Much of the quite extensive research evidence relating to the issue of noise in education has been produced by studies of the sort mentioned above, focussed on relating noise levels to particular outcomes. Notably, the results, using both experimental and observational methodologies, are remarkably consistent. The findings show that noisy conditions have direct negative effects on learning, particularly language and reading development, as well as causing indirect problems to learners through distracting or annoying them. Laboratory based cognitive psychology experiments have shown that noise affects performance on memory tasks [7,8], an effect which is at least partly explained by noise interfering with language processing. This suggests that it might be problematic to live in a generally noisy place, and real world research into the effects of chronic noise, experienced by people living near airports or busy roads, confirms this extrapolation [9]. A review of this area concluded that, ‘The evidence for effects of environmental noise on health is strongest for annoyance, sleep and cognitive performance in adults and children.’ [10, p. 253, see also 11].

Of particular concern for education is the reliable finding, which fits with the laboratory results, that chronic noise exposure impairs cognitive functioning [see 12 for an overview]. Studies have found associations between noisy environments and reading problems [13–15], deficiencies in pre-reading skills [16] and more general cognitive deficits [17]. There is the implication that in addition to interfering with processing on each occasion, the environmental noise may be contributing to developmental problems, particularly with speech and language and with reading. These studies imply that either living in a home or attending a school near a source of on-going noise will increase the likelihood of a child having educational difficulties. Clearly those attending neighbourhood schools are likely to experience raised noise levels from the same source at both home and school. Shield and Dockrell make this point in their review of research concerning noise experienced by school students [12, p. 102], also noting that many such children suffer other problems of deprivation, presumably because noisy residential areas are not desirable places to live.

Recent research, however, into the problem of noise in the school environment has tended to centre on the potentially more widespread problem of students struggling to learn because of noise generated within the school itself. In their surveys of external and internal noise in urban primary schools, Shield and Dockrell comment that ‘during lessons it was normally not possible to hear external noises’ [18, p. 734] and the noise levels they report just outside these urban schools were frequently lower than might have been expected due to schools being located in side streets or separated from roads by playgrounds.

Possibly problematic noise comes from several sources, as demonstrated by studies where measurements of classroom noise levels were made [18–21]. Firstly, there is noise intrusion, which can come from outside the school, but seems to be increased more by noise from other parts of the school leaking into the classroom. Secondly, there is background noise from within the room, often due to heating and ventilation systems but also caused by equipment such as projectors and computers. Finally there is the noise generated by students engaged in learning, which, unsurprisingly, varies according to the nature of the activity [18, p. 735]. Noise levels are argued to be exacerbated by high levels of reverberation, since this can increase the noise level itself and make the hearing of speech more difficult [20, p. 7]. The consistent findings, across laboratory and field studies, of noise interfering with language based tasks together with concerns about speech intelligibility in the classroom has suggested to some that controlling reverberation might be as important as reducing noise levels. Indeed this is an argument that some researchers advance [19,20], although other work is more equivocal about reverberation, tending instead to emphasise noise levels and signal to noise ratios in discussions of speech intelligibility [12], standards for school building adopted or advocated always include guideline maximums for both noise levels and reverberation [12,22–24].

Potentially linking both external and classroom noise directly with cognitive processing deficits, there have been some experimental studies which have investigated how well children perform various learning tasks when exposed to recorded real world noises played at a range of levels. Dockrell and Shield [25] used recordings of classroom ‘babble’ alone and together with environmental noises such as sirens and road noise. These noises were played at a level to match average ambient noise levels recorded in classrooms, and tended to impair performance on a range of literacy, numeracy and speed tests. There was, however, complexity in the results relating to the impacts of the differing sorts of noise on the processes underlying performance on the various tests.

Hygge [26] conducted a series of experiments where school children read texts in either noisy or quiet conditions, with the noise provided by recordings of aircraft, train and road noise or irrelevant speech. Tests conducted a week later in quiet conditions showed deficits in memory for these texts when they were studied in noisy conditions, with this being more pronounced for recall, as opposed to recognition, of information. Interestingly, this deficiency in learning continued from session to session when children had experienced noise first, associated with motivation for the task dropping off as a result of the noise experience.

Thus, these sorts of experiments demonstrate not only that children generally have more difficulty performing cognitive tasks when it is noisy, but also suggest that noise tends to undermine long term learning, corroborating the findings of the observational studies of chronic environmental noise. Yet, in both the experimental studies discussed above there were complications in the patterns of results which leave uncertainties about precisely how noise interferes with particular cognitive processes relevant to learning.

Other research has been conducted that takes a somewhat wider view of noise, sometimes also attempting to understand the mechanisms for the effects reviewed above. As Hygge and Knez point out, there is “theoretical and practical value in knowing how the physical parameters of the indoor environment may combine or interact” [27, p. 291], but very little work puts this into practice. Their own studies of noise in combination with heat and lighting [27,28] are an exception to this, but have failed to clarify when and to what extent problems due to these other physical factors interact with deficiencies caused by noise.

An important interaction with absolute noise levels are the attempts that teachers and learners make to accommodate or compensate for noisy surroundings. Early research observed that teachers pausing during bursts of external noise leads effectively to a reduction in teaching time, sometimes estimating this loss to be as high as 11% of teaching time [15]. More recently, researchers in the US have found evidence of voice fatigue being a particular health concern for teachers [29], and this is identified as another reason to reduce classroom noise levels [20, p.3,23]. Turning now to the actions of the learners in noisy situations, it has been suggested that apparently conflicting experimental results may sometimes be due to the positive effect of arousal [27]. In discussing the results of their noise experiment, Dockrell and Shield argue that children in the environmental noise condition may have been provoked ‘to actively focus on the task, possibly by redirecting attention’ [25, p. 522], with this heightened attention producing better than expected performance. From these experimental studies, however, it is not possible to know the limits of such improved performance over time, different tasks and noises, or across differing learners.

Much work does not differentiate between individuals in their reactions to noise or between different sorts of noises. Where individual differences are investigated, there is a body of evidence that some individuals might be more sensitive to noise than others [30,31]. The more immediate problem for school design in terms of inclusion is the finding from some studies that the performance of students with Special Educational Needs (SEN) might be more susceptible to noise [25], although impacts on particular children may not be predictable [32,33]. Also, learners for whom the language spoken at school is an additional, rather than first, language may be further disadvantaged by noise in the classroom [11]. Nelson et al. [20] review research evidence showing that such children perform less well than native speakers on tests of speech comprehension against noise. On the written tasks used by Dockrell and Shield [25], however, the performance pattern of the children with English as an additional language (EAL) did not suggest that they suffered additional difficulties, probably because the tasks did not depend on speech intelligibility. Considering different types of noise, Hygge [26] presents evidence of aircraft noise being particularly disruptive to learning, traffic noise less disruptive but still impairing performance, while train noise did not cause a problem. Yet, confusingly, this study also found no effect for verbal noise, despite Knez and Hygge [27] arguing that irrelevant speech is a particularly distracting noise.

These apparent contradictions may be explained by research findings that factors such as predictability, control and judged necessity influence how annoying people find particular noises [10,34], although the relationship of noise annoyance and performance deficits is far from clear. There is interest in noise annoyance [34,35] and links to mood [36,37], but Hygge [26] argued that interference with the encoding stage of memory which the noise conditions seemed to cause in his experiment was not mediated by distraction or mood. Related to this, research by Dockrell and Shield [38] into the perceptions of primary school children regarding actual environmental noise showed that annoyance may not be related in a straightforward manner to either noise level or frequency of occurrence.

Within mainstream education literature there appears to be more awareness of noise generated within schools, rather than externally, and the impact is usually considered through investigations of perceptions of this noise and any annoyance or sense of distraction. For example, Flutter has spent more than a decade involved in projects where students are consulted about their learning and has found that a central concern is with excessive noise: ‘Noise and distracting behaviour were the most frequently mentioned problems and many students said they would like a calmer and quieter environment’ [39, p. 184]. In answer to this, a book about effective teaching argues that a whole school policy addressing corridor noise levels has the potential to improve ‘classroom climate’, making it more conducive for learning [40, p. 61]. Recent rebuilding and refurbishment programmes within the UK [41,42] have also alerted some educators to the contribution of qualities of the school building itself to experiences of noise. In English secondary schools awaiting Building Schools for the Future (BSF) rebuilding work, problems with acoustics are considered by over three-quarters of head teachers to be affecting to ‘a large extent’ or ‘to some extent’ the school’s provision of education [41, p. 65]. Such reports of inadequate acoustics are likely to be broadly accurate since these schools will have been constructed according to less stringent guidelines, which were not then legally compelling [see 12, pp. 108–109]. Furthermore, many of these schools were built in the second half of the twentieth century using lightweight construction methods with inadequate sound insulation. Since similar designs and construction methods extend across the sectors of education and beyond the UK, this is not a problem specific to UK secondary schools although the BSF programme has highlighted it here.

As stated at the beginning of this section, the research evidence relating to noise is remarkable for its broad consistencies. Noise interferes with learning, through direct effects on information processing, and via indirect effects on teachers, learners and communication in the classroom. The variety of these routes suggest why noise is a serious problem for schools but also why the causal mechanisms proposed, principally interference with speech intelligibility, distraction and annoyance are each unlikely to provide complete explanations of how noise interferes with learning. It is clear that high background noise levels and poor acoustics decrease speech intelligibility but educational activities are not all equally dependent on understanding speech, as will be discussed in more detail below. Although noises may (sometimes) annoy people, or cause distractions that could impair learning, neither of these causal links seems completely necessary for problems with learning, and are likely to be only part of the problem, or perhaps are only significant in some cases.

The range of noise sources and the various ways in which noise can impair learning, which the research base, and the differing interpretations of its findings, illuminate are indicative of a serious problem for schools. Yet the nature of this research also suggests the difficulty of enacting solutions to these noise problems. Firstly, it is clear that different solutions will be required depending on the main source of the noise. Concerns about external environmental noises affecting health, wellbeing and learning in school have implications for the location and design of the school structure. This issue is discussed, for example, in the World Health Organisation (WHO) guidelines, where recommendations for planning bodies and architects are laid out [23, p. 10, p. 19]. Noise leaking between rooms in a school can similarly be understood as a problem of the design and construction of the premises, while background noise due to heating and ventilation machinery suggests a need for quieter models [20, p. 8]. Across these noise sources, however, the emphasis of the solutions will depend on judgments made about the main mechanism for interference with learning. This is shown in relation to the proposal that teachers’ voices should be amplified [43]. This makes sense if the impact of the signal to noise ratio on speech intelligibility is considered central, but may be disputed if absolute noise levels are argued to be more important [20, p. 3].

Of distinct importance in identifying solutions to noise problems is the recognition that in educational settings, as in other real world situations, noise is not the isolated factor which it sometimes appears to be in the context of experimental and observational studies focused on noise. This means that proposed solutions for improving acoustic elements of the school environment might have detrimental effects on other aspects, while changes made to the physical setting to enhance other factors might worsen noise problems. A particular tension is evident between noise and air quality. Some suggested acoustics improvements, such as carpets [21] or ceiling hangings [16] may collect dust and so worsen air quality, while excessive background noise in classrooms is often due to open windows [23, p. 8] or mechanical ventilation units [19]. Yet research demonstrating the impact of poor air quality on health, absence from school [44], and therefore on learning, suggests that good ventilation is vital in a classroom, a conclusion that reviews of the physical education environment tend to reach [45–49].

Other interactions of noise with various environmental factors, and the potential for solutions to conflict, are evident from considering studies in these separately researched areas. User satisfaction with buildings has been found to be linked to the control of temperature [50, p. 133], but this again suggests teachers opening windows and turning on heaters, with the associated implications for classroom noise levels. There is also research suggesting that more spacious classrooms result in more positive feelings amongst students [51] and, very specifically, that higher ceilings in classrooms decrease perceptions of crowding, and increase teacher satisfaction [52]. Yet larger spaces present more acoustic challenges, particularly in relation to reverberation. Measurements reported by Knecht et al showed that ‘rooms with the longest reverberation times were those with the largest volumes’, and the classrooms with lower ceilings were more likely to have ‘acceptable’ reverberation times [19, p. 69]. The inevitable difficulties of resolving these tensions and interactions satisfactorily are very rarely suggested by the building guidelines relating to noise. A notable exception to this omission is the discussion in the WHO pamphlet about the need for school canteens to be both hygienic and acoustically satisfactory, yet ‘the materials available rarely meet the two criteria simultaneously’ [23, p. 15]. The authors then go on to make practical suggestions, such as the use of tablecloths and rubber pads on chair legs, which will reduce noise levels without compromising hygiene.

Adding another level of complexity to the competing demands of the physical environment is the fact that inhabitants of a setting are not passive recipients, and their actions will interact with aspects of the surroundings, sometimes in unpredictable or complicated ways. This has already been suggested by descriptions above of teachers pausing during occasions of external noise and suggestions from some noise experiments that potentially distracting noise may sometimes enhance performance as people make more effort to concentrate. Other instances of changed behaviour due to noise might be more predictable, but are no less problematic. The WHO guidelines describe the ‘vicious circle’ that occurs when inadequate acoustics force speakers to raise their voices, resulting in more noise and so yet louder voices [23, p. 9].

These examples indicate the potential for more extended modifications to behaviour to have pronounced effects on the learning environment, and indeed there is some research evidence of this occurring. In the 1960s and 1970s when many new schools in the US and UK were built around open plan principles, some investigations of teaching practices in these new settings found that teachers were being more, rather than less, conservative in their methods. Fears of disturbing other classes made them less likely to encourage active, and so more noisy, independent learning than were teachers in traditional cellular classrooms [52,53]. Similarly, some research into behaviour in open plan offices shows such settings can discourage, rather than facilitate, social interactions, as people are concerned about disturbing their colleagues [54]. These observations suggest that sometimes in schools an organisational approach to noise nuisance may be appropriate so that a peaceful working environment can be maintained without the loss of other school activities that generate more noise. One of the WHO recommendations asserts that schools ought to ‘adapt their timetables in order to set aside certain periods of the day and areas of the school for noisy activities’ [23, p. 19].

It is worth remembering, however, in connection with organising and managing a school community to control noise that such institutional management necessitates value judgments being made about different sorts of noise. Also, in this context, it is worth noting the considerable disagreement within the noise literature about the importance of conscious noise annoyance and the range of factors which appear to affect individual judgments [e.g., see 38, p. 2965]. Although these judgments, both those made by individuals and those reached at an institutional level, are undoubtedly more true to the complexity of human experience of noise than purely quantitative measurements of noise level, they do leave more room for argument. For example, in a recent paper a British education researcher argues that existing structures in schools result in ‘weak silences’, consisting of imposed quiet which disempowers students through silencing them, and an absence of ‘strong silences’, when non-coercive quiet allows concentration, gaze and self-understanding [55]. This suggestion that there exist bad sorts of quiet in school is balanced by understandings, common among educationalists, that there may be good noise. The classroom activities which Shield and Dockrell found to be noisiest [18, p. 735], ‘group work’ and ‘group work and movement’, are considered by many educators to be desirable, through facilitating learners’ engagement and collaboration. Furthermore, theorists about educational practices often argue for the centrality of dialogue, both between teacher and learner and between learners, to successful learning [56,57].