Sage Journals: Discover world-class research

Abstract

Acoustic comfort is a concept hardly described in the literature. But it has been used in engineering typically to refer to low noise or annoyance in order to invoke no discomfort. Current standardized methods for airborne and impact sound reduction are deployed to assess acoustic comfort in dwellings. However, the measured sound pressure levels do not represent comfort. The latter should include further the human perception of the acoustic environment. Therefore, this article reviews studies that approached acoustic comfort through the association of objective and subjective field data, combining in situ acoustic measurements and survey responses from residents. We evaluated the studies using Bradford Hill’s criteria. Most researches focus on self-reported noise annoyance while some others on satisfaction responses. Many studies were found incomprehensibly described: often vital data of statistical evaluation or study design are lacking. The results indicate that noise is a significant issue in living environments, especially certain impact noise types. The use of extended low-frequency spectra down to 50 Hz was suggested for impact measurements in order to predict better self-reported noise response. Greater problems with low-frequency transmission are displayed in lightweight structures which perform inefficiently compared to heavyweight components. Harmonization of presented results and study design details should be taken into account for future articles.

Keywords

Acoustic comfort field measurements subjective responses association indicators

Introduction

This article concerns a review of acoustic comfort evaluation during field studies in residencies, which include acoustic measurements of building structures and surveys or interviews with residents in their actual living environment. The scope of this review is to collect and examine those studies which combine acoustic data and subjective responses in order to approach acoustic comfort.

Despite being an important concept in engineering, acoustic comfort is vaguely defined and explored in the literature. So far, the term has been used in a general sense by engineers and designers, usually to refer to conditions with little noise and disturbances in a certain space. However, most publications do not offer a concept description, even when they use the term “acoustic comfort” or quality in their title.¹

A first definition in the literature is provided by Rindel² and then repeated in some following articles.^3–5 The description offered for acoustic comfort is “a concept that can be characterized by absence of unwanted sound and opportunities for acoustic activities without annoying other people.” This definition offers a user’s perspective rather than a relation to merely acoustic measured data: acoustic comfort, for a certain person, is a combination of the person as a receiver of sound as well as a source. This means that a person can be disturbed by his or her own sounds because the sounds are truly disturbing or just because others might be disturbed, and dissatisfaction or conflicts might arise.

Past research has shown that measurements and metrics that acousticians use in order to assess building acoustic conditions may not always be representative of how residents perceive acoustics in their living environment. For example, tenants might have problems with impact noise types or vibration transmission from neighboring flats in the low-frequency range that is partially omitted from measurement spectra.⁶

Developments in the construction industry, such as the use of wood as a building material, create a demand for higher standards to be met in dwellings. Various regulations exist in several countries to assess sound insulation issues from noise inside or outside a building.⁵ Residents still report complaints about noise from neighbors, outside road traffic, indoor technical installations, or other sources.^6–15 A central concern is thus how well the perception of residents corresponds to the results acquired by acoustic measurements and the descriptors of sound insulation in buildings. The latter are defined in a list of related standards, and variations of these are sometimes proposed in order to achieve better levels of agreement. Statistical methods have been used to examine how well building acoustic descriptors correlate to the subjective ratings of tenants, in field or laboratory studies. If they do, it is possible to formulate models for prediction of satisfaction and comfort for the building users.

For all the above reasons, sound perception and noise annoyance issues remain popular. However, the available research results usually come from small studies and remain insufficient (small samples) suggesting a demand for further research. Consequently, the idea of associating and comparing data of human sound perception to technical acoustic data seems essential. This article provides a systematic review in the association of subjective responses and acoustic data, in field studies in buildings. The overall purpose for examining this association is to evaluate, simulate, and maybe predict the response of residents and approach the concept of acoustic comfort.

Methods

The following databases were used to search for peer-reviewed publications and conference proceedings, offering investigation or comparisons between acoustic field data and subjective responses relevant to building acoustic studies: ScienceDirect, AIP Scitation, Ingenta Connect, ResearchGate, PubMed, Scopus, and Google Scholar. Several keywords in different orders and combinations were used, such as objective, subjective, acoustic, psychoacoustic, self-report, rating, score, comfort, quality, airborne, impact, sound, insulation, noise, annoyance, assessment, association, correlation, evaluation, comparison, building, and dwellings. Many times, the searches did not return a useful outcome for the scope of this review. Then, some studies were found in the references of the relevant publications, which were found initially.

Review criteria

Requirements for inclusion of articles in this review were the comparison of results between field acoustic data and subjective responses in the actual living environments as well as the use of statistical methods for the association of those data. The subjective data are obtained from residents with questionnaire surveys or interviews. In the end, 50 articles were found during the search in databases or relevant references; 24 of them were decided to be included in this review,^1–24 which correspond to 10 complete studies. The excluded studies^25–51 concerned mostly laboratory studies and not field studies.

The exclusion criteria, besides article context, were the year of publication and language. A threshold was set to 1985, because the earlier research studies found were few and very limited in results. Also, few of them were national publications, written not in English but in German, for example, so we could not translate and analyze them properly. The review search took place from April 2015 until September 2017.

Summary of methods, metrics, and quantities in the reviewed studies

Many different indicators (or descriptors) have been used to represent different quantities in acoustic measurements. They are all standardized in international ISO or other national standards, which usually comply with ISO. Many variations of them exist as well, since experimental research has been done to acquire better indicators than the standardized ones. A tabulation of all indicators presented in this review as well as the standards in which they are defined is presented in Table 1.

Table 1.

Descriptions of the acoustic indicators used in the reviewed studies.

Indicator	Description	Standards	References
${R^{'}}_{w}$	Apparent airborne sound reduction index (same as $R_{w}$ , for field measurements)	ISO 717-1, ISO 140-4, EN ISO 12354-1, ISO 16283-1	6 –13, 24, 52 –55
$S T C$	Airborne sound transmission class, calculated similar to $R_{w}$	ASTM E413	16 –19
$D_{n T, w}$	Apparent standardized level difference index	EN ISO 12354-1	14, 52, 53
${L^{'}}_{n, w}$	Weighted standardized impact sound pressure level	ISO 717-2, ISO 140-7, EN ISO 12354-2, ISO 16283-2	6 –18, 20 –24, 52 –55
$C$	C is an A-weighted pink noise spectrum adaptation term	ISO 717-1, -2, EN ISO 12354-1, -2	6 –12
$C_{50 - 3150}$	$C$ adaptation terms, frequency range 50–3150 Hz	Same as $C$	6, 14, 24
$C_{I, 20 - 2500}$	$C$ adaptation terms, frequency range 20–2500 Hz	Same as $C$	6 –12
$C_{I, A k u L i t e, 20 - 2500}$	$C$ adaptation terms, frequency range 20–2500 Hz	Same as $C$	6 –12
$L_{A F m a x}$	The Japanese rubber ball impactor index	JIS A 1418-2	6, 13, 56

For the calculation of indicators such as ${R^{'}}_{w}$ or ${L^{'}}_{n T, w}$ , after the actual airborne or impact sound measurements, a predefined reference curve in the 1/3 octave spectra is used for comparison. The latter curve is shifted higher or lower with steps of 1 dB until the sum of deviations (between the measured and the reference curve) is maximum without exceeding 32 dB. Then, the value of the shifted reference contour at 500 Hz is used as the single number indicator of the measured building component, wall or floor.^57–60 Detailed analysis and review of the descriptors for airborne and impact sound insulation can be found in the existing literature, with a complete study of European indicators and comparison with suggested values in national regulations presented by Rasmussen and Rindel.⁵ Note that the sound transmission class index $S T C$ used in the US standards is actually a reduction index which is calculated similar to the airborne sound reduction index $R_{w}$ .

Several statistical methods are used in the studies, by means of statistical correlations and regression analyses, which associate acoustic data and subjective responses. The quality of statistical association is described with typical parameters:

The correlation coefficient, denoted as $r$ , $ρ$ , or $R$ , is a measure of the strength and the direction (slope) or the linear relationship between two variables, taking values $0 < | r | < 1$ . If r is positive, the linear relation is positive and upward, that is, the higher value for the independent variable X means the higher value for the dependent variable Y. If r is negative, the opposite occurs, that is, higher X means a lower dependent response Y. A coefficient of value 1 describes a perfect positive linear relationship between the data.

The coefficient of determination, denoted as $R^{2}$ , is the squared correlation coefficient in the case of simple regression with one dependent variable. It is a measure of how well the regression line model represents the data. It is defined as the ratio of the explained variation to the total variation, so it indicates the percentage of variables positioned closest to the fitting line in terms of statistical significance. Both coefficients take values between zero and one ( $0 ⩽ | r | ⩽ 1$ ) and can be also expressed as a percentage, for example, r = 85% instead of 0.85. In this review, the correlation as percentage is preferred, to avoid confusions between the positive or negative slopes in different cases of airborne or impact sound insulation which differ.

The p-value and the confidence intervals (CIs) are measures of statistical significance, meaning the probability for the real result (which we approach with statistical methods) to be different than the observed, that is, the outcome of the statistics.

Evaluation of included studies

The quality of evidence for studies in this review was evaluated using Bradford Hill’s criteria,⁶¹ which is an evidence classification method often used in epidemiology and health studies. The fulfilled criteria are rated in a scale of high (+++), moderate (++), and low (+). The results are tabulated in Table 2, while the criteria used for evaluation are as follows:

Strength of association: it refers to the causality proven by the association between the studied variables (cause, effect size, and confounding factors).

Consistency: indicates the degree of certainty when similar results are observed by different studies in different tests.

Specificity: specific factors and effects on a specific population lead to a more likely causal relationship.

Temporality: it is based on temporal relations between effects and used as an indicator for causality, meaning one effect occurring after an exposure.

Biological gradient: it refers to the relation between exposure and effect; usually bigger exposure leads to greater effect, but not always, while the opposite outcome can occur as well.

Plausibility: it means that a biological explanation of why a cause leads to a certain effect supports a reasonable causality.

Coherence: it is a condition meaning that a stated causal relationship should not contradict with other accepted results or knowledge.

Experiment: it refers to the study design parameters that guarantee a reasonable causation, such as randomization.

Analogy: the possibility of having or predicting analogous effects from similar factors without total evidence.

Publication type: an additional criterion that we added in order to rank the reviewed studies. Scientific journal articles are thoroughly peer reviewed, while conference articles are usually less well reviewed. There are also study reports from research organizations that may be scientifically conducted but not reviewed. Thus, publications were evaluated as follows: scientific journal, +++; conference article, ++; and technical report, +.

Table 2.

Evaluation of the presented studies according to selected criteria.

References	Publication type	Strength of association	Consistency	Specificity	Temporality	Biological gradient	Plausibility	Coherence	Experiment design	Analogy
1	++	++	+	+	+	++	+	+	+	+

6	+++	+++	+++	++	++	++	++	++	++	++
7	++	++	++	++	++	++	++	++	++	++
8	++	++	++	++	++	++	++	++	++	++
9	++	++	++	++	++	++	++	++	++	++
10	+	++	++	++	++	++	++	++	++	++

11	+++	++	++	++	++	++	++	++	++	++
12	++	++	++	++	++	++	++	++	++	++

13	++	+	+	+	+	+	+	+	+	+

14	+++	+++	++	++	++	++	++	++	++	++
15	++	++	++	++	++	++	++	++	++	++

16	+++	+++	++	++	++	++	++	++	++	++

17	++	+++	+++	++	++	+++	+++	+++	++	++
18	+++	+++	+++	++	++	+++	+++	+++	++	++

20	+	++	++	+	+	++	++	++	+	+
21	++	++	++	+	+	++	++	++	+	+

22	+++	+++	+++	++	++	+++	+++	+++	+++	++

23	+	+++	++	++	++	+++	+++	+++	+++	++

24	+++	++	++	++	++	+++	++	++	++	++

++: scientific journal; ++: conference article; +: report.

The references are grouped according to the research study they present:^{1,6–10,11–12,13,14,15,20,21–23,24,25,26,27,28.}

The evaluation of the included studies was conducted by the authors. The presented data were chosen according to their relation and importance for this review’s context. In Table 3, an overview is presented for all the selected studies, which are tabulated with important details on study design, variables, and summary of results. The studies are analyzed in the next section. In Table 2, the evidence evaluation rating of the studies is presented according to Bradford Hill’s criteria,⁶¹ which were chosen because they focus on causation between exposure and effects. Other evaluation such as the GRADE approach⁶² was not preferred as it focuses a lot on the study design; in building acoustics, most studies are cross-sectional or experimental, so they would be rated very low in such a case. Readers who would like to have a deeper insight in any specific study results, and conclusions may read the original publications using the references.

Table 3.

Overview of the selected studies.

References	Sample and experiment details	Variables as defined in the study	Models and results	Parameters for evaluation	Summary of results
1	Synthesis of results from building surveys and measurements from other complete studies^18,19,61	$P$ : poor evaluation $G$ : good evaluation $F$ : fair evaluation by the residents	Exposure–effect curves and models not stated. Suggested relation: $P + F + G = 100 %$	No details provided	Dose–response curves have an average slope of 4% in all studied cases
6 –10	251 respondents to survey 10 Swedish multistory buildings for field measurements 2 HW structure 8 LW constructions Results in acoustic measurements and subjective responses were averaged per test building	Independent: ${R^{'}}_{w} + C_{50 - 3150}$ : weighted normalized airborne reduction ${L^{'}}_{n T, w}$ : weighted standardized impact sound Dependent: V1: mean annoyance toward airborne sounds from neighbors V2: mean annoyance toward impact sounds from neighbors above in a 11-point scale	Linear regression: ${R^{'}}_{w}$ to V1 ${R^{'}}_{w} + C_{50 - 3150}$ to V1 ${L^{'}}_{n T, w}$ ${L^{'}}_{n T, w} + C_{50 - 3150}$ ${L^{'}}_{n T, w} + C_{I, A k u L i t e, 20 - 2500}$	Coefficients: $r = 76 %, R^{2} = 0.58$ $r = 85 %, R^{2} = 0.73$ $r = 51 %, R^{2} = 0.26$ $r = 57 %, R^{2} = 0.32$ $r = 92 %, R^{2} = 0.85$	Low-frequency spectra down to 20 Hz are essential for impact sound measurements and correlation to annoyance. Annoyance predicted well with descriptor ${R^{'}}_{w} + C_{50 - 3150}$ , not well with ${L^{'}}_{n, w}$ but sufficiently with suggested descriptor: ${L^{'}}_{n, w} + C_{I, 20 - 2500}$
11, 12	Continuation of previous study^6–10 800 extra respondents to survey 13 Swedish multistory buildings for field measurements, so 23 in total 6 HW 6 cross-laminated timber 11 LW constructions Results in acoustic measurements and subjective responses were averaged per test building	Independent: ${L^{'}}_{n T, w}$ : weighted standardized impact sound pressure level Dependent: V2: mean annoyance toward impact sounds from neighbors above in a 11-point scale	Linear regression: ${L^{'}}_{n T, w}$ ${L^{'}}_{n T, w} + C_{50 - 2500}$ ${L^{'}}_{n T, w} + C_{20 - 2500}$ ${L^{'}}_{n T, w} + C_{I, A k u L i t e, 20 - 2500}$	Coefficients: $r = 42 %, R^{2} = 0.18$ $r = 70 %, R^{2} = 0.49$ $r = 84 %, R^{2} = 0.71$ $r = 80 %, R^{2} = 0.65$	Results from previous study confirmed, inclusion down to 20 Hz is essential for impact sound measurement to correlate well with self-reported annoyance. The low-frequency inclusion does not affect much the results from HW constructions in contrast to the LW ones
13	Unknown number of respondents to survey 10 French multistory or detached buildings	Independent: ${L^{'}}_{n T, w}$ : weighted standardized impact sound pressure level Dependent: V2: mean annoyance to impact sounds from neighbors above in a 11-point scale	Linear regression: ${L^{'}}_{n T, w} + C_{I, 50 - 2500}$ to V2 $L_{A F m a x}$ to V2 ${L^{'}}_{n T, w}$ ${L^{'}}_{n T, w} + C_{I}$	Coefficients: $r = 89 %, R^{2} = 0.79$ $r = 85 %, R^{2} = 0.73$ $r = 86 %, R^{2} = 0.74$ $r = 85 %, R^{2} = 0.73$	${L^{'}}_{n T, w} + C_{I, 50 - 2500}$ correlated best to ratings. Suggestions: ${L^{'}}_{n T, w} + C_{I, 50 - 2500} = 52 dB$ and $L_{A F m a x} = 54 dB (A)$
14, 15	600 Norwegian dwellings Nearly 720 respondents: 83% in concrete slab constructions and 17% in LW constructions	Independent: ${L^{'}}_{n, w}$ , ${L^{'}}_{n T, w}$ : weighted normalized and standardized impact sound pressure level Dependent: V1: annoyance to airborne sounds from neighbors V2: annoyance to impact sounds from neighbors above in a 5-order scale	Exposure–effect curves used but regression models not stated $D_{n T, w}$ offers best correlation with V1 ${L^{'}}_{n, w} + C_{i, 50 - 2500}$ correlates very well with V2 ${L^{'}}_{n T, w} + C_{I, 50 - 2500}$ correlates best with V2	Curves within 95% CI in all cases	Noise is a big issue for residents, especially low-frequency problems. ${L^{'}}_{n, w} and {L^{'}}_{n T, w}$ do not correlate with V2 without correction terms
16	398 participants (350 flats) in Sweden 22 floors, concrete, or timber joist were measured. Results in acoustic measurements and subjective responses were averaged per building block in the study	Independent: $I_{i}$ : old-type impact sound index ${L^{'}}_{n, w}$ , ${L^{'}}_{n, A}$ : weighted and A-weighted normalized impact sound pressure level Dependent: $S$ : mean satisfaction response of tenants in a building block from 1 (quite unsatisfactory) to 7 (quite satisfactory)	Linear regression: $\bar{I} = 86.3 - 5.48 S$ ${\bar{L^{'}}}_{n, w} = 80.6 - 5.48 S$ ${\bar{L^{'}}}_{n, A} = 85.2 - 5.09 S$ Suggested: ${\bar{I}}_{S} = 86.3 - 5.53 S$ $S = 4.4$ corresponds to 51% of the resident sample	Coefficients: $r = 73 %, R^{2} = 0.53$ $r = 75 %, R^{2} = 0.56$ $r = 72 %, R^{2} = 0.52$ $r = 87 %, R^{2} = 0.76$	Good association exists between impact noise and subjective response. Low frequencies from 50 Hz should be considered in the measurement spectra. New reference curve suggested for ISO standard method correction to have better correlation levels
17, 18	600 participants interviewed, 300 party walls measured in Canada. Results in acoustic measurements and subjective responses were clustered in 8 groups according to $S T C$ values	Independent: $S T C$ : weighted airborne sound reduction index, similar to ${R^{'}}_{w}$ Dependent: V1: want to move out? V2: satisfaction with building V3: considerate neighbors V4: neighbors’ noise either side V5: neighbors’ voices V6: neighbors’ TV sound V7: neighbors’ music V8: awakening from neighbors’ noise V9: sound insulation satisfaction	Linear regression of $S T C$ with: V1: yes to move out V2: satisfaction for house V3: neighbors consideration V4: neighbors’ noise general V5: neighbors’ voice sounds V6: neighbors’ TV sound V7: neighbors’ music V8: sleep awakening V9: sound insulation rating (linear and Boltzmann equation models not stated)	Coefficients: $R^{2} = 0.56, p = 0.033$ $R^{2} = 0.83, p = 0.002$ $R^{2} = 0.86, p = 0.001$ $R^{2} = 0.82, p < 0.002$ $R^{2} = 0.94, p < 0.001$ $R^{2} = 0.77, p < 0.004$ $R^{2} = 0.92, p < 0.001$ $R^{2} = 0.60, p = 0.024$ $R^{2} = 0.92, p < 0.001$	Suggested value STC = 60 dB would solve most annoyance problems. If STC = 50 dB, then annoyance from most noise types decreases significantly; above that level, there is some important. protection from music sounds. If STC = 55 dB, then circa 10% of the subjects are disturbed by general noise from neighbors
20, 21	Synthesis of results from building surveys and measurements from the original and another study⁶	Independent: $I_{i}$ : old impact sound index⁶ ${L^{'}}_{n T, w}$ : weighted standardized impact sound pressure level Dependent: $S$ : mean satisfaction response of tenants in a building block $T$ : percentage of satisfied tenants	Linear regression: ${L^{'}}_{n T, w} = 80.6 - 5.48 S$ ${L^{'}}_{n T, w} + C_{I, 50 - 2500} = \bar{I} - 6.4$ Suggested combination of equations: ${L^{'}}_{n T, w} + C_{I, 50 - 2500} = - 0.25 + 68.3$	Coefficients: $r = 75 %, R^{2} = 0.56$ $r = 96 %, R^{2} = 0.92$	Low frequencies below 100 Hz are important for accurate measurements. Unsatisfactory self-reports if ${L^{'}}_{n, w} < 48 dB$ . Minimum 53 dB suggested for and ${L^{'}}_{n T, w} + C_{I, 50 - 2500}$
22, 23	198 participants in Sweden 22 floors of several structure types were measured; 12 building data taken from Bodlund.¹⁶ Results in acoustic measurements and subjective responses were averaged per building block in the study as in Bodlund¹⁶	Independent: ${L^{'}}_{n, w, n e w, i}$ : weighted impact sound index combined with suggested new reference curves Dependent: $S$ : mean satisfaction response of tenants in a building block from 1 (quite unsatisfactory) to 7 (quite satisfactory)	Linear regression: ${\bar{L^{'}}}_{n, w, n e w, 01} = 76.29 - 4.10 S$ ${\bar{L^{'}}}_{n, w, n e w, 02} = 77.69 - 4.12 S$ ${\bar{L^{'}}}_{n, w, n e w, 03} = 79.28 - 4.09 S$ ${\bar{L^{'}}}_{n, w} = 75.35 - 4.58 S$ ${\bar{L^{'}}}_{n, w} + C_{I} = 73.31 - 4.22 S$ $\begin{array}{l} {\bar{L^{'}}}_{n, w} + C_{I, 50 - 2500} = \\ 74.4 - 4.17 S \end{array}$	Coefficients: $r = 85 %, R^{2} = 0.73$ $r = 86 %, R^{2} = 0.74$ $r = 87 %, R^{2} = 0.76$ $r = 74 %, R^{2} = 0.55$ $r = 79 %, R^{2} = 0.62$ $r = 84 %, R^{2} = 0.71$	Good association between impact noise and subjective response. New reference curve suggested for ISO standard method correction to achieve better correlation with subjective ratings. Low frequencies from 50 Hz need to be considered in the measurement spectra
24	159 residents in Finland 4 HW buildings, 2 LW buildings were included in a survey to compare responses from various structures	Independent: ${R^{'}}_{w}$ and ${R^{'}}_{w} + C_{50 - 3150}$ were equal for different structure types HW, LW Dependent: Several variables	Mann–Whitney U-test	Values within 95% CI in all cases	No significant differences found based on the responses of residents from different building types

HW: heavyweight; LW: lightweight; CI: confidence interval.

Reviewed results

The first extensive research took place in Europe in 1985, and it is reported by Bodlund.¹⁶ It concerns the evaluation of the sound conditions in Swedish buildings, specifically for impact sound insulation. That study proposed a basic method set for further research in building acoustics, using objective and subjective assessment of noise, and it was cited in many following publications. A wide sample of 350 Swedish dwellings was used, and acoustic measurements of impact sound transmission took place in building blocks of houses. The residents were interviewed in order to provide their ratings on acoustic behavior of their home using a satisfaction scale ranging from 1 (quite unsatisfactory) to 7 (quite satisfactory). About 464 scores were collected from 398 participants. The constructions tested were 22 floors, concrete, or timber joist floors in a sample of both attached houses and multistory residencies. All the data were grouped and averaged according to the actual urban building blocks, which consisted of similar constructions. There were at least 6 different floor measurements and about 20 interviews per block.

The average impact sound index $I_{i}$ (as defined in the old ISO Recommendation 717-2:1982) of the block’s measurements was compared with the average rating score of the block’s tenants, using linear regression analysis. For the mean impact sound index within a block, the typical standard deviation (SD) was reported at 3.7 dB. The study found that a mean rating of 4.4 of reported satisfaction corresponds to a 51% of the building tenants who regard their home acoustic conditions as good or very good. Thus, a lower response than that is considered not satisfactory for building standards in this article. The linear regression analysis of the data acquired a model of average $I_{i} = 86.3 - 5.4 S$ , where $S$ stands for the subjective mean score with r = 73%; this gives a determination coefficient of R² = 0.53. Bodlund compared his results also using the other ISO-weighted indices, ${\bar{L^{'}}}_{n, w}$ and A-weighted levels ${\bar{L^{'}}}_{n, A}$ which is used nowadays and then he acquired different values for correlation $r$ of 75% (R² = 0.56) and 72% (R² = 0.52), respectively.

Bodlund mentions that the tapping machine spectrum is different than the one excited by a person running; the tapping machine gives significantly higher amplitudes in middle and high frequencies, while walking excites mostly low frequencies on the floor structures. Furthermore, this effect is more intense in wooden structures, an argument which is supported in other following studies.⁶

Finally, Bodlund suggested that a new reference curve for the ISO 717 corrections with an emphasis on the low and middle frequencies would correlate better to subjective ratings. He used the study’s results to calculate a new curve, that being a straight line from 50 Hz to 1 kHz with a slope of 1 dB/octave. In this case, the regression model for the average suggested index was $I_{s} = 86.3 - 5.53 S$ and offered the best regression of 87%, with r = 0.87 (R² = 0.76). The suggestions for reference curves are part of a trend for ISO method corrections in the past literature, but it is not further described in this article.

In the same time with the previous studies, a similar study took place in Canada by Bradley^17,18 for investigation of airborne sound insulation in a wide sample of 300 constructions, row housing and multifloor buildings, in three cities. Acoustic measurements were performed in the party walls between houses, and interviews were taken face to face with 600 tenants. Responses to questions were given using a 7-point scale. The association of airborne sound reduction index and personal responses was analyzed by fitting linear regression models or sigmoidal Boltzmann equations. However, the models are not stated but only their $R^{2}$ and p-values, while the results with the fitting models are presented in graphs. The measured apparent $S T C$ values had a range of 38–60 dB (mean: 49.8 dB) as measured according to ASTM.¹⁶

When the residents were asked if they want to move out of their home due to noise, more than 94% replied positive, indicating that neighbor noise is a serious issue (fitted line slopes down with increasing $S T C$ reduction, $R^{2} = 0.56, p = 0.033$ ). The $S T C$ values were found to be significantly related to the residents’ satisfaction for their buildings ( $R^{2} = 0.83, p = 0.002$ ), as well as the feeling of having neighbors who consider to avoid making noise ( $R^{2} = 0.86, p = 0.001$ ). Specifically, $S T C$ values were associated with dissatisfaction from neighbors’ general noise from either side of the wall ( $R^{2} = 0.82, p < 0.002$ ), neighbors voice sounds ( $R^{2} = 0.94, p < 0.001$ ), neighbors’ TV sounds ( $R^{2} = 0.77, p < 0.004$ ), and also neighbors’ music ( $R^{2} = 0.92, p < 0.001$ ). It is stressed that annoyance depends not only on airborne sound reduction but also on noisy behavior of neighbors and how often neighbors make noise. Then, the $S T C$ values were associated with sleep awakening due to neighbor’s noise ( $R^{2} = 0.60, p = 0.024$ ) and the subjective rating of the tenants for the building’s sound insulation ( $R^{2} = 0.92, p < 0.001$ ). It is concluded in the study that after a minimum value of 50 dB for $S T C$ , the annoyance coming from most noise types decreases importantly. Above STC = 50 dB, there is some protection from music sounds transmission as well. Based on their results, the authors suggest a value of STC = 60 dB as an optimal solution since very few people would be annoyed then.

In the previous studies,^{1–2,20–22} the conclusions presented deal with a synthesis of results of previous studies in different countries which took place between the years 1972 and 1997. A short review of those studies is included in Hveem et al.²⁰ and Rindel and Rasmussen.²¹ Sound insulation data from buildings and self-reported noise annoyance data were compared in order to assess the satisfaction perception of building tenants.

Some further analyzed results from Bodlund¹⁶ are used in Hveem et al.²⁰ and Rindel and Rasmussen²¹ where another regression model was developed which was finally expressed as ${L^{'}}_{n, w} = 80.6 - 5.48 S$ and offered a correlation of 75% (R² = 0.56). The study of Hagberg²³ is referred, which evaluated Bodlund’s curve during impact sound measurements in 146 buildings using also the low-frequency spectrum adaptation terms $C_{I, 50 - 2500}$ . Another relation presented is ${L^{'}}_{n, w} + C_{I, 50 - 2500} = L_{B} - 6.4$ , where $L_{B}$ is the average impact sound index $I_{i}$ from Bodlund,¹⁶ and the acquired correlation between the two metrics is 96% (R² = 0.92). Finally, using the combination of the last equation and others from previous studies,^2,18,16 a new relation is derived, which can be used as a prediction model for subjective satisfaction of inside acoustic conditions: ${L^{'}}_{n, w} + C_{I, 50 - 2500} = - 0.25 T + 68.3$ with the same parameters r = 96% (R² = 0.92) that seems the most successful regression model developed in all selected studies.

In Hveem et al.,²⁰ another self-report assessment is published where 17 floor structures in multistory buildings were rated as satisfactory/good, barely satisfactory, or unsatisfactory. The observations indicated that the overall subjective response is satisfactory or good when the ${L^{'}}_{n, w}$ value is 5–10 dB lower than the Nordic minimum requirement of 58 dB. Therefore, a minimum suggestion is made for both ${L^{'}}_{n, w} ⩽ 53 dB$ and ${L^{'}}_{n, w} + C_{I, 50 - 2500} ⩽ 53 dB$ . The authors highlight also the need to include the low-frequency range below 100 Hz in the impact sound assessment and stress that human-induced vibrations affect the overall acoustic sense of floor structures as well.

In the study of Rindel,¹ a subjective satisfaction model is suggested, after observations from the studies examined before,²⁰ which is $P + F + G = 100 %$ . The variables $P$ , $F$ , and $G$ stand for subjective evaluation by the inhabitants as poor, fair, and good, respectively. It is also observed that the percentage for F typically approaches 30%. The linear regression slopes $A$ from past studies are compared, in cases where the correlation coefficients are |r| > 0.7 to be considered as good correlations for the researchers. It is summarized that dose–response curves have an average slope of 4% per decibel in the examined cases, after comparing insulation levels and subjective annoyance. The latter means that for every additional decibel in the noise pressure levels, airborne or impact, the satisfaction of the tenants decreases 4%, in the satisfaction scales. The satisfaction scales are not the same for every combined result from different studies, although those results are claimed to be valid for all cases of airborne and impact noise generated from co-habitants, as well as road traffic noise in dwellings. Besides dissimilar scales, shortcomings are also reported due to different definitions of subjective parameters (poor, fair, good, satisfaction, etc.) in different surveys (questionnaires and oral interviews). No specific information is reported on how to handle those discrepancies.

Additional conclusions by Rindel and Rasmussen²¹ stress the need for low-frequency adaptation terms for improved correlation between airborne and impact sound insulation and subjective responses of tenants. This problem concerns mostly timber structures due to their poor performance in low frequencies down to 20 Hz. There is also an ad hoc suggestion mentioned for airborne sound insulation to be satisfactory, at least 2/3 of the tenants should consider it good which corresponds to a minimum of ${R^{'}}_{w} = 60 dB$ .

In a following study by Hagberg,^22,23 the results from Bodlund¹⁶ were enriched with new data, and they were reprocessed. Another 10 Swedish buildings of various structures were tested with impact sound measurements; 198 new participants were interviewed with the same method. The new data were combined with previous measurements from 12 buildings from a previous study¹⁶ to make up a total sample of 22 buildings. Linear regression models were performed again to test the data association between residents’ satisfaction and impact sound index values. All data were averaged again per building block as before in Bodlund.¹⁶ New reference contours were tested too, as well as the previous suggested Bodlund’s reference curve, which was found insufficient to associate well with the new sample data.

The apparent airborne sound reduction index ${R^{'}}_{w} + C_{50 - 3150}$ was very well predicted by the user’s ratings of perceived airborne noise transmission, having a correlation of 85% (R² = 0.73) but not so well when omitting the adaptation terms: without $C_{50 - 3150}$ , it was 76% (R² = 0.58). Contrary to airborne sound, the standardized impact sound index ${L^{'}}_{n, w}$ was poorly correlated to the relevant questions for impact noise annoyance r = 51% (R² = 0.26), even when using adaptation terms ${L^{'}}_{n, w} + C_{50 - 3150}$ (r = 57%, R² = 0.32). One explanation for that is the lack of important low-frequency content in the measurements, which affects the final value of the descriptor ${L^{'}}_{n, w}$ . However, the suggested improved index ${L^{'}}_{n, w} + C_{I, 20 - 2500}$ which has a spectrum adaptation expanding down to 20 Hz was much better correlated r = 86% (R² = 0.74) indicating that impact sound assessment should include the very low-frequency content as well. Finally, another adaptation term curve is proposed within this project, the $C_{I, A k u L i t e, 20 - 2500}$ , which follows the same weights of ISO 717-2 between 50 and 400 Hz, increases 2 dB per 1/3 octave bands below that and increases 1 dB per 1/3 octave band after 400 Hz. The optimal results acquired in regression between occupants responses and ${L^{'}}_{n, w} + C_{I, A k u L i t e, 20 - 2500}$ had a correlation coefficient of r = 92% (R² = 0.85).

Previous studies^6–15 have many aspects in common; since they are contemporary, they follow the same methodology and occurred about the same period. First, they all deal with the subject of evaluation of acoustic comfort in multistory family dwellings, based on the combination of objective and subjective data. They use standardized procedures of airborne and impact sound (standardized tapping machine and Japanese rubber ball) following the relevant standards (ISO 717-1 and -2,^57,58 ISO 140-4 and -7,^59,60 EN 12354-1 and -2,^52,53 and JIS A 1418-2⁵⁶), for the characterization of sound insulation of building elements. The acoustic measurements took place in selected living rooms or bedrooms in the study buildings in every case. Questionnaires were developed, for the rating of noise assessment into the participants’ apartments based on a common methodology described in Simmons⁶³ and following ISO 15666.⁶⁴ In all cases, the question formulations included annoyance due to noise and vibration from neighboring apartments, noise from neighbors in common or collective spaces, noise from technical installations or equipment, outdoor noise, and noise inside the tenant’s apartment. Then, the results between acoustic measurements and perceived noise annoyance were compared, and the degree of association among the collected data was investigated.

In previous articles,^6–10 the AkuLite research program is presented, a study with a sample of 10 Swedish multistory buildings: a typical heavy concrete building and 9 lightweight (LW) structures (4 wooden, 4 made of cross-laminated timber, and 1 of steel framework). A total of two typical rooms one above another were measured acoustically in each test building. A total of 251 responses were collected from participating tenants (reported response rate circa 30%) of the test buildings. The AkuLite questionnaire consisted of 15 questions concerning noise annoyance inside apartments. The measurement data and subjective responses, grouped in mean values for every building, were evaluated statistically using linear regression analysis within 95% CI.

The results of the AkuLite project were enriched in a continuation study presented by Ljunggren et al.^11,12 Acoustic measurements and surveys in another 13 Swedish buildings took place, since the previous sample of buildings was limited according to the authors. The same methodology was followed, and about 800 responses were collected; the associations of standardized impact noise levels measured with tapping machine to self-report annoyance were explored from the total sample number of 23 buildings. Again, the standardized impact sound index ${L^{'}}_{n T, w}$ was poorly correlated with subjective annoyance from footstep noise (R² = 0.18). Better results were acquired with the inclusion of lower frequencies from 50 Hz in the index calculations, offering a determination coefficient of R² = 0.49 for ${L^{'}}_{n, w} + C_{I, 50 - 2500}$ . Even better results were acquired from 20 Hz, with values of R² = 0.71 for the standardized ${L^{'}}_{n, w} + C_{I, 20 - 2500}$ and with R² = 0.65 for the indicator ${L^{'}}_{n, w} + C_{I, A k u L i t e, 20 - 2500}$ . The previous results from Ljunggren et al.⁶ were confirmed. This study also included in the sample six heavyweight (HW) concrete buildings instead of only two as in the previous study. It was highlighted that for HW cases, inclusion of low frequencies does not change the results drastically as for the LW cases, which show an undesired acoustic behavior in the low-frequency range.

Summing up for the AkuLite project, the perception outcome of the tenants’ responses indicates that the noise annoyance due to airborne sound transmission is generally low, and the correlation of objective and subjective data is sufficient. In contrast to that, low-frequency noise induced by impact sound was found to be the highest recorded source in both acoustic measurements and self-reported noise annoyance. The indicators ${L^{'}}_{n T, w}$ and ${L^{'}}_{n, w} + C_{I, 50 - 2500}$ were found poorly associated with self-reported annoyance. To achieve a sufficient correlation between impact sound index and subjective assessment, the low-frequency range down to 20 Hz is considered essential. This study did not only confirm the importance of low-frequency range in the measurements but further claimed expansion down to 20 Hz, instead of 50 Hz as in Bodlund¹⁶ and Hagberg.^22,23

In Guigou-Carter et al.,¹³ 10 various construction buildings were measured for the French study project Acoubois, some multistory ones and some attached houses. The survey included questions about annoyance from several noise types, similar to Ljunggren et al.⁶ The sample size and response rate are not stated in the publication (reported only 57% females, age span 26–59 years), as well as other essential data about the study design. In the study results, 85% of the tenants reported sound insulation to be very important. Overall, more than 50% did not report any annoyance, which is considered as a satisfactory result for the French regulations according the authors.

The correlation coefficients were calculated for some questions corresponding to the measurements of airborne and impact sounds. According to that study, the best correlation found for the impact sound index ${L^{'}}_{n, w} + C_{150 - 2500}$ to the self-reported impact noise annoyance with (r = 0.89, R² = 79%). ${L^{'}}_{n, w} + C_{150 - 2500}$ is assessed as better correlated than the Japanese ball impactor index $L_{A F m a x}$ (r = 0.85, R² = 73%), ${L^{'}}_{n, w}$ alone (r = 0.86, R² = 74%), or ${L^{'}}_{n, w} + C_{I}$ (r = 0.85, R² = 73%). Some optimal values of ${L^{'}}_{n, w} + C_{150 - 2500}$ = 52 dB and $L_{A F m a x} = 54 dB (A)$ are suggested after the result evaluation. However, no statistical significant levels were presented for the results. The authors conclude that impact noises are more annoying than others for the tenants’ comfort. More than 50% were from quite to very annoyed by impact noise from neighbors walking and 30% quite to very annoyed by vibrations from neighbors walking, moving, or dropping objects.

In the articles of Milford et al.¹⁴ and Høsøien et al.,¹⁵ another study is presented, which took place in Norway for the evaluation of subjective sound quality ratings in newly built dwellings (2002–2015). Field measurements in 600 buildings were done alongside a socio-acoustic survey with a questionnaire sent to the occupants of the buildings. In total, 702 residents answered to 35 questions, similar to the ones developed in JIS A 1418-2.⁵⁶ The articles elaborate on the responses from questions regarding annoyance due to airborne and impact sounds coming from neighbors from the above floor in a slightly differentiated scale from 1 (not annoyed) to 5 (extremely annoyed).

The results indicate that 65% of the occupants are at least slightly disturbed, and the authors emphasize on the wide problem of low-frequency noise; 33% report worried about their own TV, music, or speech annoying other occupants; 20% of the occupants report at least moderately annoyed by traffic noise or impact sounds from neighbors above. The articles mention that impact sound annoyance from neighbors above, especially footfall noise, is reported as stressing as road traffic annoyance. Bad correlation is reported between subjective ratings and the weighted impact sound indices ${L^{'}}_{n, w}$ and ${L^{'}}_{n T, w}$ unless the correction term $C_{I, 50 - 2500}$ is included. ${L^{'}}_{n T, w} + C_{I, 50 - 2500}$ is thus reported as the best predictor of subjective annoyance within 95% CI. For airborne sounds annoyance, $D_{n T, w}$ is reported as the best descriptor to predict occupants’ annoyance. Good agreement of the slopes of the dose–response curves with results from Rindel¹ is mentioned as well. Finally, it is reported that more than half of the residents questioned would be willing to pay an extra cost for better acoustic conditions.

Another building survey was setup in Finland²⁴ to compare acoustic satisfaction in different multistory building structures with similar airborne sound insulation of walls. Specifically, four HW concrete buildings with measured ${R^{'}}_{w} = 66 dB$ (floors) and ${R^{'}}_{w} = 56 dB$ (walls) and two LW building structures with ${R^{'}}_{w} = 63 dB$ (floors) and ${R^{'}}_{w} = 57 dB$ (walls) were included. All buildings fulfilled the Finish Building Code requirements ( ${R^{'}}_{w} ⩾ 55 dB$ and ${L^{'}}_{n, w} ⩽ 53 dB$ ). However, the sound insulation in low frequencies was significantly lower for the LW cases. A sample of 159 residents (72 in HW and 87 in LW) replied to a wide questionnaire, and they offered responses on noise annoyance to certain sources, satisfaction, noise sensitivity, and other variables. The associations between the variables were tested with nonparametric Mann–Whitney U-test, and results were reported within 95% CI. No significant statistical differences were found between the groups of residents, except for two control variables: education and extraversion, which were neglected. No significant differences were found for the examined building types (HW and LW) and dependent variables: willingness to move out (yes/no), distraction of performance, inconvenience from neighbor noises, and disturbance from certain noise sources. Loud speech, TV, and music were reported as the highest annoyance for airborne sound. Also, impact noise types such as footsteps, moving furniture, and closing doors were reported as the most disturbing. Inconvenience from outside noises, as well as disturbance to outside traffic noise, was higher for residents of LW buildings: that is reasonable concerning low acoustic performance at low frequencies for LW structures. However, technical installations noise was significantly higher in HW buildings.

Discussion

This review article concerns studies which include acoustic data from in situ measurements and self-reported responses from the residents, collected with surveys or interviews in test buildings. The selected field studies explore acoustic comfort in buildings through the association of objective and subjective data. Few researches were found to fulfill the aims of this review, namely, 10 separate studies reported in 24 articles. Noticeably eight of the analyzed field studies are Scandinavian, then one is French, and one is Canadian.

Most of the studies found during our search were conducted in laboratory tests, and subjective assessment was evaluated with listening experiments. The laboratory tests are easier to set up, but laboratory measurements ignore the interaction of the whole building structure during sound propagation. In contrast, field measurements capture the real acoustic behavior of structures. The mental state of a participant can be also different in a laboratory than being in the actual living environment and offering spontaneous judgments. Therefore, only results from field studies were chosen to be investigated in this review article.

From all the articles dealing with acoustic comfort and even including the term “comfort” or “quality” in their title, only two of them provide an actual definition of those concepts.^2,5 More definitions could be reported, and the writers should generally elaborate more on the concept of comfort.

The review revealed that noise issues in residential buildings are significant for acoustic comfort, especially impact noise sounds produced by neighbors that include many low-frequency components.^6–15 An overview of the noise types reported (and which ones were found most important) in the different studies is given in Table 4. Many studies report that specifically impact noise leads to high disturbance according to human perception results, mostly cases of impact sound from neighbors walking, either barefoot or not. Neighbors’ steps from the above floor are reported as the most annoying noise source for residents.^6–13,16 Some study results are even more specific for noise types, such as in Ljunggren et al.¹² where footfall walking is stressed to be the biggest disturbance due to the excitation of many low frequencies that propagate through flanking transmission paths too, that is, through connected floors and walls.

Table 4.

Noise types reported in the different case studies.

Noise types	References (one for each study)
Airborne noise in general from neighbors	14, 17
Airborne noise from neighbors in general (daily living, talking, audio, and TV)	6, 13, 14, 17, 24
Airborne noise from neighbors’ music (low frequencies)	6, 14, 17
Impact noise in general from neighbors	6^a, 13, 14, 17^a
Impact noise from neighbors moving/dropping objects	13, 14, 22, 24
Impact noise by footsteps (neighbors walking barefoot)	6, 13, 14^a, 16, 22, 24
Impact noise by neighbors walking on heels/hard sole shoes	6, 13, 14, 16, 22
Traffic noise	6, 13, 14, 24
Noise in common areas	13
Outdoor noise	13
Noise within a flat	13
Vibration induced from machinery in other flats	13
Vibration induced from neighbors’ walking	13

Reported as most annoying.

Many studies also conclude that extended frequency spectra which include frequencies below 100 Hz correlate better with self-reported responses on noise annoyance, especially for the impact sound cases.^{1,6–14,16,17} This finding further underlines the observation that low frequencies could offer results for better prediction of human perception in living environments. Besides being an overall suggestion, it is considered a necessity for LW structures, where the most problematic noise propagation occurs in low frequencies, due to resonances of structural elements and coupling among them.⁶ In LW building structures, the impact sound insulation standards can be met, and the ${L^{'}}_{n, w}$ curve and the single value might look sufficient, but the residents might still complain firmly for low-frequency noise transmission. Further expansion of the whole frequency range down to 20 Hz has been recently suggested,^6–12 while ISO standards have 100 Hz as the lowest frequency limit; thus, researchers do not measure any lower frequencies. An exception there is in Sweden, where the national standards comply with ISO ones, but they demand measurements down to 50 Hz. In contrast to the results for impact sounds, the results for airborne sound insulation are not that ambiguous. Few complaints have been reported,¹³ but generally occupants offer subjective ratings which indicate overall satisfaction.

Noise content with intense low-frequency characteristics can be more disturbing while propagating through LW building components. LW structures offer better sound reduction than HW ones but not in the low-frequency range.^6–10 Below around 100 Hz, the performance is expected to change, with poorer insulation of LW walls as indicated by the in situ studies.

In some articles, the authors suggest specific values for building acoustic indicators, which came up as efficient to represent a good level of acoustic conditions in every case study. The suggested values usually correspond to 50% satisfaction of residents, and they are presented in Table 5. The most important suggestions are as follows:

${R^{'}}_{w} ⩾ 60 dB$ , since it is suggested by Rindel and Rasmussen²¹ and confirmed in a different study with the similar $S T C$ indicator by Bradley.¹⁷

${L^{'}}_{n T, w}$ and ${L^{'}}_{n, w} + C_{I, 50 - 2500} ⩽ 56 dB$ , coming from results of a complete wide study by Hagberg²² published in a journal.

Table 5.

Suggested acoustic indicator values corresponding to satisfaction.

Indicator	Minimum/maximum requirements	References	Publication type
$S T C$	⩾60 dB	17	++
${R^{'}}_{w}$	⩾60 dB	19	++
${R^{'}}_{w}$	⩾55 dB	14	++
${L^{'}}_{n, w}$ and ${L^{'}}_{n, w} + C_{I, 50 - 2500}$	⩽48 dB	18	++
${L^{'}}_{n, w}$ and ${L^{'}}_{n, w} + C_{I, 50 - 2500}$	⩽53 dB	18	++
${L^{'}}_{n, w}$ and ${L^{'}}_{n, w} + C_{I, 50 - 2500}$	⩽56 dB	20	+++
${L^{'}}_{n, w}$	⩽53 dB	12	++

++: scientific journal; ++: conference article; +: report.

Concerning prediction models developed from the results, some of them are good with high determination coefficients, that is, high $R^{2}$ values, which explain the variance of the model. Best regression model from field measurements observed so far is ${L^{'}}_{n, w} + C_{I, 50 - 2500} = - 0.25 T + 68.3$ (r = 0.96, R² = 92%), which came up as a synthesis of results from several articles.^16,20,21 The term $T$ corresponds to the percentage of satisfied tenants.

Although statistical methods are used to compare and associate results from objective and subjective data, in the examined literature, many shortcomings take place in the study designs and reporting methods and results. First, the biggest problem observed in many studies is the lack or misuse of basic statistic indicators; some of them do not even mention the sample size of participants¹³ or other parameters such as p-values. Furthermore, the presentation of the outcome is not always successful, even if it is important. In many studies, the regression models are presented with the independent variable (usually airborne or impact sound levels) on the y-axis and the dependent on the x-axis, while the opposite is the usual way for statistical data representation. In few cases, the regression models and parameters represent the opposite relationship between dependent and independent variables.^16,22 That makes the comparison or regression models and their parameters cumbersome. For most studies, there is no assumptions analyzed for the used methods and any tests of statistical significance; few of them provide sufficient information on the test design and mention parameters such as p-values on their results. The insufficient statistical background of the studies can be clearly seen in the evaluation criteria fulfillment in Table 2.

Conclusion

The study review highly indicates that there are serious annoyance issues which affect acoustic comfort in dwellings. There exist especially problems with impact noise types from neighbors, which include a high degree of low-frequency content. Specifically, walking noise has been reported as the most disturbing noise source. Also, the lack of very low-frequency content in the impact sound measurements leads to weak statistical association with subjective response of residents. Therefore, most studies suggest that measurements should include extended frequencies down to 50 Hz (or even down to 20 Hz), instead of 100 Hz which is the present lowest limit in the ISO standards. The greatest problems with impact noise and related low-frequency transmission are found in LW structures, while concrete buildings have better overall insulation against noise transmission, airborne or impact.

Many studies included in this review lending data to these suggestions lack rigorous scientific presentation of results and statistical methods leading to a risk of misinterpretation. We suggest a harmonized description of methods and results using common acoustic and statistical indicators, sufficient reporting of statistical evaluation parameters, and the testing for statistical significance.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article was written within the research project “Acoustic Comfort in Building Apartments” funded by Saint-Gobain Weber.

References

Rindel

. Acoustic quality and sound insulation between dwellings. Build Acoust 1999; 5: 291–301.

Rindel

. Acoustical comfort as a design criterion for dwellings in the future. In: Proceedings of the sound in the built environment, Aucklandm, New Zealand, 21–23 November 2002.

Rasmussen

Rindel

. Sound insulation of dwellings—legal requirements in Europe and subjective evaluation of acoustical comfort. In: Proceedings of the DAGA’03, Aachen, 18 March 2003.

Rasmussen

Rindel

. Concepts for evaluation of sound insulation of dwellings—from chaos to consensus? In: Proceedings of the 2005 forum acusticum, Budapest, 29 August–2 September, 2005.

Rasmussen

Rindel

. Sound insulation between dwellings—requirements in building regulations in Europe. Appl Acoust 2010; 71: 373–385.

Ljunggren

Simmons

Hagberg

. Correlation between sound insulation and occupants perception—proposal of alternative single number rating of impact sound. Appl Acoust 2014; 85: 57–68.

Ljunggren

Simmons

Hagberg

. Findings from the AkuLite project: correlation between measured vibro-acoustic parameters and subjective perception in lightweight buildings. In: Proceedings of the inter-noise 2013, Innsbruck, 15–18 September 2013.

Hagberg

Bard

. Low frequency sound transmission in multifamily wooden houses. In: Proceedings of the inter-noise 2014, Melbourne, VIC, Australia, 16–19 November 2014.

Simmons

Ljunggren

. Aku20—searching for optimal single number quantities in EN ISO 717—2 correlating field measurements 20—5000Hz to occupant’s ratings of impact sounds—new findings for concrete floors. In: Proceedings of the EURONOISE 2015, Maastricht, 31 May–5 June 2015.

10.

Simmons

Hagberg

Backman

. Acoustical performance of apartment buildings—resident’s survey and field measurements: AkuLite report 2, 2011, http://www.diva-portal.org/smash/get/diva2:962643/FULLTEXT01.pdf

11.

Ljunggren

Simmons

Öqvist

. Correlation between sound insulation and occupants perception—proposal of alternative single number rating of impact sound—part II. Appl Acoust 2017; 123: 143–151.

12.

Ljunggren

Simmons

Öqvist

. Evaluation of impact sound insulation from 20Hz. In: Proceedings of the 24th international congress on sound and vibration, ICSV24, London, 23–24 July 2017.

13.

Guigou-Carter

Balanant

Villenave

. Acoustic comfort evaluation in lightweight wood-based buildings. In: Proceedings of the forum acusticum, Krakow, 7–12 September 2014.

14.

Milford

Høsøien

Løvstad

et al . Socio-acoustic survey of sound quality in dwellings in Norway. In: Proceedings of the inter-noise 2016, Hamburg, 21–24 August, 2016.

15.

Høsøien

Rindel

Løvstad

et al . Impact sound insulation and perceived sound quality. In: Proceedings of the inter-noise 2016, Hamburg, 21 August 2016.

16.

Bodlund

. Alternative reference curves for evaluation of the impact sound insulation between dwellings. J Sound Vib 1985; 102: 381–402.

17.

Bradley

. Deriving acceptable values for party wall insulation from survey results. In: Proceedings of the inter-noise 2001, Hague, 27–30 August 2001.

18.

Bradley

. Acoustical measurements in some Canadian homes. Can Acoust 1986; 14(4): 1–40.

19.

ASTM E413-16:1999. Classification for rating sound insulation.

20.

Hveem

Homb

Hagberg

et al . Low-frequency footfall noise in multi-storey timber frame buildings. NKB Nordic Committee on Building Regulations, NKB Committee and Work Reports 1996:12 E, 1997. Helsinki: Monila Oy.

21.

Rindel

Rasmussen

. Assessment of airborne and impact noise from neighbors. In: Proceedings of the inter-noise 97, Budapest, 1997, pp. 1739–1744.

22.

Hagberg

. Evaluating field measurements of impact sound. Build Acoust 2010; 17: 105–128.

23.

Hagberg

. Ljudkrav mestod av ISO/DIS 717. Nordic Committee on building Regulations, NKB Committee and Work Reports 1996:02, 1997. Helsinki: Monila Oy.

24.

Hongisto

Mäkilä

Suokas

. Satisfaction with sound insulation in residential dwellings—the effect of wall construction. Build Environ 2015; 86: 309–320.

25.

Yeon

. Subjective evaluation of floor impact noise based on the model of ACF/IACF. J Sound Vib 2001; 241(1): 147–155.

26.

Yeon

Jeong

. Objective and subjective evaluation of floor impact noise. J Temp Des Architect Environ 2002; 2(1): 20–28.

27.

Jeon

Jeong

Vorländer

et al . Evaluation of floor impact sound insulation in reinforced concrete buildings. Acta Acust United Ac 2004; 90: 313–318.

28.

Jeon

Ryu

Jeong

et al . Review of impact ball in evaluating floor impact sound. Acta Acust United Ac 2006; 92: 777–786.

29.

Lee

Kim

Jeon

. Psychoacoustical characteristics of impact ball sounds on concrete floors. Acta Acust United Ac 2009; 95: 707–717.

30.

Jeon

Sato

. Annoyance caused by heavyweight floor impact sounds in relation to the autocorrelation function and sound quality metrics. J Sound Vib 2008; 311: 767–785.

31.

Jeon

Lee

Kim

et al . Subjective evaluation of heavy-weight floor impact sounds in relation to spatial characteristics. J Acoust Soc Am 2009; 125(5): 2987–2994.

32.

Kim

Ryu

Jeon

. Effect of temporal decay on perception of heavy-weight floor impact sounds. J Acoust Soc Am 2013; 134(4): 2730–2738.

33.

Jeon

. Psychoacoustical evaluation of heavyweight floor impact sounds in apartment buildings. In: Proceedings of the inter-noise 2014, Melbourne, VIC, Australia, 16–19 November 2014.

34.

Ryu

Sato

Kurakata

et al . Relation between annoyance and single-number quantities for rating heavy-weight floor impact sound insulation in wooden houses. J Acoust Soc Am 2011; 129(5): 3047–3055.

35.

Gover

Bradley

Schoenwald

et al . Subjective ranking of footstep and low-frequency impact sounds on lightweight wood-framed floor assemblies. In: Proceedings of the forum acusticum 2011, Aalborg, 26 June 2011.

36.

Gover

Bradley

Zeitler

et al . Objective and subjective assessment of lightweight wood-framed floor assemblies in response to footstep and low-frequency impact sounds. In: Proceedings of the inter-noise 2011, Osaka, Japan, 4 September 2011.

37.

Späh

Hagberg

Bartlomé

et al . Subjective and objective evaluation of impact noise sources in wooden buildings. Build Acoust 2013; 20(3): 193–214.

38.

Kylliäinen

Hongisto

Oliva

et al . A laboratory listening experiment on subjective and objective rating of impact sound insulation of concrete floors. In: Proceedings of the inter-noise 2016, Hamburg, 21–24 August 2016.

39.

Kylliäinen

Hongisto

Oliva

et al . Subjective and objective rating of impact sound insulation of a concrete floor with various coverings. Acta Acust United Ac 2017; 103: 236–250.

40.

Öqvist

Ljunggren

Johnsson

. Listening test of walking noise from 20 Hz in dwellings. In: Proceedings of the inter-noise 2017, Hong Kong, China, 27–30 August 2017, pp. 3935–3942.

41.

Monteiro

Machimbarrena

De la Prida

et al . Subjective and objective acoustic performance ranking of heavy and light weight walls. Appl Acoust 2016; 110: 268–279.

42.

Rychtáriková

Muellner

Chmelík

et al . Perceived loudness of neighbour sounds heard through heavy and light-weight walls with equal R_w+ C_(50–5000). Acta Acust United Ac 2016; 102(1): 58–66.

43.

Rychtáriková

Roozen

Müllner

et al . Listening test experiments for comparisons of sound transmitted through light weight and heavy weight walls. Akustika 2013; 19: 1–10.

44.

Rychtáriková

Müllner

Urbán

et al . Influence of temporal and spectral features of neighbour’s noise on perception of its loudness. In: Proceedings of the EURONOISE 2013, Innsbruck, 15 September 2013.

45.

Pedersen

Antunes

Rasmussen

. Online listening tests on sound insulation of walls—a feasibility study. In: Proceedings of the EURONOISE 2012, Prague, 10–13 June, 2012.

46.

Vian

J-P

Danner

Bauer

. Assessment of significant acoustical parameters for rating sound insulation of party walls. J Acoust Soc Am 1983; 73(4): 1236–1243.

47.

Tachibana

Hamado

Sato

. Loudness evaluation of sounds transmitted through walls—basic experiments with artificial sounds. J Sound Vib 1988; 127(3): 499–506.

48.

Park

Bradley

Gover

. Evaluating airborne sound insulation in terms of speech intelligibility. J Acoust Soc Am 2008; 123(3): 1458–1471.

49.

Park

Bradley

. Evaluating standard airborne sound insulation measures in terms of annoyance, loudness and audibility ratings. J Acoust Soc Am 2009; 126(1): 208–219.

50.

Hongisto

Oliva

Keränen

. Subjective and objective rating of airborne sound insulation—living sounds. Acta Acust United Ac 2014; 100: 848–864.

51.

Virjonen

Hongisto

Oliva

. Optimized single-number quantity for rating the airborne sound insulation of constructions: living sounds. J Acoust Soc Am 2016; 140(6): 4428–4436.

52.

ISO 12354-1:2017. Building acoustics—estimation of acoustic performance of buildings from the performance of elements—part 1: airborne sound insulation between rooms.

53.

ISO 12354-2:2017. Building acoustics—estimation of acoustic performance of buildings from the performance of elements—part 2: impact sound insulation between rooms.

54.

ISO 16283-1:2014. Field measurement of sound insulation in buildings and of building elements—part 1: airborne sound insulation.

55.

ISO 16283-2:2014. Field measurement of sound insulation in buildings and of building elements—part 2: impact sound insulation.

56.

JIS A 1418-2:2000. Acoustic—measurement of floor impact sound insulation of buildings—part 2: method using standard heavy impact sources.

57.

ISO 717-1:1996. Acoustics—rating of sound insulation in buildings and of buildings elements—part 1: airborne sound insulation.

58.

ISO 717-2:1996. Acoustics—rating of sound insulation in buildings and of buildings elements—2: impact sound insulation.

59.

ISO 140-4:1998. Acoustics—measurement of sound insulation in buildings and of building elements—part 4: field measurements of airborne sound insulation between rooms.

60.

ISO 140-7:1998. Acoustics—measurement of sound insulation in buildings and of building elements—part 7: field measurements of impact sound insulation of building elements.

61.

Höfler

. The Bradford hill considerations on causality: a counterfactual perspective. Emerg T Epidemiol 2005; 2005: 2–11.

62.

Schünemann

Hill

Guyatt

et al . The GRADE approach and Bradford Hill’s criteria for causation. J Epidemiol Comm Health 2011; 65: 392–395.

63.

Simmons

. Chapter 6. Developing a uniform questionnaire for socio-acoustic surveys in residential buildings. In: Fausti

Carrascal

Ingelaere

et al . (eds) Building acoustics throughout Europe, vol. 1: towards a common framework in building acoustics throughout Europe. 2013, pp. 102–124.

64.

ISO/TS 15666:2003. Acoustics—assessment of noise annoyance by means of social and socio-acoustic surveys.

References	Publication type	Strength of association	Consistency	Specificity	Temporality	Biological gradient	Plausibility	Coherence	Experiment design	Analogy
1	++	++	+	+	+	++	+	+	+	+

6	+++	+++	+++	++	++	++	++	++	++	++
7	++	++	++	++	++	++	++	++	++	++
8	++	++	++	++	++	++	++	++	++	++
9	++	++	++	++	++	++	++	++	++	++
10	+	++	++	++	++	++	++	++	++	++

11	+++	++	++	++	++	++	++	++	++	++
12	++	++	++	++	++	++	++	++	++	++

13	++	+	+	+	+	+	+	+	+	+

14	+++	+++	++	++	++	++	++	++	++	++
15	++	++	++	++	++	++	++	++	++	++

16	+++	+++	++	++	++	++	++	++	++	++

17	++	+++	+++	++	++	+++	+++	+++	++	++
18	+++	+++	+++	++	++	+++	+++	+++	++	++

20	+	++	++	+	+	++	++	++	+	+
21	++	++	++	+	+	++	++	++	+	+

22	+++	+++	+++	++	++	+++	+++	+++	+++	++

23	+	+++	++	++	++	+++	+++	+++	+++	++

24	+++	++	++	++	++	+++	++	++	++	++

References	Publication type	Strength of association	Consistency	Specificity	Temporality	Biological gradient	Plausibility	Coherence	Experiment design	Analogy
1	++	++	+	+	+	++	+	+	+	+

6	+++	+++	+++	++	++	++	++	++	++	++
7	++	++	++	++	++	++	++	++	++	++
8	++	++	++	++	++	++	++	++	++	++
9	++	++	++	++	++	++	++	++	++	++
10	+	++	++	++	++	++	++	++	++	++

11	+++	++	++	++	++	++	++	++	++	++
12	++	++	++	++	++	++	++	++	++	++

13	++	+	+	+	+	+	+	+	+	+

14	+++	+++	++	++	++	++	++	++	++	++
15	++	++	++	++	++	++	++	++	++	++

16	+++	+++	++	++	++	++	++	++	++	++

17	++	+++	+++	++	++	+++	+++	+++	++	++
18	+++	+++	+++	++	++	+++	+++	+++	++	++

20	+	++	++	+	+	++	++	++	+	+
21	++	++	++	+	+	++	++	++	+	+

22	+++	+++	+++	++	++	+++	+++	+++	+++	++

23	+	+++	++	++	++	+++	+++	+++	+++	++

24	+++	++	++	++	++	+++	++	++	++	++

Review of acoustic comfort evaluation in dwellings—part I: Associations of acoustic field data to subjective responses from building surveys

Abstract

Keywords

Introduction

Methods

Review criteria

Summary of methods, metrics, and quantities in the reviewed studies

Evaluation of included studies

Reviewed results

Discussion

Conclusion

Footnotes

Declaration of conflicting interests

Funding

References

References	Publication type	Strength of association	Consistency	Specificity	Temporality	Biological gradient	Plausibility	Coherence	Experiment design	Analogy
1	++	++	+	+	+	++	+	+	+	+

6	+++	+++	+++	++	++	++	++	++	++	++
7	++	++	++	++	++	++	++	++	++	++
8	++	++	++	++	++	++	++	++	++	++
9	++	++	++	++	++	++	++	++	++	++
10	+	++	++	++	++	++	++	++	++	++

11	+++	++	++	++	++	++	++	++	++	++
12	++	++	++	++	++	++	++	++	++	++

13	++	+	+	+	+	+	+	+	+	+

14	+++	+++	++	++	++	++	++	++	++	++
15	++	++	++	++	++	++	++	++	++	++

16	+++	+++	++	++	++	++	++	++	++	++

17	++	+++	+++	++	++	+++	+++	+++	++	++
18	+++	+++	+++	++	++	+++	+++	+++	++	++

20	+	++	++	+	+	++	++	++	+	+
21	++	++	++	+	+	++	++	++	+	+

22	+++	+++	+++	++	++	+++	+++	+++	+++	++

23	+	+++	++	++	++	+++	+++	+++	+++	++

24	+++	++	++	++	++	+++	++	++	++	++