated how interference from noise and reverberation influences the recognition of syllables, words, or sentences in classrooms. Kindergarten and first and second grades are the main years in which children learn to break written words into their phonetic components and acquire the ability to read. Careful listening is needed to develop the ability to discriminate among minor differences in words such as pet, pit, pot, put, and pat (Anderson, 2004). Such differences can be lost in a noisy environment, so young children require the higher signal-to-noise ratios provided by quieter conditions.
The impacts of excessive noise vary according to the age of the students, because the ability to focus on speech is a developmental skill that evolves and does not mature until ages 13 to 15 years. As children mature they tend to develop strategies to cope with noise levels. Accordingly, young children require better acoustical environments than do adult listeners to achieve equivalent word recognition scores (Elliott et al., 1979; Elliott, 1979; Neuman and Hochberg, 1983). Classrooms of younger children are also found to be noisier (Picard and Bradley, 2001).
A student’s difficulty in understanding speech in noisy situations may not be recognized by teachers, building designers, or other adults. That is, adults cannot rely on their own perception of speech under adverse listening conditions to recognize a child’s difficulty under the same conditions. Elliott et al. (1979) found that the ability to recognize sentences in noisy environments improves systematically with age for children who are 5, 6, 7, 8, and 10 years old. Similar effects of children’s age on speech perception in noisy environments were reported by Finitzo-Hieber and Tillman (1978) and Marshall (1987).
Although it is clear that children need quieter conditions than adults to achieve high speech recognition scores (Elliott, 1979) and that the younger the children, the quieter the conditions should be, the results of the various studies are very different and do not agree with the results of similar tests carried out in actual classrooms (Bradley and Sato, 2004). This is thought to be so because the early laboratory studies used monaural (one-ear) headphones to present speech and noise signals, which increases the negative effects of noise and reverberation. Listening naturally with two ears confers a binaural advantage, enabling higher speech recognition scores in noisy environments. Nábĕlek and Robinson (1982) found binaural advantages of a similar magnitude for 10-year-olds and adults. However,
Neuman and Hochberg (1983) found greater binaural advantages for the youngest children (5-year-olds). That is, the speech recognition scores of the 5-year-olds improved when they listened normally with two ears rather than monaurally. It is clear that the older studies in which monaural headphones were used exaggerated the negative effects of noise on speech recognition. Nonetheless, it is well established that children need quieter conditions than adults to understand speech well and that their ability to understand speech in everyday noise improves with age.
There are also studies suggesting that the negative effects of excessive reverberation are more acute for younger listeners (Nábĕlek and Pickett, 1974; Nábĕlek and Robinson, 1982; Neuman and Hochberg, 1983). Some early studies used monaural headphone presentation of the test signals (Finitzo-Hieber and Tillman, 1978), which would exaggerate the negative effects of reverberation on speech recognition scores (Moncur and Dirks, 1967). Because shorter reverberation times led to improved speech recognition scores in these tests, they caused some to recommend very short reverberation times for classrooms. However, such conclusions are based on an incomplete understanding of room acoustics in that it is not possible to have both increased speech level (to maximize signal-to-noise) and reduced reverberation time. For example, adding sound-absorbing material to a room will reduce reverberation times (which usually helps speech intelligibility) but at the same time will reduce speech levels, leading to reduced signal-to-noise ratios and reduced intelligibility scores. Although reverberant speech sound degrades the intelligibility of speech, the early arriving reflections of the speech sound (those that arrive at the listener within about 0.05 seconds after the direct sound) significantly enhance the apparent loudness of the speech and improve speech intelligibility (Bradley et al., 2003). In many situations, it is the early reflection energy that makes it possible to understand speech in rooms (e.g., children listening farther from the teacher or children listening while the teacher’s voice is directed away from them). Thus adding too much absorptive material to reduce reverberation time will reduce speech loudness and impair intelligibility. This is why acoustics textbooks have for many decades referred to the need for optimum rather than minimum reverberation times (Knudsen and Harris, 1950).
Because too much noise and the reduced signal-to-noise ratios that accompany it are a more or less ubiquitous problem in classrooms, there must be enough reflected speech sound to maintain adequate loudness. Excessive noise is typically a more significant problem than is too much reverberation. Noise levels in classrooms can easily be 10 dB (or much more) too loud, indicating 10 times too much noise energy. In contrast, it is almost impossible for reverberation times to be 10 times too long. An optimum reverberation time will ensure that there are adequate speech
levels without excessive reverberation. A few studies have looked at this problem (e.g., Reich and Bradley, 1998; Hodgson and Nosal, 2002), but none have focused on the particular needs of young children. Although these studies are not conclusive, they indicate that a reverberation time in the range close to conventional recommendations (0.4 to 0.7 s) is probably acceptable. Following American National Standards Institute (ANSI) Standard 12.6, “Acoustical Performance Criteria, Design Requirements and Guidelines for Schools,” which recommends designing for a reverberation time of 0.6 s or a little less, and appreciating that a small deviation from this probably does not have a large effect on speech recognition scores in classrooms, seems to be the best advice now available.
EXCESSIVE NOISE AND STUDENT ACHIEVEMENT
In addition to degrading children’s ability to recognize speech sounds, excessive noise can also interfere with the performance of various tasks. For example, children have been found to be more likely to give up on solving difficult puzzles in noisier situations (Cohen et al., 1980). Interfering noise can be more distracting for younger children (Higgins and Turnure, 1984), and speech sounds can be more disturbing than neutral, ventilation-type noises (Carhart et al., 1969; Elliott et al., 1979). This latter result suggests that speech sounds from adjacent classrooms are much more distracting than many other types of noise. The interference with the recognition of speech sounds and with various tasks may explain, in part, reported impaired student achievement in noisy environments. Green school guidelines should address the design of HVAC systems and walls and doors separating classrooms and corridors and the acoustic quality of windows and walls to the outdoors.
Transportation Noise Sources
The most substantial body of research related to noise and student performance in the classroom examines the impacts of noise from road traffic, trains, and aircraft. Since the 1970s, a number of studies have been conducted that compare the reading skills of students in schools exposed to transportation noise with the reading skills of students in schools in quieter areas. A study in the early 1970s looked at the performance of children in a New York school that was next to the tracks of an elevated train. Over a 3-year period, the aggregate scores of students in grades two, four, and six on the train side of the school were compared with those of students on the nontrain side of the school. Students on the noisy side lagged 3 to 4 months behind in reading compared with students on the quieter side. After the train tracks were treated to abate the noise, read-
ing levels of children on that side of the building improved (Bronzaft and McCarthy, 1975; Bronzaft, 1981).
A 1982 study of students in New York schools under and not under flight paths matched the students for socioeconomic status, race, gender, hearing loss, mother’s education level, and English as a second language (Green et al., 1982). The study found that high levels of environmental noise were inversely related to reading ability in elementary school children. The authors were also able to conclude that the reading deficits were due to chronic noise exposure and not to the noise levels at the time of the test. A later study of students in New York matched students/schools for low socioeconomic status, student absenteeism, and teacher experience and then analyzed reading achievement test scores for grades two through six (Evans and Maxwell, 1997). The analysis found that a higher percentage of students in noisy schools were reading 1 to 2 years below their grade level.
A study of schools near Munich airport looked at the cognitive effects on children before and after the airport was moved to a new location. The study found impaired reading comprehension in third- and fourth-grade children before the move. Children attending schools near the old airport had significantly more errors on a standardized reading test than students from quieter communities. Further, reading comprehension deteriorated in children in schools near the new airport (Hygge et al., 1996).
A 1997 study comparing students from two schools near Heathrow Airport found a significant association between noise and reading comprehension that could not be accounted for by annoyance, social class, or other factors (Haines et al., 2001a,b).
In one of the most comprehensive and rigorous studies to date, Stansfeld et al. (2005) conducted a cross-national, cross-sectional study to assess the effect of exposure to aircraft and road traffic noise on cognitive performance (reading comprehension) and health in children. The study assessed 2,844 children ages 9 and 10 in 89 schools in the United Kingdom, Spain, and the Netherlands in 2002. Schools in all three countries were selected to represent varying levels of exposure to aircraft and traffic noise. The selected schools were matched by students’ socioeconomic status, the primary language spoken at home, and other factors. External noise was measured in decibels, and reading comprehension was assessed using standardized and normalized tests routinely used in each country.
Tests were also conducted to measure students’ recognition and recall (episodic memory), sustained attention, working memory, and prospective memory. Socioeconomic characteristics were assessed as potential confounding factors, and pilot studies were conducted beforehand to assess the feasibility, reliability, validity, and psychometric properties of the cognitive tests to be used. The pooled data gathered through the
study were analyzed statistically using multilevel modeling, and the final results were adjusted for a number of factors, including children’s long-standing illness, parental support for schoolwork, and home ownership. The authors noted the study’s limitations: it was cross-sectional, not longitudinal; restricted to 9- and 10-year-olds; and did not focus on noise exposure in the students’ homes. Moreover, the noise assessment techniques differed from country to country.
This study found that chronic exposure to aircraft noise “was associated with a significant impairment in reading comprehension…. [A] 5-decibel difference in aircraft noise was equivalent to a 2-month reading delay in the United Kingdom and a 1-month delay in the Netherlands” (Stansfeld et al., 2005, p. 1946). This outcome was consistent with findings from other studies on the effects of aircraft noise on reading comprehension. Because it was a cross-sectional study, the effect of long-term noise exposure to aircraft noise could not be measured. Socioeconomic status was not found to be a factor in the size of the effect. The study also found that aircraft noise was “not associated with impairment in working memory, prospective memory, or sustained attention” (Stansfeld et al., 2005, p. 1946).
The authors also looked at the effect of traffic noise on the children. In contrast to the impact of airport noise on reading comprehension, the authors noted “no effects of road traffic noise on reading comprehension, recognition, working memory, prospective memory, and sustained attention (Stansfeld et al., 2005, p. 1947).
More Sensitive Groups
Although excessive noise and reverberation in classrooms was a problem for all children, there were several especially sensitive groups for whom the problems were more acute. These groups included hearing-impaired listeners, second language learners, and children with learning difficulties.
It seems obvious that children with a hearing impairment will experience greater difficulties in conditions with excessive noise and reverberation and this was supported by a number of studies (Finitzo-Hieber and Tillman, 1978; Nábĕlek and Pickett, 1974). In addition, it has been reported (Reichman and Healy, 1983) that a very high proportion of students with attention and/or learning difficulties had significant hearing loss. Clearly, hearing impairment is an additional impediment to learning for children in typical classroom environments. Where a hearing-impaired child is present in a classroom, the rationale is even stronger for controlling noise levels and reverberation. Such a child can also be helped by the use of an amplification system, whereby the teacher uses an FM radio microphone
that transmits directly to the listening child’s hearing aid. This approach is successful because it bypasses the noise and reverberation problems of the classroom.
There are also many children who have varying degrees of fluctuating or temporary hearing impairment. These conditions are frequently due to medical problems such as ear infections or the common cold. Although temporary, while these conditions prevail, younger children, who are already at a disadvantage, will find it even more difficult to understand speech in environments with unwanted noise or reverberation.
Students for whom English is a second language are another more sensitive group that will be more strongly affected by the negative effects of noise and reverberation on their speech recognition capabilities. Nábĕlek and Donahue (1984) found that reverberant conditions (reverberation times of 0.8 and 1.2 seconds) reduced speech recognition scores by 10 percent in listeners for whom English was a second language. However, their use of a monaural headphone presentation technique may have exaggerated the effect. They found that second language listeners who had learned English very early in life were better able to recognize speech in noisy environments than those who had learned it later in life. Those learning English later in life need a 5 dB higher signal-to-noise ratio to perform as well as the early learners of English.
Children with learning difficulties have also been found to experience greater difficulties in understanding speech in noisy environments. Elliott et al. (1979) demonstrated that children with learning problems require louder speech (i.e., higher signal-to-noise ratios) to achieve speech recognition scores similar to those of normally progressing children. They also concluded that this effect was not due to poorer attention or hearing loss in these children. Bradlow et al. (2003) similarly found that children with learning disabilities were more adversely affected by decreased signal-to-noise ratios than were children in a control group without learning disabilities.
In classrooms having a child from one of these more sensitive groups, the rationale is even stronger for reduced noise levels and strict control of classroom reverberation to ensure that all the children can understand the words of the teacher and of their fellow students.
EXCESSIVE NOISE AND TEACHERS’ HEALTH
Teachers who work in noisy classrooms must constantly raise their voices to be heard over other sounds. Over time, speaking in noisy environments can lead to vocal fatigue and other voice problems. More serious effects include the growth of nodes and polyps on the vocal cords: Often, these can be treated by voice therapy, but in some cases they require sur-
gery (Williams and Carding, 2005). A 1993 study found that four out of five teachers who participated in the study reported some problems with vocal fatigue (Gotaas and Starr, 1993). A 1995 study of populations in the U.S. workforce that rely on voice as a primary tool of their trade found that teachers constitute more than 20 percent of the voice-clinic load, or five times the number expected by their prevalence in this segment of the workforce (that is, the segment that relies heavily on the use of their voices) (Titze et al., 1996). Similar results were found by a Swedish study (Fritzell, 1996).
Work by Preciado et al. (1998) is one of the few efforts to attempt to relate vocal disorders in teachers to environmental factors. Voice problems occurred more frequently for teachers of the lowest grades, those in larger classrooms, those with more students, and those in classrooms with higher noise levels. Smith et al. (1998) found that while 20 percent of the teachers had missed work owing to voice problems, only 4 percent of other professionals (nonteachers) had done so. They said their findings suggest that “teachers are at a high risk of disability from voice disorders and that this health problem may have significant work-related and economic effects” (p. 480).
There is clear evidence that noise and reverberation in typical classrooms interfere with children’s ability to understand speech. These problems are not well recognized by those unfamiliar with the many research studies on the topic. They are also not well recognized because adults do not suffer the same communication problems in classrooms as do young children. Further, it is reasonable to suppose that there are connections between impaired communication conditions in classrooms and the many reports of decreased student achievement in noisy situations. These problems can be mitigated by school designs that employ quiet mechanical systems (heating, ventilating, plumbing, etc.) and isolation against noise from adjacent spaces and outdoors. Effective and supportive classroom acoustics can be achieved by paying attention to the construction of a classroom, the sources—external and internal—of noise, and the children’s language and learning needs.
When planning schools, criteria for ambient noise level should be set, designed for, and eventually verified through a commissioning process. Noise criteria can be based on A-weighted sound levels or on several other common measures of noise such as values of noise criterion (NC), room criterion (RC), or balanced noise criterion (BNC). In addition, the desired room reverberation time should be established to achieve room acoustical quality appropriate for speech communication.
Careful HVAC design and installation are needed to meet requirements for background noise from HVAC sources, especially where sound-absorbing duct liners are not being used. Designers should consider using quiet fans, fan silencers, duct vibration “breaks,” and duct systems that deliver air to classrooms at sufficiently low velocities to minimize flow noise. In general, unit ventilators are not nearly quiet enough to meet the ambient noise criteria.
Achieving adequate sound isolation of a classroom against noise transmitted from adjacent spaces is also important. Designing to reduce unwanted sounds from adjacent spaces is done by specifying appropriate sound transmission class (STC) values for the partitions separating a classroom from adjacent spaces. Reducing noise transmission from outdoor sources is equally important and should be achieved through appropriate site selection (e.g., away from persistent transportation noise). In addition, HVAC equipment should be selected to meet acceptable outdoor noise levels and should be located away from classrooms. Windows will be a critical component in controlling the intrusion of outdoor noises.
Specifying the acoustical design criteria requires some special effort. The criteria can most easily be determined by following the recommendations of ANSI Standard 12.60, “Acoustical Performance Criteria, Design Requirements, and Guidelines for Schools.” Tables 6.1 and 6.2 summarize some of the recommendations of this standard.
TABLE 6.1 Required Acoustic Conditions in Classrooms
Recommended Allowable Maximum
Ambient noise level
SOURCE: ANSI S12.60.
TABLE 6.2 Sound Isolation Requirements Between a Classroom and Various Types of Adjacent Spaces
Type of Adjacent Space
STC Requirement for Separating Partitions
aIn the case of outdoor noise, a site-specific design is recommended.
SOURCE: ANSI S12.60.
In addition to the requirements for the various physical components of the classroom and other parts of the school building, it is also very important that teachers appreciate the effects of noise and poor acoustics, how these effects vary with the age of children, and how they can manage their teaching activities to best communicate with the students.
CURRENT GREEN SCHOOL GUIDELINES
Current green school guidelines typically indicate some recognition of the importance of outdoor and indoor noise management in classrooms. They recognize the measurable impact of noise on the academic performance of students and its interference with speech communication in classrooms.
In some cases, green school guidelines call for the reduction of background sound levels to NC35 or less, to NC30. However, only NC30 approximately meets the requirements of ANSI S12.60, which is intended to ensure that a teacher’s voice is clearly understood by younger children against a background of other local noise-generating equipment or activities.
Some green school guidelines recommend meeting ANSI S12.60 standards for sound isolation between classrooms and adjacent spaces so that noise from these spaces does not compromise the quiet background sound levels needed for a teacher to be clearly understood.
Current green school guidelines do not fully address controls on outside noise generation. Given the results of research on the impact of traffic and airplane noise on student performance, these sources and the selection of school sites to avoid loud outdoor sound need to be considered.
FINDINGS AND RECOMMENDATIONS
Finding 6a: Most learning activities in school classrooms involve speaking and listening as the primary communication modes. The intelligibility of speech in classrooms is related to the levels of speech sounds relative to the levels of ambient noise and to the amount of reverberation in a room.
Finding 6b: Sufficient scientific evidence exists to conclude that there is an inverse association between excessive noise levels in schools and student learning.
Finding 6c: The impacts of excessive noise vary according to the age of students, because the ability to focus on speech sounds is a developmental skill that does not mature until about the ages of 13 to 15 years. Thus, younger children require quieter and less reverberant conditions than do
adults to hear equally well. As adults, teachers may not appreciate the additional problems that excessive noise creates for younger students.
Finding 6d: Excessive noise is typically a more significant problem than is too much reverberation in a classroom. It is not possible to have both increased speech level (to maximize signal to noise) and reduced reverberation times. Good acoustical design must be a compromise that strives to increase speech levels without introducing excessive reverberation.
Finding 6e: The most substantial body of research related to excessive noise and learning in the classroom addresses the impacts of road traffic, trains, and airport noise.
Finding 6f: Some available evidence indicates that teachers may be subject to voice impairment as a result of prolonged talking in noisy school environments. However, there is no information to quantify a relationship between specific noise levels in classrooms and potential voice impairment.
Recommendation 6a: To facilitate student learning, future green school guidelines should require that new schools be located away from areas of higher outdoor noise such as that from aircraft, trains, and road traffic.
Recommendation 6b: Future green school guidelines should specify acceptable acoustical conditions for classrooms and should require the appropriate design of HVAC systems, the design of walls and doors separating classrooms and corridors, and the acoustic quality of windows and walls adjoining the outdoors. This recommendation is most easily achieved by requiring that green schools comply with American National Standards Institute (ANSI) Standard 12.60, “Acoustical Performance Criteria, Design Requirements, and Guidelines for Schools.”
Recommendation 6c: Additional research should be conducted to define optimum classroom reverberation times more precisely for children of various ages.