Research Methods for Understanding Aircraft Noise Annoyances and Sleep Disturbance (2014)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Appendix H. Annoyance Literature Review EXECUTIVE SUMMARY AND CONCLUSIONS A literature review has been conducted to support plans for a new national survey of aircraft noise in the United States. The review supports the planning by identifying methods for estimating a national dose- response relationship (the survey’s primary announced goal), evaluating issues about non-noise factors that are hypothesized to affect noise annoyance, and identifying unresolved noise annoyance issues that could be secondary goals for the national survey. The measure of impact for this survey, as for all dose- response noise regulations, is the privately-expressed noise annoyance that is measured in social surveys, not the visible, publicly-expressed actions such as complaints to authorities, lawsuits or public protests. Primary Goal The announced primary goal is to form an accurate nation-wide estimate of the dose-response relationship: the function that predicts the proportion of the population annoyed (to varying degrees) by aircraft noise from acoustical data that characterizes the aircraft noise at their residences.20 The literature review supports a decision to use two noise annoyance questions from an ISO technical specification that are now widely used in noise annoyance surveys around the world. The review found that when multiple cities have been surveyed in a single study, the residents in different geographic areas (neighborhoods, cities, airports, etc.) have significantly different annoyance reactions to the same noise level. Despite many hypotheses, the causes of these differences have not been established. One of the implications of such geographical effects is that the national survey design should be based on a large number of randomly selected airports that are drawn with probability selection methods. The literature review identified one major uncertainty concerning plans for estimating the dose-response relationship: the mode of administering a questionnaire. Noise annoyance surveys have always varied as to whether they are interviewer-administered (face-to-face or telephone) or self-administered (usually mail-in), but more recently the economics of interviewing have resulted in some large surveys being self- administered. There is enough uncertainty about the effect of interviewing mode to mean that the results of a new US survey could not be compared with many other surveys unless effects of the survey mode are evaluated with new data from the US survey. The literature review identified 62 hypotheses about non-noise effects on annoyance and attempted to locate summaries that could be expected to provide evidence about each of the hypotheses. With over 600 noise surveys and over 1,000 publications it was not possible to conduct new summaries of evidence on each of the 62 issues. As a result the findings from previous published summaries are presented when available. Summaries are needed because individual studies often report contradictory findings and, as a result, conclusions can only be reached about average results when many studies are combined. These summaries are presented for the 30 of the 62 issues for which a results-neutral search strategy produced the summary. For the remaining issues some study results are identified but the primary focus is on the implications for the new US survey. Some broad conclusions about the 30 summarized hypotheses were reached. In general, demographic characteristics of residents (gender, age, education, socio-economic status, etc.) have no important impact on noise annoyance. As a result demographic characteristics do not explain differences between annoyance reactions in different geographical areas. Selected attitudes, on the other hand, have a consistently strong effect: fear of danger from the noise source, perception that authorities could better control the noise, and self-reported general sensitivity to noise. A change in noise exposure affects 20 Throughout this document, italics are used to identify significant issues likely to shape the survey implementation. H-1

reactions for road traffic and railway noise, but the effect on aircraft noise annoyance is uncertain. Ambient noise levels and time spent at home do not have an important effect on annoyance. Conclusions on many other issues are not clear for a variety of reasons: results are contradictory, study methods are weak, too few studies have been conducted, or no surveys have been conducted. In general, characteristics of geographic areas have been only occasionally studied, often with weak methodologies and almost no non-survey information about the areas. No studies have been located that presented evidence on how annoyance is affected by airport authorities’ actions, activities, or community relations programs. Studies have not examined the correlation between public complaints and private annoyance. Secondary Goals The secondary goals for a new national survey could be derived from the evidence from summary analyses in this literature review, knowledge about whether or not there are new methodological developments, estimates of costs, and a unique strength of the sample design - the large number of airports and neighborhoods to be studied. Of course all of these scientific and technical considerations must be weighed against policy judgments about the practical value of particular study goals. These various considerations suggest: a focus on the following factors that may help to explain the surveyed annoyance: characteristics of geographical areas, authorities’ actions, characteristics of community relations programs, relations to aircraft operations (landing/take-off/flight path location), and complaint rates. A focus on these types of goals should help to answer a major policy question: Is a single, national dose-response relationship justified because it is not possible to objectively predict deviations from a national average for local geographic areas? Or, alternatively, are there readily- available variables that predict differences between geographical areas and form a legitimate basis for local exceptions to national policies? Progress could also be made in identifying the size of the geographic units that are associated with unexplained variations in annoyance reactions. With a nested, clustered sample design it would be possible to begin to estimate the portion of the unexplained variance that is due to individuals’ situations (the variation between people in adjacent homes with identical exterior noise exposures), the portion that might be due to local factors (variation between nearby neighborhoods), the portion attributable to common aircraft operations (variation between larger areas exposed to similar aircraft operations) and the portion attributable to airport or city characteristics (the remaining differences between airports). Additional variables will need to be estimated for the national study, even if they have been studied before or can only be imprecisely measured, in order to provide some evidence as to whether such variables’ effects have or have not substantially biased the study results. Examples of such variables include demographic characteristics, recent changes in aircraft noise levels, estimates of ambient noise levels, meteorological conditions, and sensitivity to noise generally. Table 16 summarizes the data collection elements identified by the literature review. The remainder of this report develops these elements. Table 16 can also serve to organize a discussion of what elements should or should not be included in a new national annoyance survey. H-2

Table 16 Possible Elements of a Complete, Complex Annoyance Survey A. Primary Study Goal: Items that are required for the primary study goal 1. Resident questionnaire (Required to measure annoyance and the various demographic and attitudinal issues.) 2. Flight operation data and analysis (Required to estimate noise exposure and identify the characteristics of the operations near a respondent – relation to flight path, landings, etc.) B. New Contributions to Understanding Community Annoyance: Elements that contribute to understanding community differences 3. Clustering of respondents into blocks in the sample design. (Needed to identify the size of geographical areas associated with different reactions.) 4. Field experiments comparing questionnaire administration mode. (Used to provide a basis for comparisons to other surveys and dose-response standards that are based on different questionnaire administration modes.) 5. Questionnaire for airport officials (Used to identify characteristics of community relations programs, perceptions of community differences, the existence of legal actions or other major community conflicts with the airport) 6. Content analysis of local media (Used to estimate the impact of media and to rate the visibility of airport officials’ actions) 7. Analyze data from previous US surveys to create comparison for current survey (This will answer the question as to whether noise reactions have changed.) C. Controlling for Previously Identified Explanations: Elements that control for commonly assumed explanations for community differences 8. Meteorological data – short -term and long-term 9. Detailed airport complaint data on number of complainants, number of complaints, and location of complainants 10. Aircraft noise exposure estimates for each of the past five or ten years (Needed to control for changes in aircraft noise exposure) 11. Aircraft noise exposure data for the previous days or months (Used to test hypothesis that annoyance reactions are actually directed at a short period.) 12. Indicators of ambient noise levels, especially traffic noise level (Rough controls for ambient noise might come from geo-coded information about road type or population density.) D. Other Elements 13. Estimate outdoor to indoor attenuation of aircraft noise for each residence (A model to estimate attenuation based on respondent-provided information would need to be developed or, if already available, assessed.) 14. Supplemental sample of aircraft noise complainants (This would ensure sufficient complainants to compare complainants to non-complaining members of the population.) 15. Questionnaire data from knowledgeable authorities about the political activism of communities 16. Aircraft ground operation noise predictions H-3

Introduction Plans are being developed for a national survey of aircraft noise in the United States as part of an ACRP sponsored study. This report reviews the community noise-response literature for information that can contribute to the design of that survey. Since there have been over 600 prior social surveys of noise annoyance, a major objective is to determine what issues have been substantially settled by previous research and which unresolved issues should be studied to strengthen the new national survey. Primary study goal: Updating the United States dose-response relationship for aircraft noise The primary goal of a new national survey of aircraft noise in the United States is to update previous estimated dose-response relationships and provide a best estimate of the bivariate relationship between aircraft noise exposure and the annoyance of residents for the nation as a whole. At the present time the estimates that are used are based on averaging the results from transportation noise surveys that have been conducted since 1965 in at least eleven counties (Miedema and Vos, 1998; Schultz, 1978; American National Standards Institute., 2005; Federal Interagency Committee on Noise (FICON), 1992; International Standardization Organization, 2003). Some summaries of the dose-response relationship combine studies of all noise sources while others are based on only aircraft noise studies. No nationally representative survey of aircraft noise has been conducted in the US. The largest previous coordinated studies in the US were conducted at nine airports from 1967 to 1970 (Connor and Patterson, 1972; Patterson and Connor, 1973; Tracor Inc., 1971) Consequences of the research strategy needed to meet the primary goal Dose-response relationships are estimated by relating residents’ noise annoyance to the noise levels experienced at their home. The residents’ annoyance is measured in a questionnaire that is administered as part of a large social survey. The estimates of long-term aircraft noise environments are derived from measurements of aircraft noise or from aircraft noise estimation models. To obtain a statistically-sound national average it is necessary to obtain data from an adequate number of airports that have been selected on a probability (random) basis. Such an unbiased probability selection procedure also makes it possible to estimate the likely precision of the final measured dose-response relationship. In order to provide such an estimate, a sample of at least 30 airports is needed. The primary, national dose-response goal thus dictates a sample with a large number of airports. As is true of most aircraft noise studies, large and moderate size airports will be studied, not airports where there are large proportions of general aviation traffic. In addition the study will be confined to relatively high noise areas and thus to areas around airports. No other aircraft noise survey has been conducted at such a large number of airports. The large number of airports provides this study with a unique opportunity to study differences between geographic areas such as areas surrounding different airports, local areas near each airport, and neighborhoods. It is suggested that a secondary goal of the survey be to determine the extent to which dose-response relationships are uniform or different in different geographically-defined areas and, to the extent that areas differ, to identify factors that explain the differences. This goal addresses a fundamental noise policy issue: Are there local conditions that create important deviations from the average national dose- response relationship? Previous studies have only collected noise data and other data about individual respondents and have thus largely ignored airport and neighborhood differences that might or might not create different dose-response relationships. Very little attention has been directed at the effects of such geographically-specific variables as types of flight operations, airport authorities’ actions, community relations programs, media coverage, community organization, and meteorological conditions. With only 30 airports statistical tests of purely between-airport differences will be imprecise and there are likely to be confounding effects for correlated variables. In so far as possible, every effort needs to be made to base conclusions on within-airport/between geographical area contrasts, but it is likely that even though H-4

this 30-airport study provides improved, best-available evidence on many issues, the evidence will not be conclusive for many geographic area variables. Literature review strategy The literature review strategy for this report has been designed to efficiently focus on issues that will best address the study goals that can be addressed with the unique, multi-airport study design. With over 600 previous social surveys of reactions to environmental noise it is not feasible to review and create a new synthesis of the results from the associated thousands of publications. In addition the goal of this literature review is not so much to form best summaries of findings, but rather to report multi-study summary analyses where available and to identify relevant study design challenges for issues that have not been the subject of summary analyses. Relevant literature and findings have been identified through the following procedures: • Reading the entries in a catalog of noise surveys to identify surveys with relevant designs or study goals (Bassarab, Sharp, and Robinette, 2009). (This procedure was especially helpful in identifying multi-airport and multi-city studies.) • Searching the contents of major acoustical journals and conference proceedings to identify summary analyses that synthesized the results on some topics that could relatively easily be identified with search terms. • Using search engines to find keywords in the more than 1,000 publications from previous noise annoyance surveys that are locally available in PDF files. (This search was not a perfect search of all the publications because many articles had been optically scanned and could not be perfectly read by OCR (Optical Character Recognition) programs.) • Using search engines to search the internet for publications that are not in standard acoustical journals. • Reading reference lists in articles that were identified with the above procedures to locate additional articles and reports. (This was especially useful for identifying literature that was not in standard acoustical publications.) • Corresponding with international noise survey experts to locate information or confirm the absence of information on important community characteristics that had not been the subject of summary reviews. All the studies examine residents’ reactions to environmental noise experienced at their homes. Almost all the studies concern transportation noise, primarily aircraft and road traffic noise, though some are of railway noise and a few are directed at noise from neighbors and industrial sources. Of the 62 noise annoyance issues identified in the next section, about half have been addressed in summary analyses that used a results-neutral method to select publications for analyses. Many other issues have been discussed in detail and analyzed. The summary analyses included in the present literature review, however, are ones where the literature or data were selected using procedures that were not biased toward identifying positive findings. A bias toward positive findings can occur when the publication search procedure may miss minor reports that a variable that had no effect on annoyance while identifying all instances where a variable had a major effect on annoyance. This is a variant of the ‘file drawer’ problem that is encountered in experimental studies where studies with negative findings are left unpublished in a file drawer while studies with positive findings are prominently published. A similar problem could arise for studies of noise annoyance if only article abstracts are searched. A positive H-5

finding (e.g. annoyance increases with education) is likely to be featured in an abstract, while a negative finding (e.g. annoyance is not related to education) is likely to buried in a sentence or footnote in the body of a publication. The summary analyses utilized in this literature review are based on either reading all publications that could have reported results for the issue or on reanalyses of all the original, individual social survey data sets stored in a large data archive. Issues Reviewed This report concentrates on the factors, other than the acoustical characteristics of the noise source, which have been hypothesized to affect residents’ annoyance with noise. Table 17 lists 62 issues that were compiled from consulting six publications that analyzed or listed factors that might affect transportation noise annoyance (Fields, 1993; Kroesen, Molin, and van Wee, 2008; Miedema and Vos, 1999; Vader, 2007; Vos, 2010; Wyle Research, 2011). The 62 issues are grouped under 10 broad topical headings. Some focus on the individual respondent, such as demographic characteristics and personal attitudes. Others focus on the interaction between noise officials and residents: for example, attitudes toward authorities, authorities’ actions, and complaints directed to authorities. Others focus on local conditions such as neighborhood conditions or housing characteristics. Two issues concern the possible impact of the questionnaire administration mode on a survey’s findings about levels of annoyance. Although this list does not include all the issues that have been discussed in the noise literature, it does attempt to list all the issues that are likely to be relevant for the design of the new national survey. Several facets of a new national survey are not discussed. The choice of the primary dependent variable for a dose-response survey is not discussed. It is assumed that the annoyance response questions for the survey will be survey questions that were developed as part of a multi-national effort that involved coordinated studies in more than nine languages (Fields et al., 1997). The two following, almost identical, pair of questions have been published as an ISO Technical Specification (International Organization for Standardization, 2003) and are widely used in current noise annoyance surveys around the world: Q. Verbal. Thinking about the last (12 months or so), when you are here at home, how much does noise from (noise source) bother, disturb or annoy you: not at all, slightly, moderately, very, or extremely? Q. Numeric. Next is a 0-to-10 opinion scale for how much (source) noise bothers, disturbs or annoys you when you are here at home. If you are not at all annoyed choose 0; if you are extremely annoyed choose 10; if you are somewhere in between choose a number between 0 and 10. Thinking about the last (12 months or so), what number from 0 to 10 best shows how much you are bothered, disturbed or annoyed by (source) noise? It is anticipated that the final dose-response relationship will use one of these questions as the measure of response. Reporting a relationship with a single question provides an easily understood, transparent measure of noise impact that has come to be widely accepted for setting noise policies around the world. One test-retest study that conducted 97 repeated interviews concluded that “pure annoyance” questions were less reliable and valid than “general questions” about “perceived affectedness” or “dissatisfaction” (Job, Hatfield, Carter, Peploe, Taylor, and Morrell, 2001). Due to question placement and additional question-wording differences between the “simple annoyance” and “general questions” in that study, it is difficult to know exactly why the questions performed differently. It should be noted that the suggested H-6

questions above are not purely “annoyance” questions since they introduce the general annoyance concept with the phrase “… bother, annoy, or disturb…” Using a single annoyance question for the dose-response relationship does not, however, preclude the use of multi-question indices. Such indices should provide more precise estimates of the relative effects of different explanatory variables. One source of additional reaction questions could be questions from previous surveys with which a precise comparison is needed. The list of 62 issues also does not include a discussion of the form of the dose-response curve. Recent publications have argued that the shape of a dose-response curve can be supported by theory from studies of loudness ratings (Fidell, Mestre, Schomer, Berry, Gjestland, Vallet, and Reid, 2011). Whether the data from the new national survey are better represented by this curve or a logistic curve or some other shape can, of course, be tested at the analysis stage. The issue is not discussed here because it does not affect the study design. The list of 62 issues does not include the topics that are addressed in defining a noise index such as the relative importance of noise level and number of events, the impact of nighttime noise, the correct acoustical frequency weighting, and characteristics of individual noise events such as roughness or tonality. The acoustical data that will be available for the analyses will make it possible to consider some of these issues. However, it seems unlikely that much could be learned from the relatively homogeneous, long-term environments that are dominated by passenger jet aircraft at the moderate and large air carrier airports that will be randomly selected for the national survey. Making significant improvements in our knowledge about such noise-index issues would almost certainly require a survey design that would concentrate on unusual noise environments around unusual airports where normally correlated variables are not confounded. For example the impact of varying ratios of nighttime to daytime noise could only be estimated by including unusual airports or neighborhoods which have high nighttime to daytime flight ratios. Including multiple airports with such unusual exposure conditions would weaken the ability of the national survey to derive accurate dose-response relationships for the nation as a whole. Some of the survey data sets that are available for the study of such issues have been identified (Wyle Research, 2011). A secondary analysis of available data may be the most effective next step toward addressing these issues. A close examination of the literature for many of the 62 issues would, no doubt, identify additional sub topics within the issues and identify other labels for these and other closely-related issues. The selection of labels for the issues and the divisions between closely-related issues are sometimes arbitrary. It is hoped, however, that the level of detail is sufficient for this review. The ‘Table 18 Summary?’ column in Table 17 contains a ‘Yes’ if quantitative results are presented in Table 18for the issue. It should be noted, that many issues are only briefly discussed in this report. The bibliographic references that are given for each issue can provide more detailed information. The literature on these 62 issues is summarized in the next two sections of this report. The first section, Findings, discusses conventional issues about the effect of individual characteristics, local conditions, aircraft operations, and the method of survey administration. The Frameworks for assessing authorities’ actions section discusses a subset of the issues from Table 17 that are related to the effect that airport authorities' may have on annoyance. This separate section is needed to outline the framework for developing airport authority issues. These airport authority issues have not been systematically studied in previous surveys. H-7

Table 17 List of Issues Reviewed Issue # Issue Hypothesis a Table 18 summary? A. Effects of demographic characteristics 1 Gender Women are more annoyed Yes 2 Age - (Older age) Older age increases annoyance Yes 3 Age - (Middle age) Middle age increases annoyance Yes 4 Education High education increases annoyance Yes 5 Status-occupation High status increases annoyance Yes 6 Income High income increases annoyance Yes 7 Household size Mid-sized households (2 or 3 members) increase annoyance Yes 8 Length of residence Long-term residence decreases annoyance Yes 9 Home ownership Home ownership decreases annoyance Yes 10 Dwelling unit type Residents of single unit dwellings are more annoyed Yes 11 Usage of noise source Use of the transportation noise source decreases annoyance Yes 12 Economic connection Economic dependence on the noise source decreases annoyance Yes 13 Country/culture People from different cultures and countries differ in noise/annoyance reactions - - B. Effects of home conditions that modify exposure 14 Time at home Residents spending more time at home are more annoyed Yes 15 Exposure individualized - (sum below: out-of-doors , attenuation, orientation, season) Individuals with relatively less exposure are less annoyed Yes 16 Out-of-doors usage Residents who spend more time out-of-doors are more annoyed Yes 17 House attenuation Greater outside-to-inside transmission loss decreases annoyance Yes 18 Room orientation Quiet 'escape' rooms decrease annoyance Yes C: Effects of local community conditions 19 Community differences Airports and communities differ in annoyance responses Yes 20 Ambient noise levels Low-ambient noise levels increase annoyance Yes 21 Sparsely settled areas Rural or 'peaceful' suburban environments increase annoyance - - 22 Media coverage Positive or negative media coverage of the noise source creates corresponding annoyance reactions - - 23 Activist community Politically active communities increase annoyed - - 24 Meteorological conditions Comfortable weather (average climate or date-of-interview weather) increases annoyance Yes D. Effects of attitudes - general 25 Fear/Danger Fear of danger from a noise source increases annoyance Yes 26 Sensitivity General sensitivity with noise increases annoyance Yes 27 Importance of source Belief in the importance of the noise source decreases annoyance Yes 28 Preventability by authorities Belief that authorities could reduce noise increases annoyance Yes 29 Exposure control - individual Belief that the resident can control or avoid noise exposure decreases annoyance - - H-8

Issue # Issue Hypothesis a Table 18 summary? 30 Expectations for future exposure Expecting an increase in noise exposure increases annoyance - - E. Effects of attitudes toward authorities 31 Transparency of process The perception that authorities develop policy transparently and provide relevant information decreases annoyance - - 32 Fairness of procedures The perception that authorities follow procedures in a fair manner decreases annoyance - - 33 Trust The perception that authorities can be trusted decreases annoyance - - 34 Understanding of residents' concerns The perception that authorities understand or are concerned about residents decreases annoyance - - 35 Residents can affect policy The perception that authorities' actions are influenced by residents' views decreases annoyance - - F. Effect of aircraft operations 36 Distance to flight path Being under a flight path increases annoyance Yes 37 Landing operations Exposure to landing operations increases annoyance - - 38 Ground operations Noise from ground operations (including start of take-off noise) increases annoyance beyond the levels expected from airborne operations - - 39 Airport size Small airports create greater annoyance (adjusted for noise exposure) - - 40 Predictable noise profile A regular, predictable noise event profile decreases annoyance - - 41 Vibration Vibration of structures or rattles increase annoyance - - 42 Non-noise (other) Lights, odors, or other non-noise impacts increase noise annoyance - - 43 Change-Immediate impact Residents overreact to changes in noise exposure (either increase or decrease) Yes 44 Change-Long-term adaptation With time, annoyance with a changed noise exposure decreases Yes G. Authorities= actions and activities 45 Operator noise abatement actions Officials' programs to control noise decrease annoyance beyond levels expected from exposure - - 46 Community relations programs Strong community relations programs decrease annoyance - - 47 Conflict - history A history of noise operator/community conflict increases annoyance - - 48 Compensation to residents Receiving compensation from authorities decreases annoyance - - 49 Operators' perceptions Operators can predict residents' annoyance levels - - H. Correlated noise impacts 50 Vibration The belief that the noise source also causes vibrations increases annoyance - - 51 Health The belief that the noise source affects health increases annoyance - - 52 Air pollution The belief that the noise source's fumes or dirt pollute the air increases annoyance - - 53 Activity disturbance The belief that the noise interferes with daily activities (speech, concentration, listening, etc.) increases annoyance - - H-9

Issue # Issue Hypothesis a Table 18 summary? 54 Sleep The belief that the noise interferes with sleep increases annoyance - - I. Complaints and public actions 55 Complaint rate: annoyance High community complaint rates indicate relatively high annoyance - - 56 Complaint rate: other correlates What community or event characteristics are correlated with complaint rates? - - 57 Complainant characteristics Complainants' characteristics are representative of annoyed residents - - J. Other issues 58 Self-selection by moving away High noise levels cause annoyed residents to move out of the area Yes 59 Long-term annoyance trends Over the years noise annoyance has increased for the same noise exposure Yes 60 Survey administration : Telephone vs. Face-to-face A telephone interview yields higher annoyance ratings (than face-to-face) Yes 61 Survey administration : Self- vs. interviewer-administered Self-administered questionnaires (mail) yield higher annoyance ratings - - 62 Relevant exposure period Respondents' annoyance is determined by recent experiences not by the entire previous year - - Footnote to Table 17: In the interest of brevity, the hypotheses have all been stated in the form ‘X increases annoyance’ – a form which seems to imply a causal connection. In fact these are hypotheses about associations, not causation. Thus the listed variable might only be associated with some other variable that is a real cause – for example, aging might not cause annoyance, but a particular age cohort might have experienced some event that affected annoyance. In other instances the annoyance might be the cause. For example, feelings of extreme annoyance might lead some people to conclude that their health is affected by noise. Findings This section begins with a description of Table 18 where the results for the 30 issues with summary analyses are displayed. The remainder of the section discusses each of the 62 issues listed in Table 16. Two different types of analyses provide the results that are presented in separate columns in Table 18. The two types of analyses are: meta-analysis and secondary analysis. Both types of analyses provide conclusions on the single issues by systematically and uniformly drawing on the data from a large number of studies that addressed the issue. A meta-analysis is an analysis that is based on a systematic analysis of publications. A secondary analysis is a re-analysis of the original, individual- level questionnaire responses from a number of different surveys using a single comparable analysis strategy for all the surveys. One type of summary analysis in Table 18 is a count of the number of surveys that support or do not support a hypothesis. For a single study a hypothesis is considered to be supported if an effect is large enough to be “important” or is statistically significant. For most summaries an “important” effect is the equivalent of a 3-decibel change in noise exposure or a 5% difference in annoyance. The summaries based on this survey counting approach are provided in the column labeled “Do most studies support?” If more than 50% of the studies support a hypothesis a “YES” appears in that column. The next two columns report the number of studies [Studies (N=)] and number of respondents from those studies [Respondents (N=)]. Thus for Issue 4, the hypothesis that “High education increases annoyance” was not supported by more than 50% of the 18 studies (Studies N=18) and these 18 studies contained a total of 29,893 respondents. The source of the finding (Fields, 1993) is cited in the next-to-last column (Source H-10

for study support count). This meta-analysis by Fields counted ‘important’ effects where a study’s reported effect was considered to be ‘important’ “if it was the equivalent of a three-decibel difference in noise exposure or statistically significant.” Additional details for the analysis methods are given in the cited publication which gives the objective rules followed for combining diverse published results. The publication also describes sensitivity analyses that determined whether considering the relative sizes or methodological quality of the analyzed surveys would lead to different conclusions. For the particular summaries prepared on the issues in Table 17 in the meta-analysis columns, the study counts are based on meta-analyses, except for the counts for the noise-change Issues #43 and #44 which are based on secondary analyses. The other summary analysis method in Table 17 is to calculate a measure of the size of the effect across all of the diverse data sets. For noise annoyance surveys, these types of summaries are only available from secondary analyses of the original study questionnaire data. The results for this summary method are given in the “Average effect (dB equivalent)” column of Table 17. These results are based on the number of studies and respondents given in the two following columns. The source of the summary analysis is then provided in the last column of Table 17. For Issue 4, “Education”, these columns show that this secondary analysis examined the answers from 30,427 respondents drawn from 26 different surveys. When the difference in the responses of the low education (not completed secondary school) and high education (completed university or higher) respondents were compared it was found that the percentage difference in annoyance scores was the same percentage difference that would be generated by a 2- decibel difference in noise exposure. This small effect is thus consistent with the meta-analysis results in the “Do must studies support?” column which also found that most studies did not find an important effect. The summary secondary analyses that are drawn upon in this report consider the studies to be substantially equivalent in the sense that they do not test to determine whether there are statistically significant interactions between country or survey and the variables that are the basis for the issues in this review. For example, there is no test of the possibility that education might affect annoyance in some countries but not in others. A simple comparison of results from diverse surveys from different countries could be misleading because the country-of-survey could be confounded with the many methodological differences between surveys. This literature review provides only the simplest, condensed results from the summary analyses. For example, Table 17 reports support for a hypothesis that two-member and three-member households are more annoyed by noise (Issue 7). The summary publication gives some important additional information such as the fact that this household-size effect was tested and found to not be due to correlations with age and that single-person households were considerably less annoyed than the large households of four or more members (Miedema and Vos, 1999). Additional details about the analysis methods used in the secondary analyses, including the definition of the subgroups that are compared for each variable, are not provided in the present report but are in the cited publications. All of the analyses in Table 18 are based on analyses that have controlled for noise exposure. In every case, the findings should be interpreted as evidence about the effect of a variable on residents at the same noise level. H-11

Table 18 Summary from Meta-analyses and Secondary Analyses Issue # Issue Hypothesis Meta-analysis results Secondary analysis results Source for study support count Source for average effect Do most studies support?a Studies (N=) Respon dents (N=) Average effect (dB equivalent)b Studies (N=) Respon dents (N=) A. Effects of demographic characteristics 1 Gender Women are more annoyed NO 47 62,479 0 dB 34 38,255 Fields, 1993 Miedema and Vos, 1999 2 Age - (Older age) Older age increases annoyance NO 63 77,122 Fields, 1993 3 Age - (Middle age) Middle age increases annoyance 3 dBc 47 62,983 Van Gerven, Vos, Van Boxtel, Janssen, and Miedema, 2009. 4 Education High education increases annoyance NO 18 23,983 2 dB 26 30,427 Fields, 1993 Miedema and Vos, 1999 5 Status-occupation High status increases annoyance NO 22 33,701 2 dB 23 27,247 Fields, 1993 Miedema and Vos, 1999 6 Income High income increases annoyance NO 10 15,846 Fields, 1993 7 Household size Mid-sized households (2 or 3 members) increase annoyance 2 dB 27 29,993 Miedema and Vos, 1999 8 Length of residence Long-term residence decreases annoyance NO 44 61,322 Fields, 1993 9 Home ownership Home ownership decreases annoyance NO 23 25,327 2 dB 25 29,463 Fields, 1993 Miedema and Vos, 1999 10 Dwelling unit type Residents of single unit dwellings are more annoyed NO 14 18,463 Fields, 1993 H-12

Issue # Issue Hypothesis Meta-analysis results Secondary analysis results Source for study support count Source for average effect Do most studies support?a Studies (N=) Respon dents (N=) Average effect (dB equivalent)b Studies (N=) Respon dents (N=) 11 Usage of noise source Use of the transportation noise source decreases annoyance NO 6 12,089 2 dB 12 16,800 Fields, 1993 Miedema and Vos, 1999 12 Economic connection Economic dependence on the noise source decreases annoyance NO 12 16,364 2 dB 14 21,516 Fields, 1993 Miedema and Vos, 1999 B. Effects of home conditions that modify exposure 14 Time at home Residents spending more time at home are more annoyed NO 17 19,765 Fields, 1993 15 Exposure individualized - (summary of Issues 16, 17, 18 & five season studies)i Individuals with relatively less exposure are less annoyed YES 30 39,119 Fields, 1993 16 Out-of-doors usage Residents who spend more time out-of-doors are more annoyed NO 2 1,000 Fields, 1992 17 House attenuation Greater outside-to-inside transmission loss decreases annoyance Mixed support 14 18,725 Fields, 1992 18 Room orientation Quiet 'escape' rooms decrease annoyance YES 9 8,522 Fields, 1992 C: Effects of local community conditions 19 Community differences Airports and communities differ in annoyance levels YES 19 55,000 7 dB 19 55,000 See cell to right Fields, Ehrlich, and Zador, 2000 20 Ambient noise levels Low-ambient noise levels increase annoyance NO 16 15,512 1 dB d 20 57,000 Fields, 1996; p.37-39 Fields, 1996; 24 Meteorological conditions Comfortable weather (average climate or date-of-interview weather) increases annoyance YES 7 18,043 1-3 dB 41 51,130 Fields, 2004 Miedema, Fields, and Vos, 2005 H-13

Issue # Issue Hypothesis Meta-analysis results Secondary analysis results Source for study support count Source for average effect Do most studies support?a Studies (N=) Respon dents (N=) Average effect (dB equivalent)b Studies (N=) Respon dents (N=) D. Effects of attitudes - general 25 Fear/Danger Fear of danger from a noise source increases annoyance YES 21 44,713 19 dB 12 17,494 Fields, 1993 Miedema and Vos, 1999 26 Sensitivity General sensitivity with noise increases annoyance YES 23 36,435 11 dB e 29 33,977 Fields, 1993 Miedema and Vos, 1999, Miedema and Vos, 2003 27 Importance of source Belief in the importance of the noise source decreases annoyance YES 4 5,882 Fields, 1993 28 Preventability by authorities Belief that authorities could reduce noise increases annoyance YES 11 19,462 Fields, 1993 E. Effects of attitudes toward authorities F. Effect of aircraft operations 36 Distance to flight path Being under a flight path increases annoyance NO 3 3,230 Fields, 1993 43 Change-Immediate impact Residents overreact to changes in noise exposure (either increase or decrease) YES 43 - - f Brown, van Kamp, 2009 44 Change-Long-term adaptation With time, annoyance with a changed noise exposure decreases NO 14 - - f Brown, van Kamp, 2009 G. Authorities= actions and activities H. Correlated noise impacts H-14

Issue # Issue Hypothesis Meta-analysis results Secondary analysis results Source for study support count Source for average effect Do most studies support?a Studies (N=) Respon dents (N=) Average effect (dB equivalent)b Studies (N=) Respon dents (N=) I. Complaints and public actions J. Other issues 58 Self-selection by moving away High noise levels cause annoyed residents to move out of the area NO 7 5,877 Fields, 1993 59 Long-term annoyance trends Over the years noise annoyance has increased for the same noise exposure 8 dB g 22h 42,078 Janssen,Vos, van Kempen, Breugelman s, and Miedema, 2011 60 Survey administration : Telephone vs. Face- to-face A telephone interview yields higher annoyance ratings (than face-to-face) NO 4 3,393 Fields, 1993 Footnotes to Table 18: a. Most of these summary counts of study results come from one report (Fields 1992). For this report a study was counted as “supporting” a hypothesis if the published data indicated an “important” effect. The definition of an ‘important effect’, in order of precedence is: a difference in annoyance reactions that is the equivalent of the difference created by a 3-decibel difference in noise level, a difference of at least 5% in the percent annoyed, an explanation of at least 1% of the variance in annoyance, an effect that is statistically significant at the p<.05 level. A complete explanation of the counting protocol is available in the report. b. The average decibel equivalent effect is the difference in noise exposure that creates the same difference in annoyance as that which is observed between subgroups for the variable being studied. c. For the ‘middle age’ hypothesis those in the 30 - 49 age group were more annoyed than either younger or older respondents (Van Gerven, Vos, Van Boxtel, Janssen, and Miedema 2009). This most recent and comprehensive review, cited in Table 18, supports this pattern and provides a model to predict age effects but does not provide observed annoyance scores for different age groups and does not provide a measure of the decibel equivalent of the differences between age groups. As a result the estimated differences in Table 18 are the differences estimated between the age 30-40 age group and the age 70+ age group that were presented in an earlier summary secondary analysis (Miedema and Vos 1999). d. The best estimate from this secondary analysis is that a 20-decibel lower ambient noise creates higher annoyance that is the equivalent of a 1-decibel increase in noise from the rated noise source. H-15

e. The estimate that sensitivity has an effect that is the equivalent of an 11 decibel difference in noise exposure comes from an analysis that analyzed 15 surveys with 14,294 respondents which assumed that the impact of sensitivity was the same at all noise levels (Miedema and Vos 1999). When the analysis was repeated at a later date with additional surveys, and a total of 29 surveys and 33,977 respondents, an interaction of noise level and sensitivity was found that indicated that the influence of sensitivity increased with increasing noise level (Miedema and Vos 2003). f. The number of respondents included in the meta-analysis is not reported (Brown and van Kamp 2009). g. The value of 8 dB is the estimated decibel equivalent of the increase in annoyance scores over the 30 years between 1970 and 2000.Using the year coefficient (0.55) in Table 18from the analysis that controls for survey administration mode (Janssen, Vos, van Kempen, Breugelmans, and Miedema 2011), it is estimated that annoyance increases by a value of about 16.5 (30*0.55) annoyance points. This is the equivalent of approximately an 8dB increase in noise level as the estimate from a linear regression equation is that annoyance increases at a rate of approximately 1.98 annoyance score points per decibel (Personal communication with Sabine Janssen, January 27, 2012). h. Some of the 22 studies included 34 airports because some studies included more than one city. i. The summary for Issue 15 is based on the x studies reported for Issues 16, 17 and 18 as well as on 5 studies of seasonal effects that were available when this summary was reported in 1992. H-16

Topic A: Effects of demographic characteristics Issues #1 to #12: The first 12 issues in Table 18 concern the basic demographic characteristics of the population: gender, age, education, social status (as measured by occupation), income, household size, length of residence, home ownership, type of dwelling unit (single family or multiple-family), extent of usage of the transportation noise source, and having an economic connection to the noise source. All of these issues have been examined in systematic summary analyses that could be expected to identify prominent as well as less prominent reports of findings. The results for all 12 are summarized under Topic A of Table 18 . These demographic variables have very small effects on noise annoyance. In no case did more than half of the individual studies find an important effect of a demographic variable. When all of the results were pooled in the secondary analyses the estimated average difference between subgroups was never more than three decibels for any variable. Such small differences mean that any differences in annoyance in different neighborhoods or cities are not likely to be explained by demographic differences such as differences in income, education, employment by the noise source, or home ownership. While a new survey in the US is not likely to change these conclusions, there is a strong case for measuring most of these demographic variables in the questionnaire. By measuring the variables it will be possible to convince skeptics that the variables are not important in the US and to state that any demographic differences between areas have been controlled. In addition, these are relatively cheap and accurate variables, new questions do not need to developed, the questions take very little interview time, and they are easily analyzed. Issue #13: Country/Culture No systematic reviews have been located that examine all the evidence on cultural effects either between countries or, within one country between subcultures or immigrants from different countries. The Netherlands’ TNO data archive contains data from many countries but no analyses have been published that examined differences between countries. Differences in survey methodology and differences between locations within a single country make accurate comparisons between countries uncertain. Though questionnaires do sometimes record ethnicity and country of origin, this review has not attempted to locate relevant findings. In the absence of large ethnic groups the number of minority group respondents is likely to be too small to provide accurate estimates. Between country difference could be important. Summaries of surveys in Western countries have consistently found a ‘railway bonus’ – annoyance with railway noise is less severe than that with aircraft or road traffic noise at the same level (Fields and Walker, 1980;Miedema and Vos, 1998b). However, surveys in Japan have consistently found that railway noise is more, not less, annoying than other transportation noise sources (Yano, Sato, and Morihara, 2007). Topic B: Effects of home conditions that modify exposure Issue #14: Amount of time at home The results in Table 18 for the first issue under Topic B, time at home, indicate that the annoyance with aircraft noise at home is not reduced for individuals whose total exposure at home is less because they are away from home more hours a week. Issues #15 – #18: Other indicators of total noise exposure while at home The second issue under Topic B, Issue 15, is an overall summary of all the studies summarized in Issues 16 to 18 as well as five studies of season effects. The common thread in all of these issues is the H-17

hypothesis that an individual’s reaction to the noise level as represented in official regulations (the noise level outside the house at a fixed position relative to the noisiest location near the dwelling) is modified by the sum of the noise that is received at an individual’s ear. The summary for all indicators in Table 18 finds that individualized exposure does affect annoyance. In an attempt to better understand the phenomena, the component studies are divided into smaller groups by their type of individualized exposure in the three following lines in Table 18 (Issues 16 – 18). For Issue 16 in Table 18, there are only two studies and the evidence is weak: one study finds time spent outside increases annoyance the other study finds outside exposure has no effect. For Issue 17, the effect of acoustical isolation of the house structure, the evidence is very mixed and weak. Whether half the evidence supports the hypothesis depends upon whether the number of studies is counted or whether the studies are weighted according to their sample sizes. However, these studies may not provide a strong test because most of the measures of acoustical attenuation are weak. The respondent’s report on whether there is any double glazing is often the only basis for estimating attenuation. In other studies, some respondents had received free insulation as a part of a noise reduction program. However, as one review of insulation effects notes (Amundsen, Klaeboe Ronny, and Aasvang, 2011), for such noise reduction programs it is unclear whether residents’ annoyance is reduced because of the increased acoustical insulation or the fact that the residents were beneficiaries of a program that demonstrated the authorities’ concern about noise. A substantial improvement in the method of measuring acoustical attenuation could provide strong new evidence on this topic. At least two additional studies have been conducted since the meta-analysis in Table 18 was published. One showed no benefit from a noise insulation scheme (Fidell and Silvati, 1991). The other, a before/after study, showed a reduction in annoyance that was consistent with the reduction in noise level (Amundsen, Klaeboe Ronny, and Aasvang, 2011). This should not be regarded as a thorough up-dating of the previous meta-analysis since the keyword search of publication abstracts and titles might not have identified some studies, especially ones with negative findings. For the remaining issue under Topic B (Issue 18, Room Orientation) , Table 18 presents evidence that annoyance is reduced in dwellings where major rooms are acoustically sheltered by being on the side of the dwelling that does not face a noise source. This issue is mainly relevant for ground-based noise sources where one side of the dwelling often has much higher exposure than the other. As a result the room-orientation issue is not particularly relevant for planning a new national survey of aircraft noise. Not surprisingly, none of the nine studies summarized in Table 18 for this issue are of aircraft noise. Topic C: Effects of local community conditions Issue #19: Community differences There is clear evidence for differences in reactions in different geographical areas, but the extent to which these differences occur at a city level, airport level, or local neighborhood level has not been quantified. Comparisons of dose-response relationships from different surveys are often cited as evidence for differences in reactions (Fidell, Mestre, Schomer, Berry, Gjestland, Vallet, and Reid, 2011), however, it is not clear whether differences between these studies are due to genuine differences in reactions or to methodological differences between studies which were conducted by different survey organizations using different modes of administration ( for example mail or telephone) with different questionnaires that asked different survey questions which were administered in different seasons in different years in different languages. Much stronger evidence comes from studies which gather survey data from many communities as part of a single study. The results presented in Table 18 for this issue come from analyses of 19 surveys, each of which compared community response between subareas within their single survey (Fields, Ehrlich, and Zador, H-18

2000). These 19 surveys were all the surveys that the researchers could obtain from archives which sampled at least 20 geographical areas, identified those areas on the respondent-level data file, and provided the individual respondents’ ratings of annoyance on a multi-point annoyance scale. Six of the surveys measured noise reactions to two sources. The total resulting 55,000 annoyance reactions were analyzed and it was found that after controlling for noise level there were greater differences between annoyance rates in geographic areas than would be expected from the differences between respondents within areas. The study found community differences in every one of these 19 studies. The measure of the size of the community difference in Table 18 comes from the same summary analysis. The 7 dB average is the average of the 20 surveys’ estimates of the standard deviation of the subareas’ annoyance scores, expressed as the decibel equivalent value of the annoyance scores. Most of the publications that propose explanations for community differences using a single survey are based on contrasts of a small number of communities. Many of these find the annoyance is different in two communities, then identify a single other difference between the areas, but do not consider other possible explanations for differences. Differences in communities demographic characteristics are readily available from census data and are often reported to be associated with differences in community reactions. However, as the results from summaries in Table 18 demonstrate, social surveys have repeatedly found that an individual’s demographic characteristics (age, income, education, gender) have too small an effect on annoyance to explain large community differences. For example, it is often assumed that high income can explain the difference between annoyance reactions of neighborhoods, but analyses of individual respondents income, education or other socio-economic status show only small effects (no more than 2 dB for individuals in Table 18) – much too small to explain major differences in the average annoyance in different communities. The difficulties in explaining annoyance differences between small numbers of airports are well illustrated by social surveys that were conducted around nine US airports around 1970 (USA-022, USA- 032, USA-044). After conducting social surveys at seven large airports, the last phase of the study turned to two smaller airports to determine whether annoyance reactions might be different at smaller airports (Patterson and Connor, 1973). After adjusting for noise exposure, it was found that respondents were less annoyed at these last two airports, but when the researchers considered many other characteristics of the airports they realized that these last two small airports’ differed in other ways: the surveys were conducted at a cooler time of year (winter not summer), the populations were exposed to different types of operations (noise from take-offs dominated the noise environment at only the two small airports) and there was much less social interaction focused on aircraft noise. In short, though the two additional airports were originally chosen to represent a particular characteristic (airport size), the other differences between the two airports and the seven original airports made it impossible to come to a conclusion. A study of six airports in Spain came to a similar conclusion when authors offered hypotheses for airport differences that considered recent changes in exposure, the extent of military usage and the proportion of tourism flights (García, Faus, and García, 1993). For a national survey, it is clear that there needs to be a large number of airports and that, when possible, airport differences should be controlled by using identical survey methods at all locations and, when possible, testing hypotheses by comparing neighborhoods or small communities within airports. Issue #20: Ambient noise Table 18 presents the evidence that in normal residential areas, ambient noise levels do not have an important effect on noise annoyance with a specified noise source, such as aircraft. The best estimate from this secondary analysis of 20 surveys with 57,000 interview responses is that a 20-decibel lower ambient noise creates the equivalent of a 1-decibel higher annoyance with the rated noise source. This is such a small effect as to not be relevant for most noise regulations. This small effect for private H-19

annoyance is not inconsistent with the observation that very active and effective anti-noise organizations may come from low-ambient areas and successfully organize public actions to try to modify noise policy. At least one additional study has been published that reports an ambient noise effect (Lim, Kim, Hong, and Lee, 2008), but an updated comprehensive review of the literature has not been conducted. Given the results from the previous 20 studies, it seems unlikely that there have been enough new studies with different conclusions to alter the judgment that ambient noise has little or no effect. Issue #21: Sparsely settled areas Although ambient noise does not by itself create higher reactions, it may still be hypothesized that annoyance will be much higher for rural populations or in areas in which there are widely spaced country estates. Three of the ambient noise studies in the previous meta-analysis classified some areas as rural although the definition of a rural area was not provided and might have included densely settled villages within a rural countryside. None of the three studies reported a 5% or greater level of annoyance in the rural areas (Borsky, 1965; Hawkins, 1980; Wehrli, Nemecek, Turrian, Hofmann, and Wanner, 1978). A more recent study of wind farms in the Netherlands reported, contrary to the hypothesis, that annoyance was lower, not higher in rural areas without main roads (Pedersen, van den Berg, Bakker, and Bouma, 2009). Consultants often see apparent support for the hypothesis because there are otherwise quiet, sparsely settled areas where well-organized residents complain about low levels of noise exposure from a single source. However, no studies have determined whether the private annoyance levels are particularly high in these areas or even if the number of such areas that complain about aircraft noise is large relative to the possibly enormous total numbers of such areas exposed to similar low-aircraft noise environments. Issue #22: Media coverage Although publications often note that media reports and other publicity about noise exposure may affect a community’s reaction to noise, these publications almost always concern a single, complex situation and are not able to separate the effects of media attention from other factors. For example, a report on reactions to changed aircraft noise exposure around a Sydney, Australia airport notes that pre-change media reports may have affected annoyance reactions (Job, Topple, Carter, Peploe, Taylor, and Morrell, 1996a). However the principle author, Soames Job cautions that this study of a single situation does not provide proof (personal correspondence with Soames Job, December 2011). A suggestion that media may not be important comes from the finding that annoyance did not increase between a 2001 survey and a 2003 social survey around Zurich-Kloten airport despite greatly increased media attention due to planned changes at the airport (Brink, Wirth, Schierz, Thomann, and Bauer, 2008). Several studies provide evidence that does not rely on a single city or location for information about the possible effect of publicity. In the only multi-city study that was located in the literature, a study about reactions to sonic booms found a weak relationship between measures of media coverage and residents’ attitudes toward sonic booms (Tracor Inc., 1970). This sonic boom study could prove useful for future planning since it describes the methods used to analyze the content of media reports. More indirect evidence comes from a social survey in which residents’ annoyance responses were weakly correlated with their self-reports as to whether they were “… aware of any recent comments or articles in the newspapers or on TV concerning the local airport” (Le Masurier, 2007). This is not, however, strong evidence about the effect of media coverage because it relies on the respondent’s recall of media exposure to a specific topic. Some support for a media effect comes from an experiment in Hong Kong in which subjects were given either a negative or positive fact sheet on noise exposure from a railway line; the two versions having been compiled from a collection of negative or positive newspaper articles about noise-mitigation measures for the newly opened railway line (Lam and Chan, 2007). When the subjects were tested a month later it was found that those who had read the positive fact sheet were slightly less annoyed. H-20

Other support comes from a German study in which two communities that experienced very small reductions in railway noise from rail grinding differed in whether or not they were exposed to positive publicity from a leaflet, press report or other local contact. Only those exposed to the positive publicity expressed a statistically significant reduction in railway noise annoyance (Liepert, Hegner, Möhler, Schreckenberg, Schümer-Kohrs, and Schümer, 1999). Indications of the limits of media effects come from an airport noise study in England where residents were asked about a special nighttime operation trial. Despite a publicity campaign and media coverage of the sensitive night-flight trials, the social survey found that most residents were not aware of the trials (Flindell and Witter, 1999). Although a comprehensive, detailed review of the literature has not been conducted, it appears that media may have some effect on reactions. This review also suggests that strong evidence could come from correlating a survey of residents’ annoyance with independent content analyses of local media coverage. Although most of this evidence could come from comparisons between airports, differences between neighborhoods might be explored if some of the media coverage was focused on issues that only affected specific neighborhoods. Issue #23: Activist community Communities and, to a lesser extent, cities develop reputations as centers of political and environmental activism. It is hypothesized that these types of communities would sensitize residents to noise and other environmental issues and thus create higher noise annoyance among individuals. No studies have been located which test this hypothesis for noise annoyance. Issue #24: Meteorological conditions From the summary reported in Table 18 meteorological conditions appear to have at least a small effect on reactions. The secondary analysis of 41 surveys conducted in different countries at different times of year estimated the effect that a 15 degree (centigrade) difference in temperature would have on noise annoyance. A 15 degree range was chosen because this is the approximate difference between summer and winter. The summary analysis estimated that the 15 degree difference created a difference in annoyance that was the equivalent of a 1-db to 3-db difference in noise exposure depending upon the type of analysis that was conducted. However, these estimates were not statistically significant (p>0.05) (Miedema, Fields, and Vos, 2005). The large confidence intervals, ranging from -2 to +6 dB for a logistic regression of “high annoyance are almost certainly created by the fact that surveys from different countries and cities are being compared rather than surveys which were conducted of the same population at different times of year.” A study which surveyed the same population continuously for seven years, the population of the Netherlands, found a consistent, statistically significant seasonal pattern with the greatest annoyance in July, August or September (Miedema, Fields, and Vos, 2005) However, the difference in annoyance reactions could not be confidently expressed in terms of the equivalent decibel impact because that survey used an unusual, unique annoyance scale and the scales’ relationship to noise level has not been analyzed. One set of assumptions generated the estimate that the 15 degree difference in temperature would be equivalent to a one-decibel difference in noise exposure. There was no evidence that long-term reactions are more strongly influenced by meteorological conditions on the interview day or the immediately preceding days than by the meteorological conditions that prevailed in the weeks and months preceding the interview. The independent effects of temperature and other meteorological variables such as precipitation or wind speed could not be accurately estimated from the data available for the summary analysis. Meteorological conditions could be an important consideration in planning a national survey, because shared conditions at a single airport could increase airport-to-airport differences and because between- H-21

airport differences could be moderated through the timing of the survey administration. For a national survey the best strategy is probably to conduct all interviews at the same time of year, when the temperature differences in the previous three months differ the least around the country. Topic D: General attitudinal variables Issue #25: Fear/Danger Every one of the 21 studies reviewed in Table 18 found that noise annoyance is associated with a respondent’s perceptions of danger from aircraft or fear of airline crashes. The difference in the reactions of high-fear and low-fear respondents is the equivalent of a 19 decibel difference in noise exposure. Though fear is clearly important, the effect may have been overestimated in this analysis because some respondents who reported they did not ‘hear’ aircraft were not asked about fear. Fear is likely to be correlated with noise level as was reported for recent surveys around Schiphol (Janssen, Vos, Houthuijs, van Kamp, Breugelmans, and Miedema, 2010). Some evidence for a linkage between fear and airline accidents comes from a study that found that residents near a recent aircraft crash site were more fearful and annoyed than other residents at the same airport who were not near the crash site (Moran, Gunn, and Loeb, 1981). A primary reason for including questions about fear in the new national questionnaire is to understand reactions under flight paths. Being under flight paths would be expected to increase respondents’ perceptions of danger to their personal safety even if it did not affect annoyance. The fear questions from previous surveys, however, are often ambiguous as to whether the fear derives from perceived threats to the respondents’ personal safety or fear about an airliner crash anywhere in the area. Some relatively complex survey questions ask about fear from “aircraft flying overhead”, but it is uncertain whether respondents limit their response to only those aircraft flying directly over them. New survey questions could distinguish between fear from all aircraft and from only the aircraft perceived as directly overhead which could directly threaten the respondent. Issue #26: Sensitivity All but one of the 24 studies reviewed in the meta-analysis summary in Table 18 found that self-reported sensitivity to noise is related to noise annoyance with a specified noise source such as aircraft noise. The estimate that sensitivity has an effect that is the equivalent of an 11-decibel difference in noise exposure comes from a secondary analysis that analyzed 15 surveys with 14,294 respondents which assumed that the impact of sensitivity was the same at all noise levels (Miedema and Vos, 1999). When the analysis was repeated at a later date with additional surveys and a total of 29 surveys with 33,977 respondents, sensitivity was still found to be an important variable but an interaction of noise level and sensitivity was found that indicated that the influence of sensitivity increased with increasing noise level (Miedema and Vos, 2003). This interaction effect does not appear to be consistent across surveys. The same analysis that found a very large interaction effect for the secondary analysis of 28 of the 29 surveys, found a much smaller interaction effect for one of the 29 surveys, an especially large survey of 10,939 respondents around Schiphol airport (Miedema and Vos, 2003). Other investigators examined large surveys from three different airports and found no evidence of interaction at any of the three airports (van Kamp, Job, Hatfield, Haines, Stellato Rebeccas K., and Stansfeld, 2004). Whether there is a strong interaction effect or none, it is clear that sensitivity has a strong effect on noise annoyance. The 1993 Miedema and Vos article summarizes other important finding about sensitivity. These additional analyses of sensitivity confirm that the degree of sensitivity is only weakly, if at all, related to noise level. The social survey results are all based on respondents’ answers to survey questions. The three most common measures are based on 1) a single global question about whether the respondent H-22

believes he/she is sensitive to noise generally, or 2) a set of questions about annoyance with common sounds (e.g. household appliances, fingernails on a blackboard, etc.) or 3) reactions to noises in specific situations. All measures are shown to affect annoyance with transportation noise in the residential situations. At least one laboratory study has shown that measures of auditory sensations or perceptions of noise levels are not related to the type of affective, self-reported sensitivity that is measured in survey questionnaires (Ellermeier, Eigenstetter, and Zimmer, 2001). Some early social surveys assumed that noise sensitivity was a form of neuroticism (McKennell, 1963). Miedema and Vos reviewed the current psychological literature and concluded that sensitivity and neuroticism and sensitivity are correlated but independent traits (Miedema and Vos, 2003). Noise sensitivity is an important determinant of aircraft noise annoyance and should probably be measured the new US questionnaire. However, noise sensitivity appears to be too evenly spread a personality trait to help explain differences in annoyance in different areas and thus should probably not be a major focus for the new US study. Issue #27: Importance of noise source For Issue 27 in Table 18, three of the four studies found that the belief that the airport or flights are important reduces noise annoyance. Most of the ‘importance’ survey questions either imply or directly refer to the economic importance of the airport by asking about the importance for the region or city or other unit. Issue #28: Preventability (authorities) A range of labels are used for a variety of social survey questions that tap into residents’ general feelings about whether authorities have failed to take steps which could reduce aircraft noise levels. This concept has been given such labels as ‘preventability’, ‘malfeasance’, ‘misfeasance’, and ‘considerateness’. Support in Table 18 for this hypothesis is drawn from 11 surveys. All of these surveys found that the belief that the noise could be reduced was associated with greater annoyance. Some surveys have attempted to determine which air transport authorities might be perceived as being able to reduce noise by asking parallel questions about such groups as pilots, airlines, government, airport operators, or the designers of aircraft. However, the present literature search has not identified detailed discussions about theories or correlates of preventability. Attitudes toward preventability are likely to contain components of various attitudes toward authorities which are discussed under Topics E and G. Issue #29: Exposure control (respondents) Several studies have attempted to test the hypothesis that stress leading to annoyance occurs because residents are unable to develop an effective strategy for coping with the noise. This hypothesis is derived from Lazarus’s theory that stress occurs when individuals believe they do not have the resources to cope with environmental stressors (Lazarus, 1966). In the noise literature this set of related concepts is explored in studies of ‘coping strategy’ or ‘perceived control’. No systematic searches of the noise survey literature have been conducted that summarize relevant studies. Guski states that “Central to the coping concept is the belief and confidence of an affected person that he/she will somehow manage the problem. The coping strategy can be direct (e.g., in turning off the noise source, or negotiating with the people responsible for the stress) or indirect (mostly via cognitive control, e.g., by means of an exact knowledge of the time schedule of the noise source). Mostly, environmental noise sources cannot be turned off directly, but they could be negotiated, and indirect coping strategies can also be very effective in reducing the noise annoyance” (Guski, 1999). According to Guski, one study found that self-rated reports of being able to adjust to the noise were as closely related to noise annoyance as was sensitivity (Finke, Guski, and Rohrmann, 1980). Analysis of a large survey near H-23

Schiphol also found that perceived control and coping capacity are closely related to noise annoyance (Kroesen, Molin, and van Wee, 2008). While there is a psychological literature supporting the theory of coping and control (Lazarus, 1966), it is not clear whether some of the survey measures may be tautological. One item reported by Guski illustrates this problem: "If it is too loud outside, I simply close the windows, and then I am no longer disturbed". This questionnaire item seems to be close to stating that “I am not annoyed because I am not annoyed”. The “coping and control” questionnaire item already contains the annoyance judgment. The questions solicit respondents’ theories about why they are not annoyed, rather than directly measuring their coping strategies. If a series of questionnaire items were devoted to coping in a new US national survey, the survey might make a significant contribution to knowledge about personal factors that affect noise annoyance. However this could require a substantial commitment in terms of questionnaire space and substantial work on question wording and concept development to overcome the tautological question issue raised above. Personal coping strategies are not directly tied to either the unique aspects of the proposed study’s multi-airport design or to authorities’ actions. A related concept, perceived effect of resident’s opinions on public policy (Issue 35), is more relevant. Issue #30: Expectations for future exposure The expectation that noise exposure will increase in the future has been hypothesized as increasing annoyance. No systematic summaries have been compiled to determine whether studies consistently support this hypothesis. Two studies have supported the hypothesis (Öhrström, 1997; Schreckenberg, Schümer, and Möhler, 2001). With respect to public complaints and actions, it is often observed that complaints and other public actions around noise issues are strongest when proposals are introduce that would increase noise. Topic E: Effect of attitudes toward authorities Issues #31 to #35 Studies have not yet systematically explored the extent to which residents’ annoyance is affected by their attitudes toward authorities who are in some way related to the noise source. The preventability issue (Issue 28, above) provides a broad judgment of whether authorities are doing all they can to reduce noise but does not explore other attitudes toward authorities which might influence judgments of preventability or have an independent effect on annoyance. The current review of the literature is consistent with a 2007 review that concluded that there is no substantial social survey research about the relationship between authorities’ actions, residents’ attitudes toward those authorities, and residents’ annoyance (Vader, 2007). The literature reviewed in Section 4.0: (Frameworks for assessing authorities’ actions) identifies factors that are hypothesized to affect the relationships between airport authorities and communities’ public actions. In addition several laboratory experiments reviewed by Vos suggest that perceptions of authorities can affect ratings of noise (Vos, 2010). On the basis of that literature it is hypothesized that the following five perceptions reduce annoyance with aircraft noise: H-24

31 Transparency of process The perception that authorities develop policy transparently and provide relevant information decreases annoyance 32 Fairness of procedures The perception that authorities follow procedures in a fair manner decreases annoyance 33 Trust The perception that authorities can be trusted decreases annoyance 34 Understanding of residents' concerns The perception that authorities understand or are concerned about residents decreases annoyance 35 Residents can affect policy The perception that authorities' actions are influenced by residents' views decreases annoyance. (This is an extension of the coping and control attitudes from personal control within the home to control of public policy.) For the purposes of the proposed national study, these perceptions provide an important link for understanding the impact of local authorities’ actions and the neighborhood-to-neighborhood differences in annoyance reactions (Issue 19). Survey questions about the authorities’ actions and attitudes could be explicitly directed at the best practices for community relations programs that are described in the section 4 Frameworks for assessing authorities’ actions (Table 20) to determine whether particular aspects of community relations programs affect corresponding perceptions among residents and, in turn, affect noise annoyance. Such survey questions might also help to explain local neighborhood-to-neighborhood variations in annoyance by determining whether some of this variation is associated with neighborhood-to-neighborhood variations in perceptions of authorities or with acoustical or non-acoustical mitigation actions of authorities. Topic F: Effect of aircraft operations The acoustical dimensions of aircraft flights are captured in noise indices which, as was explained in the introduction, are not the subject of this literature review. Other aspects of aircraft operations are discussed here under Topic F: Effect of aircraft operations Issue #36: Distance to flight path It is often hypothesized that residents who live under a flight path will be more annoyed than those in a side line position, primarily because of a heightened fear of aircraft crashes. As the meta-analysis in Table 18 shows, only three studies were located in a 1993 summary report. A keyword search of reports for 628 surveys and a web-based search for additional references identified one additional report. The evidence is not consistent. An Oslo airport study found those under the flight path were more annoyed by the equivalent of about a 6-decibel difference in noise exposure (Gjestland, Liasjø, Granøien, and Fields, 1990). A survey around Toronto airport found that those under the flight path were more annoyed, but the authors were uncertain about the interpretation of the finding because those residents were not any more fearful than other residents (Hall, Taylor, and Birnie, 1980). A survey around Canadian general aviation airports did not find higher annoyance under the flight paths but the authors’ expressed uncertainty as to whether the training routes used for touch-and-go flights were accurately specified (Hall, Taylor, and Birnie, 1980). One of the first aircraft noise surveys, conducted in the US in the 1950’s, found higher annoyance under the flight paths in some areas but higher annoyance to the side of flight paths in other areas (Borsky, 1954). A new national US study could provide much stronger evidence than the previous studies. The new study will have tracks for individual flights and thus be able to much more accurately specify residents’ locations relative to the spread in actual flight tracks. The information will also make it possible to quantify the extent to which aircraft are directly above the respondents’ dwellings, unlike the previous surveys that simply classified dwellings as being below or to the side of a nominal flight path. Of course, H-25

a sample design with a large number of airports and flight paths will also reduce the possibility that correlated variables are confounding the interpretation of a relationship. Issue #37: Landing/take-off operations The sounds of take-off and landing aircraft are distinct. It is hypothesized that for the same noise exposure, sounds from take-offs will be more annoying. Although a thorough search of all the literature was not conducted, a keyword search of the content of reports for 628 surveys and a web-based search for additional references identified six studies that examined the correlation between annoyance and the ratio of take-off operations to landing operations. The findings are not consistent. Two surveys found that annoyance was higher for take-offs, but the relationship was not statistically significant for one, an Oslo survey (Gjestland, Liasjø, Granøien, and Fields, 1990), and the relationship for the other, a Heathrow survey, though statistically significant, was not strong enough to support a recommendation for a change in the noise metric (MIL Research, 1971). Two other surveys, one from Japan and one from the US, found the opposite to the hypothesized relationship: annoyance was greater for locations with relatively more landings (Nguyen, Yano, Nguyenhuy, Nishimura, Sato, Morihara, and Hashimoto, 2011; Patterson and Connor, 1973). However the US survey, as will be explained elsewhere, provided ambiguous findings because the take-off/landing ratios were strongly correlated with other airport characteristics (Patterson and Connor, 1973). In a Salt Lake City rating quasi-experiment, the groups of subjects who rated each flight passing over the test home were no more annoyed by take-off than landing noise (Dempsey, Stephens, Fields, and Shepherd, 1983). In a rating study around Frankfurt airport, the residents made hourly ratings of the aircraft noise environment over the course of a test day. It was found that the hours that contained both take-off and landing operations were more annoying than hours that contained only a single type of operation (Schreckenberg and Schuemer, 2010). The evidence on this, as on many issues, is not consistent. The proposed US national survey should be able to provide stronger evidence than previously with a new hypothesis and strengthened sample design. With improved measures of the flight path location it will be possible to determine whether the landing/take-off difference might be found under flight paths. Previous surveys have assumed that the landing/take-off effect would be found across the entire area even at extreme sideline positions where the direction of the flights may be less relevant. Another possible hypothesis is suggested by the Frankfurt study: consistency in types of operations or consistency in flight-to-flight noise signatures may reduce annoyance. The detailed flight operation data that will be available for the US survey should provide sufficiently accurate data to explore such a consistency hypothesis. The improved sample design should also support a better test for the landing/take-off hypotheses. A large number of airports should help to reduce the correlations with other airport characteristics. Airports with considerable variation in landing/take-off ratios can provide within-airport estimates that are less likely to be confounded by between-airport differences although there may be, possibly weaker, within-airport confounding factors. Issue #38: Ground operations Standard aircraft noise indicators include only the noise from flying aircraft, not the noise from ground operations such as taxing, engine testing or ground run up. These noises can be annoying, but results from such studies have not been systematically summarized in the literature. It is unlikely that the effect of ground operations can be accurately assessed within the planned new national survey due to small populations, correlated variables and requirements for expensive acoustical estimation issues. Only a small proportion of the aircraft noise population is exposed to ground operation noise. Those exposure situations occur very near the airport and are thus strongly correlated with noise from conventional operations as well as with many other impacts from the airport. The noise exposure H-26

from ground operation at residences is expensive to accurately predict due to uncertainties about the noise levels emitted by aircraft, the location and frequency of such emissions, and the attenuation of any such emissions by structures on the airport near the aircraft as well as structures in the neighborhoods near individual residences. Issue #39: Airport size A summary analysis has not been conducted of the differences between large and small airports or of the differences between air carrier and general aviation airports. Several published reports based on comparisons of small numbers of air carrier and general aviation airports come to different conclusions. A 1977 survey around a small airport in Hamble England with predominantly training operations for propeller aircraft (UKD-309) reported that reactions were about as expected at large, air carrier airports (Directorate of Operational Research and Analysis, 1982a). A 1981 survey around five general aviation airports in the United Kingdom (UKD-243) reported that the residents of four of the five study airports were less annoyed than would be expected at air carrier airports(Directorate of Operational Research and Analysis, 1982b). A later 1986 survey around five small airports in the United Kingdom (UKD-324) found some evidence that noise from general aviation was more annoying than that from air carriers (Diamond, Ollerhead, Bradshaw, Walker, and Critchley, 1988). A 1975 survey around four small airports in Germany (GER-114) concluded that annoyance was greater at these general aviation airports than would be expected at large air carrier airports (Rohrmann, 1976). The reasons for such inconsistent findings may be the problems that were identified for the1970’s comparison of small and large airports in the United States and mentioned above in the discussion of take-off/landing comparisons (USA-022, USA-032, USA-044) (Patterson and Connor, 1973): the airports of different sizes may differ in other ways and the survey and acoustical methodologies followed at the airports may have been sufficiently different to confound the comparison. The report of a Canadian survey around general aviation airports points out that small general aviation airports often pose an especially difficult, error-prone noise estimation problem because it is not feasible to accurately estimate the exposure from touch-and-go circuit flights (Hall, Taylor, and Birnie, 1980). Issue #40: Predictable noise profile No studies have been located that tested the hypothesis that noises are less annoying if the time-history course of the noise events are more predictable. This hypothesis has been offered as one possible explanation for the lower annoyance reaction to railway noise than to aircraft or road traffic noise (Miedema and Vos, 1998b). However, this is not consistent with the finding in Japan that railway noise is more annoying (Yano, Sato, and Morihara, 2007). This issue may well be becoming more relevant for aircraft noise as new advances in aircraft air traffic control make it possible to more precisely specify flight paths. In addition, many airport authorities can use aircraft noise monitoring stations as a basis for monitoring individual flights’ compliance with aircraft noise exposure norms. Whether this issue can be explored in a new national study depends upon the extent to which aircraft movement data and/or noise monitoring stations can accurately identify variations in aircraft noise events at individual respondents’ locations. Issue #41: Vibration The impact of vibration on annoyance has been studied primarily for railway and impulse noises. The effects of sonic booms and other large amplitude impulsive noises have also been studied (Fields, 1997; Schomer, 1981). While these studies examined annoyance with vibration, they have not studied how vibration levels would affect the more general annoyance with aircraft noise events. Aircraft noise can excite the structure with the possibility that residents may see, feel or hear some items move. It is not H-27

clear whether most previous research on the most studied noise sources is highly relevant for commercial aircraft noise. Unlike some railway vibration, aircraft-induced vibration is exclusively caused by acoustical emissions from aircraft. Unlike sonic booms and major impulse sounds, vibration is less likely to be associated with being startled or with perceiving danger. For commercial aircraft the most frequent effect of vibration may be windows rattling, a noise source which may or may not be included in respondents’ definitions of ‘aircraft noise’. A recent study of annoyance with vibration provides some evidence that acoustical window treatments may have the side effect of reducing annoyance with vibration, presumably by reducing the rattling of windows (Fidell, Pearsons, Silvati, and Sneddon, 2002). The study did not assess any resulting effect on noise annoyance. A major seven-year program to assess vibration and combined vibration and noise impacts is being conducted in the UK, but may be of limited relevance as it is not studying aircraft noise induced vibration (Perkins, Grimwood, Stanworth, and Thornely-Taylor, 2011). Issue #42: Non-noise other While noise and vibration are the most obvious impacts from transportation noise, neighborhoods can also experience other impacts such as air pollution, odors, dust, dirt, lights, or the visual impact from seeing the source of the noise. Though subjective perceptions of these non-noise impacts are occasionally asked about in questionnaires, no publications have been located about the relationship between annoyance and objective measures of these non-noise impacts for aircraft noise. This may be partly because such impacts are limited to only the highest noise levels for aircraft noise and because they are difficult to objectively measure. Accurate objective measurements would be expected to be too expensive to conduct for a new national survey. Issue #43, #44: Reactions to change in exposure Issues 43 and 44 concern the difference between reactions to a static, long-term noise exposure and reactions to a recently increased or decreased noise exposure. The recent meta-analysis in Table 18 concludes that most studies find that residents overreact to such changes as if the change in noise exposure had been even larger (Brown and van Kamp, 2009). The analysis also found evidence that such a reaction was persistent and continued for at least the few years after the change that had been investigated in the available studies. The authors noted however, that most evidence was from ground- based transportation noise studies where the changes were clearly visible. They stated that there were too few, inconsistent aircraft noise change studies to determine whether the hypotheses also held for aircraft noise (Brown and van Kamp, 2009). While the impacts of changes in noise exposure on private annoyance are often small, difficult to detect and not found in some surveys, the impact on complaints and public reactions are reported to be large, vigorous and almost universal. Proposals for increases in aircraft noise exposure are also reported to universally create vigorous public opposition. Even though the current evidence suggests that changes in noise levels may have only small persistent changes on annoyance with aircraft noise, a new national survey will need to obtain objective information about noise exposure trends and abrupt changes if for no other reason than to be able to test the wide-spread perception that changes in noise levels generate very large changes in annoyance. A questionnaire item about perceptions of changes in aircraft noise exposure should also be included to determine whether residents in some locations perceive changes even when airport operations are unchanged. H-28

Topic G: Authorities’ actions and activities Issue #45: Operator noise abatement actions When authorities take steps to reduce noise exposure it is hypothesized that residents may overreact with lower annoyance than would be expected from the noise level alone. This is similar to the earlier change hypothesis (Issue 43) but is based on the assumption that it is the knowledge of the authorities’ efforts that creates at least part of a reaction to the changed exposure. This might occur if residents knew about a specific change that was instituted by authorities or if residents attributed any improvements to authorities’ actions because the authorities were perceived as generally attempting to reduce noise. No summary analysis has been conducted of such changes. These actions are discussed further in the section Frameworks for assessing authorities’ actions and listed there in Table 19 Some doubt is thrown on this hypothesis by the earlier cited study in which it was found that most residents were not aware of night time trials even though there was a publicity campaign and media coverage (Flindell and Witter, 1999). Issue #46: Community relations programs Airports have a range of programs for communicating with their impacted communities. These programs are often directed at community members who are involved in public actions or complaints against airport noise. There is very little information about the extent to which these programs impact the privately felt annoyance that is measured in social surveys. One quasi-experimental field study was conducted at the airports Augsburg and Kassel-Calden in Germany (Maziul, 2005). At each of the airports a sample of residents were interviewed with standard noise survey questions and then told about a noise complaint line, “NoiseCall”. Residents who subsequently used the line were identified. In later follow-up interviews the original sample was followed up and the characteristics of those who used and those who did not use “NoiseCall” were compared. The perceived control and the coping strategies of residents were meant to be enhanced, and consequently, annoyance to be reduced and contentment with the management increased. When the data were analyzed the annoyance of Kassel residents who used the NoiseCall declined significantly and the contentment correspondingly increased – though the increase was not statistically significant. However at the other airport, Augsburg, no significant changes after the installation of the NoiseCall were detected. The evidence from this small experiment was thus mixed. One problem in assessing the impact of a community relations program is that the effects of an airport’s community relations program might be confounded with other airport characteristics. One way to at least partially address this problem is to measure the expected intermediate effects of a community relations programs on the various attitudes toward authorities that were outlined above under Topic E. Issue #47: Conflict – history A well-known study of an acoustics consulting firm’s files, found support for the hypothesis that a history of conflict between an airport and its community would create heightened reactions to aircraft noise at an airport (United States Environmental Protection Agency, 1974; Wyle Laboratories, 1971). The methods for objectively measuring previous conflict were not published for the study. The study did not attempt to measure resident’s private annoyance. No studies have been located that relate a history of conflict to residents’ annoyance with noise. Issue #48: Compensation to residents Authorities sometimes offer some type of compensation to residents related to their noise exposure. No studies have been located that relate receiving compensation to noise annoyance. H-29

Issue #49: Operator’s perceptions Airport authorities and other personnel who are involved in aircraft operations at an airport often believe that communities at similar noise levels react quit differently to equivalent noise exposures. This perception may be derived from objective data, such as counts of complaints and law suits, but may also be affected by less formal, difficult-to-quantify impressions formed from long-term contacts with some members of the communities. It is hypothesized that these operators’ perceptions of community response are related to residents’ annoyance. No studies have been located which tested this hypothesis. While it may be that such perceptions only reflect a few publically-expressed community leaders’ views, it is also possible that the operators’ experience does provide special insight into annoyance. Testing the hypothesis could determine whether noise impact on the population is best identified from statistical models of noise annoyance or whether the model’s predicted impact should be adjusted for authorities’ perceptions of local communities. Topic H: Correlated noise impacts Survey studies of environmental reactions to noise have always found that the general annoyance reaction is closely related to other attitudes which measure various specific perceived impacts of noise. Frequently studied perceived impacts include those listed in Issues #50 to #54: 50 Vibration The belief that the noise source also causes vibrations increases annoyance 51 Health The belief that the noise source affects health increases annoyance 52 Air pollution The belief that the noise source's fumes or dirt pollute the air increases annoyance 53 Activity disturbance The belief that the noise interferes with daily activities (speech, concentration, listening, etc.) increases annoyance 54 Sleep The belief that the noise interferes with sleep increases annoyance Early social surveys created noise annoyance indices by summing the answers to questions about levels of disturbance or concern about these types of perceived impacts. However, this practice is almost never followed now since a general noise annoyance question provides a clearer, more direct measure of respondents’ assessments. All of these topics have been reported upon in the noise literature, but no summary report has tabulated whether or not these perceived impacts are always related to annoyance. There does not appear to be a reason to believe that including these questions in a new national survey would strongly support the goals of the study. These are subjective, not objective, indicators of impact. In the absence of other evidence it is not even clear whether general noise annoyance influences assessments of these correlated impacts or whether, as is generally presumed, perceptions of these impacts cause general noise annoyance. Of course the sleep issue is being addressed in other ACRP studies with stronger methods. Perception of vibration could explain the linkage between the levels of vibration and noise annoyance, but in the absence of an accurate vibration estimation protocol, this national survey would not be expected to make an important contribution to understanding the effect of vibration. Topic I: Complaints and public actions The new national survey and previous noise surveys study the annoyance that people feel but may never share with others except in social surveys. This is accepted by the scientific community as the most accurate assessment of the feelings of residential populations toward noise. This private annoyance attitude measured in social surveys can be contrasted to actions that are expressed in public and directed at officials or other representatives of a noise source. The most frequent type of public action is individual complaints that are lodged with officials. Many airports have hot lines where complaints can be lodged. Stronger public actions include letters to newspapers, signing of petitions, attending meetings, H-30

meeting with officials, and lawsuits. Unlike expensive, infrequent social surveys, complaints and public actions are readily accessible to public officials and often demand their attention within the political system. A key question for public officials is whether the information that is readily available in public actions provides good information about the general impact of noise as measured by annoyance in social surveys. This question is addressed in three issues: the relationship between action (primarily complaints) and annoyance in communities, the variables which are correlated with complaint rates, and the demographic characteristics of complainants. Individual complaints rather than other types of public actions are most often studied because complaint actions are more frequent and more easily studied. Analyses have not been reported that used meta-analyses or other techniques to provide a quantitative summary of the evidence on these issues. In addition there is no evidence that extensive searches were conducted based on a results-neutral search strategy. Issue #55: Complaint rate: annoyance It is widely acknowledged that complaint rates substantially underestimate the amount of annoyance in a population. For example, a recent study around Schiphol airport found that only 19% of the highly annoyed respondents reported having complained to authorities (van Wiechen, Franssen, de Jong, and Lebret, 2002). Accepting the undercounting of annoyance on the basis of complaints, the question still remains as to whether the relative intensity of complaints in different communities is an adequate indicator of the relative levels of impact and annoyance in those communities. The same Schiphol study concluded that “Although complainants do not seem to be representative for the total population, and do not reflect the full extent of noise annoyance, their prevalence does reflect the regional distribution of aircraft noise annoyance in a noise polluted area” (van Wiechen, Franssen, de Jong, and Lebret, 2002). Given some of the correlates noted below, however, there are reasons for thinking that complaint rates may not reflect annoyance differences between areas. Issue #56: Complaint rates: other correlates The primary question is: What characteristics of local areas and events are correlated with complaint rates? The implicit related question is whether these characteristics also are correlated with annoyance. At least one publication about telephone complaints in Australia has shown that complaint rates can misrepresent the relative importance of different sources of noise pollution (Avery, 1982). The social survey found that road traffic and aircraft noise were the biggest problems and were more than twice as likely as construction or industry to be a problem. The telephone complaint rates provided a diametrically opposed indicator; construction and industry were the biggest problems and were mentioned four times more often than road traffic or aircraft noise. Complaint analyses now regularly consider details that have sometimes been neglected in the past such as the effects that a few serial complainers can have on complaint rates. None-the-less the underlying concern remains that complaint rates may not be a good indicator of annoyance because of other factors that turn annoyance into complaints in some circumstances but not in others. Examples of such factors are knowledge about how to make complaints, the belief that a complaint will be noticed by authorities, and confidence that a complaint can lead to change. Authorities often report that complaint rates increase when a change in noise exposure is proposed or occurs and when aircraft noise complaint hot lines are opened. A related question is whether the characteristics that lead to high levels of complaints are similarly correlated with private annoyance. It is widely acknowledged that plans for changes in airports create large numbers of complaints. For example, when complaint data from Manchester Airport since 1991 were examined, the annual number of complaints peaked in 1996 when the 'Manchester Airport Second Runway Public Inquiry' was a major local issue. But even though the number of flights increased since 1996 the number of complaints H-31

steadily fell from 50 to13 complaints per 1,000 movements from 1996 to 1999 (Hume, Terranova, and Thomas, 2002). An analysis of changes in complaint rates before and after a new runway opened in 2001 at Manchester Airport found that the complaint patterns followed the new traffic patterns (Hume, Morley, Sutcliffe, Smith, and Thomas, 2005. These striking patterns of reactions to changes and planned changes are in sharp contrast to the findings for the effects of changes in noise levels on annoyance where the evidence is sufficiently uncertain that a much weaker effect, if any, is likely (Brown and van Kamp, 2009). Issue #57: Characteristics of complainants Many discussions of community noise impact accept the hypothesis that complainants’ characteristics are representative of the annoyed residents generally. For example, if public officials hold meetings and find that most of those complaining about aircraft noise are well educated, articulate residents, it is easy to conclude that annoyance is also concentrated among highly educated residents. Though no summary analyses have been published on this issue, the issue has been addressed by analyzing social surveys by comparing the characteristics of respondents who do complain with respondents who do not report having taken part in complaints or other public actions. At least two studies augmented their standard population sample with a sample of complainants and found considerable differences in their demographic characteristics. A 1960’s survey in England found complainants are different than other residents at the same annoyance level in having more education, being of a higher occupation class, owning more expensive houses, participating at a higher rate in some type of organization that sought to make changes in the local area, having stronger preventability attitudes, but not being more sensitive to noise generally (UKD-008) (McKennell, 1963). A survey of aircraft noise in New York also found that complainants were more highly educated, had higher incomes, were more critical of airport operations, but were not more sensitive to noise (Tracor Inc., 1971). A more recent study around Manchester Airport in England is characteristic of the types of studies that are easily conducted with only complaint data (Morley and Hume, 2003). The postal codes for complaints were used to determine the average socio-economic status in the complainants’ postal codes. It was found that the complainants’ postal codes were inhabited by residents in the highest socio- economic category who had higher house prices, and were more likely to own more than one car. This is a classic case of the well-known ecological fallacy when a correlation observed at the population level is assumed to apply at the individual level (Robinson, 1950). More accurate information about the effect of demographic characteristics on complaints comes from the demographic characteristics of complainants and non-complainants than from comparisons of the average of the characteristics of all residents of geographical areas. Topic J: Other issues Issue #58: Self-selection through moving away It is often hypothesized that high noise areas contain a self-selected population, at least partly, because those who are most annoyed have moved out of the area. The tabulation of findings in Table 18 shows that most of the seven findings did not support this hypothesis. The available evidence does not support the selective-moving hypothesis. A secondary analysis of 28 studies (N=10,939 respondents) found that higher noise levels do not tend to be populated by more sensitive residents; if anything there is a weak relationship in the opposite direction (Miedema and Vos, 2003). There could, of course, be other mechanisms that would create an especially noise-resistant population at high noise levels, but the absence of strong relations between noise sensitivity and noise level suggests that any such mechanisms are weak. It seems likely that moving may be so difficult and expensive that moving is almost entirely determined by other factors. Noise sensitivity may have little effect on the choices of new residences if prospective tenants are unaware of noise levels, or are unable to predict their long-term reactions, have H-32

few realistic choices or are mainly affected by other factors such as price, unit size, distance to place-of- work, or other (non-sensitive) household members’ preferences. Issue #59: Long-term annoyance trends In the last ten years at least six publications have examined a hypothesized long-term increase in the dose- response relationship such that the average annoyance response at given noise levels is higher than it was several decades ago (Guski, 2004 Breugelmans, van Wiechen, van Kempen, Heisterkamp, and Houthuijs, 2004;Brooker, 2009;van Kempen and van Kamp, 2005;Wirth and Bröer, 2004). The results of the most recent and comprehensive review are summarized in the secondary analysis for Issue 59 in Table 18 (Janssen, Vos, van Kempen, Breugelmans, and Miedema, 2011). The summary analysis is based on 22 studies of 34 airports that were conducted over the 39 years from 1967 to 2005 and finds that recent social surveys provide higher annoyance responses than earlier surveys. However, it is not clear whether this is due to methodological factors, the locations studied, or to genuine changes in annoyance. Methodological factors could be especially important for this secondary analysis because the evidence comes from contrasts only between surveys (22) and cities (34) not from contrasts between respondents within surveys. For example, the secondary analysis of the effect of education contains both high and low education respondents from each survey. For the present change-over-time hypothesis, however the comparison is solely between surveys, many of which had different annoyance questions, different modes of survey administration, and were conducted in different countries at different locations. Though the recent review ruled out the number of annoyance scale points as an explanation for the increased annoyance, it was concluded that the trend might, at least partly, be explained by correlated trends in decreasing response rates and the use of self-administered surveys rather than interviewer- administered surveys. In addition, the surveys were drawn from different regions. All of the pre-1984 surveys (the first 18 years) were conducted in four English language countries (only one of which was European) and while all of the post 1984 surveys were conducted in northern European, non-English speaking countries. In the eleven years from 1985 through 1995, all three studies were conducted in Norway (from 1989 to 1992). Additional doubt about increased annoyance levels comes from the finding that there was not a corresponding trend in reported sensitivity to noise generally. Analyses that compare reactions at the same airports using similar survey methods but at different times would provide much stronger evidence. The national US study could do this by replicating a coordinated study at 9 airports that were studied around 1970(Connor and Patterson, 1972;Patterson and Connor, 1973;Tracor Inc., 1971) or by replicating some of the smaller studies that have been conducted at these and other airport since then (Fidell, Mestre, Schomer, Berry, Gjestland, Vallet, and Reid, 2011). Issue #60, #61: Survey Administration Telephone vs. face-to-face administered questionnaires: and (#61) Self-administered vs. interviewer- administered questionnaires. These two issues must be evaluated if the results of the proposed national survey are to be compared to previous surveys that used different modes of administration. Three modes of survey administration are most likely to be applicable for the national study: 1. Self-administered mail: A questionnaire that is mailed to a household which requests that an unnamed but uniquely-identified household member complete the survey. 2. Interviewer-administered – Telephone: An interviewer telephones a household, selects the respondent following standard procedures and reads the questionnaire. 3. Interviewer-administered- Face-to-face: An interviewer visits a household, selects the respondent following standard procedures and reads the questionnaire. H-33

A variety of other steps can be used to increase response rates and identify respondents, but these three examples provide a framework for considering the implications and limitations of previous noise annoyance survey research. Noise annoyance surveys have been regularly conducted since the 1950’s. Most studies have been conducted as face-to-face interview surveys. Telephone surveys have been infrequently used, perhaps because telephone surveys would not realize sufficient costs savings for these surveys, almost all of which were conducted in a single city and, within that city, within a limited number of compact neighborhoods or blocks where residents had almost identical noise environments from a transportation noise source. More recently, however, self-completion surveys have come to be more widely used. Most are mail surveys, but in some countries, most notably Japan, many are drop-off/pick-up surveys where the blank questionnaires are left at a home (with or without personal contact) and the completed questionnaires are then picked up later. Face-to-face interviews are exceedingly expensive and thus there are strong economic reasons for considering alternatives to an interview for the national study. Interviewer-administered telephone questionnaires have been compared to face-to-face questionnaires in at least the four studies that are summarized in Table 18. In these four US studies conducted before 1980 both modes were used in each study, the same annoyance question was asked in each mode, and noise levels were controlled in the analysis. None of the four studies found important differences in the answers to annoyance questions between the telephone and face-to-face questionnaires. This evidence thus suggests that the two modes will yield results that are not strikingly different. None of these trials were reported upon in sufficient detail to provide firm estimates of the precision of their results. In addition, most were conducted as part of the main study at a point when a finding of survey mode differences would have cast doubts upon the value of the study. The first report on the comparison of a self-administered mail questionnaire to a telephone questionnaire found a very large difference in annoyance responses. For this 1996 Netherlands study (NET-371) the original mail survey estimate of 18% highly annoyed residents around Schiphol was revised to an estimate of 31% highly annoyed to make the results comparable to previous interviewer-administered studies. At the highest noise levels the highly annoyed estimate was increased from 48% to 65% (Franssen, Lebret, and Staatsen, 1999; Page 18). For this study it was not clear what characteristics of the two survey administration modes created the difference. As for many self-administered surveys this mail survey differed from an interviewer-administered survey in that the mail survey less tightly controlled the within-household respondent selection, informed respondents about the survey purpose before they decided to participate, obtained a low response rate, and informed respondents about the survey purpose before they answered the aircraft noise annoyance question. In addition the mail questionnaire was used in the main study whereas the interviewer-administered telephone survey was directed at only the nonresponse households. Though the mail and telephone survey contained the same noise annoyance question, that annoyance question was preceded by different questions in the two surveys. A Dutch researcher familiar with the survey could not offer a firm judgment on whether or not mail respondents were likely to have followed the preferred within household selection instructions (personal communication with Sabine Janssen). The questionnaire itself did not specify the respondent while the cover letter contained only the statement in the middle of one paragraph that “We would greatly appreciate it if the person from your household…” with the nearest birthday “… would fill in the questionnaire"” (English translation by Sabine Janssen). Some other studies and reports point to possible differences between survey modes but find small or no mode effects. The statistical analyses from the most comprehensive secondary analysis of annoyance trends over decades of noise/annoyance surveys found that the shift from interviewer-administered to self-administered surveys was a possible explanation for an increase in noise/annoyance ratings (Janssen, Vos, van Kempen, Breugelmans, and Miedema, 2011). This was not, however, a tightly-controlled H-34

survey-mode comparison since the surveys were conducted in different countries with different noise/annoyance questions and other aspects of survey administration, including respondent selection procedures, were not considered in the analysis. A Japanese study (JPN-616) that compared web, mail and interviewer-administered surveys found a weak trend toward higher annoyance responses for the mail survey but not for the web survey (Yamada, Kaku, Yokota, Namba, and Ogata, 2008; Figure 4 ). The respondent selection methods for the mail and interview survey were not reported. Three studies that sampled named individuals did not find a differences between the mail-administered and interviewer-administered questionnaires used in their studies. Their study designs or analyses, however, may limit their relevance for a national USA study. A 1977 UK morning-after, sleep disturbance study (UKD-147) asked for ratings of the prior night, but did not have detailed noise data available at the time of the report (Directorate of Operational Research and Analysis, 1978). A 1999 UK study around 5 airports (UKD-482, UKD-489) found no differences in annoyance responses between the postal and interviewer-administered results even though the response rate from the mail survey was so low that the resulting small number of mail surveys were not used in the analysis. However, the report did not directly control for noise level in the analysis (Diamond et al., 2000). A strongly designed test in a 2003 Swiss survey (SWI-534) of named individuals which used identical questionnaires for two modes found “…no evidence for an effect of survey method (questionnaire versus telephone interviews) on annoyance…” (Brink, Wirth, Schierz, Thomann, and Bauer, 2008; Pages 2932-2933). However the report did not indicate that the dose-response relationship was directly evaluated. The applicability of the results is limited because the sample was predominantly drawn from low or moderate noise exposure areas - only 18% of the sample was above 57 Lden. One of the most often cited concerns about noise/annoyance surveys is that respondents will give biased, overly annoyed answers if they know that the primary purpose and goal of the questionnaire is to provide authorities with residents’ evaluations of a noise source in their neighborhood. As a result, most questionnaires are presented with a vague purpose such as “a study of living conditions” and the primary noise-specific annoyance question is included early in the questionnaire within matrices of questions about other environmental and noise conditions in the local area. With a mail survey it is not possible to conceal the purpose because a brief perusal of the questionnaire before it is answered reveals that most questions are about local aircraft noise issues. The possibility of such biases has been studied in at least three carefully designed interviewer-administered surveys, all of which were conducted in Great Britain in the 1970’s (UKD-071, UKD-116, UKD—157). All three studies used two questionnaire forms that varied in when (early or late in the questionnaire) the interviewer read the noise annoyance question. All three studies found that asking the question after the purpose had been revealed did not result in higher annoyance ratings (Garnsworthy, 1977; Langdon, 1976b; Griffiths, Langdon, and Swan, 1980). Less tightly designed comparisons in the pilot stages for a UK aircraft survey in 2005 (UKD-605) found that respondents whose interviews were preceded by ratings of aircraft noise acoustical recordings expressed no greater annoyance than did respondents who did not know that the survey was about aircraft noise when the survey began. This review of survey-mode effects has not explored all the relevant differences between the studies that could be hypothesized to affect the studies’ results. The evidence seems to suggest that if biases occur they are most likely to derive from some aspect of the respondent self-selection process (including nonresponse) rather than from the specified respondents’ responses being distorted by their knowledge of the survey’s purpose. With respect to within-household respondent selection, it appears to be possible that knowledge of the survey subject will result in either less annoyed households not returning a questionnaire or the most annoyed member of a household answering a questionnaire if the questionnaire does not name the sample member. One goal of pilot studies for a new national US survey could be to determine whether there are survey mode effects under the tightest feasible respondent selection procedures. The results from such pilot H-35

studies should aid in the selection of the survey mode for this study and provide a basis for comparing this survey’s results with the results of prior US and European surveys that have been conducted using different modes. Although the evidence reviewed above from previous, primarily non-US, surveys is helpful; information from new US pilot surveys is needed to draw firm conclusions. Populations in different countries could react differently to survey instructions and procedures. Experience from previous surveys may no longer be relevant if recent changes in the populations’ behavior, such as falling response rates, have modified the effect of survey mode on noise annoyance survey answers. Noise annoyance may also be a survey topic for which general professional experience with survey mode differences is not decisive. Strong within-household differences in annoyance reactions and interest in the local noise issues could bias within household selection more than for other survey topics. Issue #62: Relevant exposure period Noise annoyance survey questions generally ask for long-term annoyance either by vaguely asking about “annoyance around here” or by specifying a time period such as “this last year”. The hypothesis is then that respondents provide an assessment of their annoyance with the aircraft heard over the entire period. Brooker has argued that some data indicate that respondents actually consider only a shorter period of roughly three months (Brooker, 2008). The earlier evidence for a seasonal effect (Issue 24) also suggests that annoyance may be affected more by recent experiences than by long-term experiences. The exposure period is likely to not be relevant for the new national study because aircraft movements do not often change over a year. None-the-less the detailed data about aircraft movements available for a new national survey will make it possible to determine whether the study results could be sensitive to different assumptions about the relevant exposure period. Frameworks for assessing authorities’ actions As was noted earlier in this review, the proposed national survey with a large number of airports will provide a unique opportunity to explore the ways in which annoyance reactions are affected by airport characteristics. One of the most potentially important, but most often ignored, characteristics is the official policy environment set by airport authorities and other public institutions. For airport operators a major issue is whether their actions can reduce public complaints and residents’ noise annoyance. Most airports are concerned about how their actions might affect the public’s actions against an airport. Though airport authorities often have community relations programs and monitor the complaints and other public responses to the airport, they are not able to measure the impact of their particular programs on the residents’ annoyance. As a result one of the goals of this literature review has been to identify research on the relationship between authorities’ activities and residents’ annoyance. A major source of information on authorities’ actions is a 2006 report by Ruud Vader (Vader, 2007). After searching the available literature and after listing 50 non-acoustic mitigation measures, Vader concluded that there was: 1) not a satisfactory theoretical scheme for classifying mitigation measures and 2) virtually no empirical research on the relationship between authorities’ activities and residents’ annoyance. Another major source of information is an ACRP publication; an airport operator’s handbook for creating good community relations which is based on community relations research (Woodward, Friscoe, and Dunholter, 2009). Some recent publications by Kroesen, Bröer and their colleagues have provided a theoretical perspective for addressing one aspect of the public discourse/ private annoyance relationship (Kroesen and Bröer, 2009). They propose that dominant public discourse frameworks and individual community members’ frameworks influence each other and affect the amount of conflict. However, this does not provide a comprehensive perspective for clearly integrating noise-mitigation activities. Their approach does not appear to offer a useful hypothesis for a new national survey. Case studies have examined the acoustical and non-acoustical noise impact mitigation measures and the public’s actions. There are not, however, studies that correlate these aircraft noise mitigation activities H-36

with the noise annoyance expressed by representative samples of airport community residents. While there is an extensive community relations literature, no research protocols have been located that the planned US survey could use to systematically and objectively rate the quality of authorities’ noise- mitigation programs. The descriptions and inventories of existing programs and the guidance for designing noise impact mitigation programs do, however, provide the background material that could be used to begin to develop such a protocol Various typologies of noise-mitigation strategies have been developed. Stallen has classified the strategies by whether they require accommodations from the aviation sector or from the social environment (Vader, 2007). After examining the components of 50 noise mitigation programs Vader classified the components into eight categories: Community Programs, Compensation, Consultation, Financial, Information, Insulation, Land use and Property (Vader, 2007). These classification schemes do not provide a satisfactory theoretical scheme for predicting the impact of noise-mitigation schemes on residents’ annoyance. The lack of such a theoretical perspective for organizing the large number of disparate types of authorities’ activities is a major challenge for designing hypotheses that can be tested in the new US survey. This section attempts to provide some structure for this area by: 1. Listing types of aircraft noise mitigation activities that authorities can institute. 2. Listing some characteristics of community relations programs that have been hypothesized to promote good airport/community relations and thus to reduce annoyance. Authorities can attempt to mitigate the impact of aircraft operations through a wide range of acoustical and non-acoustical activities. Many publications have described particular airports’ activities. An especially complete description of acoustic and non-acoustic mitigation activities is available for Heathrow (Flindell and Witter, 1999). The search through the literature has identified many different mitigation activities. The FAA has provided guidance on noise mitigation through the Community Involvement Manual (Willkie, Madgwick, Sweatt, Frievson, and Carlton, 1990) since 1990 and an updated publication (Woodward, Friscoe, and Dunholter, 2009) both of which describe mitigation strategies. No publication has been located with a clear theoretical framework for rating activities by their presumed effectiveness in reducing personal annoyance with aircraft noise. For acoustical mitigation procedures, the mitigation might be measured in decibels of exposure reduced by the program for specific neighborhoods or airports. Such a purely acoustical measure raises the issue discussed under Issue 22 about publicity and residents’ knowledge and perception of the acoustical mitigation. For non-acoustical mitigation procedures it would be possible to rate community relations programs by features such as the resources expended, numbers of residents contacted, duration of the community relations efforts, or the number of different types of strategies pursued to communicate with residents. However, none of these are closely tied to a perspective that would predict the impact on annoyance. An alternative approach, to be discussed later in this section, is to rate programs by the extent to which they follow particular community relations principles. A necessary part of a framework for predicting impact is a list of acoustical and non-acoustical mitigation activities. Table 19 provides an outline of selected mitigation activities which, though not complete, indicates the scope of such activities. H-37

Table 19 Outline of Selected Aircraft Noise Mitigation Activities A. Acoustical mitigation (Controlling the aircraft noise) 1 Emission mitigation (Controlling the total amount of aircraft noise) a. Regulating the type of aircraft (Moving airlines toward the use of relatively quiet aircraft through fee structures, fines, or outright prohibitions) b. Controlling the number of flights (Limiting the total number of flights for the airport as a whole or at different times of day often through allocations of movement slots, impositions of fines or outright curfews) c. Regulating aircraft operations to reduce noise emissions (Encouraging operators to fly aircraft in a manner that reduces their noise emissions through setting goals, monitoring, delivering warnings, assessing fines, etc.) 2 Immission mitigation (Controlling the noise for individuals on the ground) a. Acoustical insulation of homes b. Purchasing property (Removing residents from impacted areas) c. Land use planning (Restricting the location of residences and other types of noise-sensitive land uses.) 3 Operational modification (Controlling aircraft flights while not necessarily affecting the types or total number of aircraft movements.) a. Locations of flight paths i. Routing flights away from densely populated or other sensitive areas ii. Concentrating flights over a relatively small number of areas to create areas that are not impacted iii. Dispersing flights over many areas to equitably distribute the impact over the affected population b. Modifying flight profiles (Changing ascent or descent profiles to reduce exposure in some areas) i. Creating steeper profiles that concentrate noise near the airport while reducing it at more distant locations) ii. Creating shallower profiles that reduce noise near the airport while exposing a larger area at more distant locations iii. Other standardizations of profiles ( e.g., creating regular, predictable aircraft noise emissions at particular locations) c. Varying flight paths over time (Creating breaks or quiet periods in areas that are otherwise exposed to aircraft noise.) d. Modifying ground-based aircraft operations B. Non-acoustic noise annoyance mitigation activities 1. Having a specially qualified staff member who manages community relations programs 2. Supporting an ongoing, routine community advisory committee or other liaison group 3. Regularly communicating with the community about airport developments through issuing press releases, holding public meetings, or attending other community meetings where airport noise might be discussed, , 4. Complaint monitoring (Processing information about individuals’ complaints about aircraft noise) 5. Compensating residents for exposure to aircraft noises For the non-acoustic noise annoyance mitigation activities and to a lesser extent for the acoustic mitigations, the important community-relations aspects of the programs may be less the types of activities than the characteristics of the relationships that are developed between the airport authorities and the airport communities. Several publications list characteristics which are presumed to contribute to reductions in aircraft noise impact (Vader, 2007; Vogt and Kastner, 2000; Woodward, Friscoe, and Dunholter, 2009). Some previous research has identified specific characteristics that residents ‘desire. For example, a survey around Düsseldorf International Airport and Dortmund Regional Airport, found H-38

that an open, personal and honest information exchange was the most mentioned desire of the residents (Vogt and Kastner, 2000). Such an exchange was ranked even more highly than acoustical noise abatement procedures. The most comprehensive and relevant framework for the proposed US national study is provided by the previously-mentioned ACRP supported community relations guidebook (Woodward, Friscoe, and Dunholter, 2009). The handbook identifies both the practices that authorities should follow and the outcomes that are desired and thus provides the basis for developing a protocol to rate authorities on the quality of their community relations’ programs. The handbook presents six “best practices”. In the first two columns of Table 20 these practices are divided into component parts and augmented by practices identified elsewhere to suggest nine distinct practices. Table 20 Best practices for community relations programs # Label Practice Indicators of success: Authorities Indicators of success: Community Members 1 Establish two-way communication Engage in two-way communication with the public (not just education, but listening) Airport officials believe that the public offers ideas that should be considered Community members believe that airport officials listen to and understand their views 2 Build trust Build trust through two- way interactions with the public Airport authorities trust representatives of the public with whom they regularly interact Community members trust airport authorities 3 Build respect Build respect through two- way interactions with the public Airport authorities respect the views of representatives of the public with whom they regularly interact Community members respect the views of airport authorities 4 Senior airport leadership out front Senior airport leadership are present and involved in interactions with the public Senior airport leaders attend airport/community interactions Community members believe that senior airport leaders are attentive to community issues 5 Use of graphics Graphics are used to illustrate and explain complex information Graphics are included in airport presentations to improve communication Community members feel they understand airport communications 6 Transparent planning process Members of the public are kept informed about the planning process and issues Airport officials inform the public as soon as possible about developments in a planning process and share information about the costs and benefits of alternative plans Community members believe that they are kept informed about the planning process 7 People skills among staff Noise staff members have a public service attitude and people skills Staff selection criterion include public relations skills Community members feel that airport staff are easy to communicate with 8 Communicate about long-term, future issues The public is engaged in considering issues that may develop far in the future Airport communications alert the public to possible changes in the far future Community members feel that they are aware of future issues 9 Include the public in decisions Include the public when making decisions Airport officials include members of the public in discussions before decisions are made Community members feel officials understood the issues and consider the community’s views before airport decisions were made At least two approaches could be taken to rate the quality of community relations programs: H-39

1) The airports’ documents and practices could be examined to determine the extent to which the best practices are being followed (indicators are listed in the third column of Table 20), or 2) The airport authorities and relevant involved community activists could be interviewed to determine whether the desired types of airport/community relationships had been established. The last column of Table 20 lists the types of community members’ perceptions that would show that the community relations program was successful. A 55-question questionnaire is included in the community relations handbook (Woodward, Friscoe, and Dunholter, 2009) that provides more detailed measures of the implementation of best practices than those proposed in the third column of Table 20. If rating of the community relations’ programs were obtained from either program examinations or citizen/activist surveys, the next steps would be to determine whether the rating of the quality of the community relations program is related to the two goals of: 1) operating the airport and planning for future airport changes with a minimum of adversarial public resistance and 2) reducing noise annoyance levels among community residents. Although a number of publications have mentioned the possibility of assessing annoyance (Flindell and Witter, 1999; Woodward, Friscoe, and Dunholter, 2009), no studies have been located that correlated the quality of a community relation program with annoyance responses of the residents. Though various case studies have analyzed the place of an airport’s community relations programs within-airport/community conflict situations, no studies have been located which attempt to trace the long-term effects of community relations programs on airport plans for expansion or development. The planned US survey could provide insight into the effects of aircraft noise mitigation programs by tracing the chain of causation from: Step #1) characteristics of the airport authorities’ community relations program [Table 20: Practices] TO Step #2) measured characteristics of the relations between airport officials and involved community activists [Table 20: Indicators of success] TO Step #3) less involved community residents’ perceptions of airport authorities actions and policies (Table 17, Issue 28, Preventability and Topic E: Attitudes toward authorities), TO Step #4) those community residents’ annoyance with aircraft noise. Summary: Considerations for Studying Each Issue This section summarizes the conclusions from the literature review with respect to the data collection for the annoyance survey. Related suggestions are based on the assumption that the primary goal is to provide a best estimate of the bivariate relationship between aircraft noise exposure and the annoyance of residents for the nation as a whole. A major secondary goal is to determine the extent to which dose- response relationships are uniform or different in different geographically-defined areas and, to the extent that areas differ, to identify factors that explain these differences. These goals guided the discussion of the questionnaire design and other data gathering activities in the text above. Table 21, below, summarizes this discussion by presenting suggestions for the 62 issues from Table 17. Some of these suggestions require data that are not available from the respondent questionnaire or aircraft noise models. Table 22 further summarizes the ancillary data collections that would be needed to address the issues. The first columns of Table 22 repeat the issue number, issue, and hypothesis from Table 17. The following column, “Include to: Contribute? Control?”, contains a “NO” if the subject is not suggested for the new study and either “Contribute” or “Control” if the topic is considered. “Contribute” indicates that the reason for studying the issue is to obtain important new findings about noise annoyance. “Control” H-40

indicates that the issue must be studied (sometimes with a single question in the questionnaire) in order to control for the effect of the variable in the analysis. For “Control” issues it is expected that no new findings will emerge, but that the study results might not be accepted unless there was evidence that the results were unbiased by the “control” variables. The following column, “Questionnaire Items” provides a broad indication of how many questionnaire items, if any, might need to be added for the issue. The next column, “Require ancillary data” indicates whether data are needed from a source other than the resident questionnaire or the aircraft noise model data. The “Cost of Studying” column gives a very general idea of whether the cost for including each issue will be “High”, Moderate” or “Low”. Finally, the “Comments” column mentions a few factors to consider in a decision about including an issue in a new study. Of course, a more complete discussion of these and other factors has been given above in the body of the report. H-41

ACRP 02-35 FINAL REPORT Table 21 Summary of Considerations for Studying Each Issue Issue # Issue Hypothesisa Include to: Contribute?, Control? Questionnaire item(s)? Require ancillary data? Cost of studying Comments A. Effects of demographic characteristics 1 Gender Women are more annoyed Control 1 No Low 2 Age - (Older age) Older age increases annoyance - See Issue 3 - - - - - - - 3 Age - (Middle age) Middle age increases annoyance Control 1 No Low Ask for date of birth. 4 Education High education increases annoyance Control 1 No Low 5 Status-occupation High status increases annoyance NO - - - - - - Occupation is too expensive to obtain and code. 6 Income High income increases annoyance Control ? ? ? Either include an income question or, if outside records are available, the value of the property. 7 Household size Mid-sized households (2 or 3 members) increase annoyance Control 1 No Low 8 Length of residence Long-term residence decreases annoyance Control 1 No Low 9 Home ownership Home ownership decreases annoyance Control 1 No Low 10 Dwelling unit type Residents of single unit dwellings are more annoyed Control 1 No Low 11 Usage of noise source Use of the transportation noise source decreases annoyance Control 1 No Low 12 Economic connection Economic dependence on the noise source decreases annoyance Control 2+ No Low May want more than one question. Must consider other members of the household. 13 Country/culture People from different cultures and countries differ in noise/annoyance reactions NO - - - - - - Do not expect enough minorities; Is not in literature; Has little value for policy. B. Effects of home conditions that modify exposure 14 Time at home Residents spending more time at home are more annoyed Control 1 No Low 15 Exposure individualized - (sum below: out-of-doors , attenuation, orientation, season) Individuals with relatively less exposure are less annoyed See Issues 16 - 18 - - - - - - 16 Out-of-doors usage Residents who spend more time out-of-doors are more annoyed Control 1 No Low 17 House attenuation Greater outside-to-inside transmission loss decreases annoyance Control or contribute ? ? ? If transmission loss were accurately predicted the study would contribute to knowledge. 18 Room orientation Quiet 'escape' rooms decrease annoyance NO - - - - - - Not relevant for aircraft. C: Effects of local community conditions H-42

ACRP 02-35 FINAL REPORT Issue # Issue Hypothesisa Include to: Contribute?, Control? Questionnaire item(s)? Require ancillary data? Cost of studying Comments 19 Community differences Airports and communities differ in annoyance responses Contribute ? Yes High Understanding community differences requires (1) special sample design (2) data about community. 20 Ambient noise levels Low-ambient noise levels increase annoyance Control 1 Yes Low/Moderate Might predict traffic noise from maps or other source. Probably need questions on other traffic and other noise to corroborate. 21 Sparsely settled areas Rural or 'peaceful' suburban environments increase annoyance Control 1 ? Moderate Could ask respondent about distance to other houses or about whether rural or farm area. 22 Media coverage Positive or negative media coverage of the noise source creates corresponding annoyance reactions Contribute 2+ Yes High Need content analysis of local media records to identify publicity. 23 Activist community Politically active communities increase annoyed Contribute 1 Yes High Need a source (experts or media) to rate the activism of the communities. 24 Meteorological conditions Comfortable weather (average climate or date-of-interview weather) increases annoyance Control 0 Yes Moderate Need temperature and other meteorological data for recent days, months and long-term. D. Effects of attitudes - general 25 Fear/Danger Fear of danger from a noise source increases annoyance Contribute 2+ No (Only flight tracks) Moderate Additional questionnaire items could clarify the relevance of the flight path position. 26 Sensitivity General sensitivity with noise increases annoyance Control 1 No Low 27 Importance of source Belief in the importance of the noise source decreases annoyance Control 1 No Low The wording of the question needs to consider attitudes toward authorities. 28 Preventability by authorities Belief that authorities could reduce noise increases annoyance Contribute (Issues 31-35) 2+ No Low Preventability questions should be coordinated with other attitudes toward authorities. 29 Exposure control - individual Belief that the resident can control or avoid noise exposure decreases annoyance NO 2+ No High Time and pretesting is required to develop items. These are not related to the study goal. 30 Expectations for future exposure Expecting an increase in noise exposure increases annoyance Control 1 No Low A questionnaire item is available (Fields, Ehrlich, and Zador 2000). E. Effects of attitudes toward authorities 31 Transparency of process The perception that authorities develop policy transparently and provide relevant information decreases annoyance Contribute 1 Yes High Linking Issues 31-35 with community relations program characteristics requires a survey of authorities. H-43

ACRP 02-35 FINAL REPORT Issue # Issue Hypothesisa Include to: Contribute?, Control? Questionnaire item(s)? Require ancillary data? Cost of studying Comments 32 Fairness of procedures The perception that authorities follow procedures in a fair manner decreases annoyance Contribute 1 Yes High (See Issue 31) 33 Trust The perception that authorities can be trusted decreases annoyance Contribute 1 Yes High (See Issue 31) 34 Understanding of residents' concerns The perception that authorities understand or are concerned about residents decreases annoyance Contribute 1 Yes High (See Issue 31) 35 Residents can affect policy The perception that authorities' actions are influenced by residents' views decreases annoyance Contribute 1 Yes High (See Issue 31) F. Effect of aircraft operations 36 Distance to flight path Being under a flight path increases annoyance Contribute 0 No (Only flight tracks) Low See discussion for Issue 25: Fear. 37 Landing operations Exposure to landing operations increases annoyance Contribute 1 No (Only flight tracks) Low A question might be asked about awareness of types of actions. 38 Ground operations Noise from ground operations (including start of take-off noise) increases annoyance beyond the levels expected from airborne operations Control 1 No (only if try to contribute) Low (High if try to contribute) Could ask a single question about awareness, but is probably too expensive to satisfactorily estimate noise exposure. 39 Airport size Small airports create greater annoyance (adjusted for noise exposure) Control 0 No Low The numbers of flights are known and would be expected to control for airport size effects. Airport annual movements are available. No general aviation airports will be studied. 40 Predictable noise profile A regular, predictable noise event profile decreases annoyance Contribute 1? Yes Uncertain A question will ask about predictability. It is not yet certain if analyses of flight data would measure the predictability of time- histories. 41 Vibration Vibration of structures or rattles increase annoyance Control 2 - - - - Questions can document vibration and rattle annoyance but with contributing to knowledge or controlling for vibration effects because it is not possible to predict exposure. 42 Non-noise (other) Lights, odors, or other non-noise impacts increase noise annoyance NO - - - - - - Exclude because this seems to be rare for aircraft noise, not an issue, and no observation data would be available. H-44

ACRP 02-35 FINAL REPORT Issue # Issue Hypothesisa Include to: Contribute?, Control? Questionnaire item(s)? Require ancillary data? Cost of studying Comments 43 Change-Immediate impact Residents overreact to changes in noise exposure (either increase or decrease) Control 1 Yes Moderate/High Aircraft noise exposure is needed for each of 5-10 previous years for each residence. A survey question would demonstrate that many people believe noise is increasing regardless of exposure changes. Probably could not contribute because not enough changed environments. 44 Change-Long-term adaptation With time, annoyance with a changed noise exposure decreases Control See Issue 43 See Issue 43 See Issue 43 (See previous.) G. Authorities= actions and activities 45 Operator noise abatement actions Officials' programs to control noise decrease annoyance beyond levels expected from exposure Contribute 2+ Yes High Official records may have some information. Otherwise interviews with officials are required as for Issue 31. Also need resident questions about awareness of actions or programs. 46 Community relations programs Strong community relations programs decrease annoyance Contribute See issues 31-35 Yes High (See Issue 45) 47 Conflict - history A history of noise operator/community conflict increases annoyance Contribute Yes High Media search and interviews with authorities as needed as for Issue 31. 48 Compensation to residents Receiving compensation from authorities decreases annoyance Contribute 2+ Yes High Need to develop questionnaire item and use authorities’ records. 49 Operators' perceptions Operators can predict residents' annoyance levels Contribute 0 Yes High (See issue 31) It may be difficult to establish community boundaries for judgments. H. Correlated noise impacts 50 Vibration The belief that the noise source also causes vibrations increases annoyance Demonstrate confounding 1 No Low (See Issue 41) 51 Health The belief that the noise source affects health increases annoyance NO - - - - - - The remaining impact issues 50 to 54) provide little or no objective information about impact and require questionnaire time. 52 Air pollution The belief that the noise source's fumes or dirt pollute the air increases annoyance NO - - - - - - (See Issue 51 and Issue 42.) 53 Activity disturbance The belief that the noise interferes with daily activities (speech, concentration, listening, etc.) increases annoyance NO - - - - - - (See Issue 51.) Questions would not contribute to study goals, but can illustrate types of impact. H-45

ACRP 02-35 FINAL REPORT Issue # Issue Hypothesisa Include to: Contribute?, Control? Questionnaire item(s)? Require ancillary data? Cost of studying Comments 54 Sleep The belief that the noise interferes with sleep increases annoyance Control 1 No - - (See Issue 51). A question would provide a weak indicator of between- site differences in nighttime noise impact. I. Complaints and public actions 55 Complaint rate: annoyance High community complaint rates indicate relatively high annoyance Contribute 0 Yes High Complaint data are needed for areas from airport records for collation with annoyance. 56 Complaint rate: other correlates What community or event characteristics are correlated with complaint rates? Contribute 0 Yes High Need information about communities 57 Complainant characteristics Complainants' characteristics are representative of annoyed residents Contribute 2+ No Low This is low cost if it only asks the random sample. It is high cost if there is a special sample of complainants. J. Other issues 58 Self-selection by moving away High noise levels cause annoyed residents to move out of the area NO - - - - - - Information about actually moving will not be available. Questions about intent could be asked but probably have low validity. 59 Long-term annoyance trends Over the years noise annoyance has increased for the same noise exposure Contribute 1-2 Yes High Need data from previous surveys and the ability to reproduce acoustical estimates. Also include previous survey annoyance questions. 60 Survey administration : Telephone vs. Face-to- face A telephone interview yields higher annoyance ratings (than face- to-face) Contribute 0 Yes High These data provide a basis to make comparisons to previous surveys. 61 Survey administration : Self- vs. interviewer- administered Self-administered questionnaires (mail) yield higher annoyance ratings Contribute 0 Yes High These data provide a basis to make comparisons to other surveys. 62 Relevant exposure period Respondents' annoyance is determined by recent experiences not by the entire previous year Contribute 0 No Moderate This requires calculations of exposure for additional time periods (e.g. previous month, 3-months). H-46

Table 22 identifies 16 data collection elements of a complex, multi-objective study design, including the 14 that are not required by the basic dose-response goal. Although these 16 elements are implicit in Table 21, they are easier to isolate in Table 22. The elements are listed under four headings according to the function they would perform in the study. Table 22 Possible Elements of a Complete, Complex Annoyance Survey A Primary Study Goal: Items that are required for the primary study goal 1. Resident questionnaire (Required to measure annoyance and the various demographic and attitudinal issues.) [Issues 1 to 13, 25-35, 57] 2. Flight operation data and analysis (Required to estimate noise exposure and identify the characteristics of the operations near a respondent – relation to flight path, landings, etc.) [Issues 36, 37, 39, 40] B. New Contributions to Understanding Community annoyance: Elements that contribute to understanding community differences 3. Clustering of respondents into blocks in the sample design. (Needed to identify the size of geographical areas associated with different reactions.) [Issue 19] 4. Field experiments comparing questionnaire administration mode. (Used to provide a basis for comparisons to other surveys and dose-response standards that are based on different questionnaire administration modes.) [Issues 60, 61] 5. Questionnaire for airport officials (Used to identify characteristics of community relations programs, perceptions of community differences, the existence of legal actions or other major community conflicts with the airport)[Issues 45 to 49] 6. Content analysis of local media (Used to estimate the impact of media and to rate the visibility of airport officials actions) [Issue 22] 7. Analyze data from previous US surveys to create comparison for current survey (This will answer the question as to whether noise reactions have changed. For TRACOR studies around 1970 a CNR to DNL transfer function is needed. For other surveys the social survey data would need to be obtained.) [Issue 59] C. Controlling for Previously Identified Explanations: Elements that control for commonly assumed explanations for community differences 8. Meteorological data – short -term and long-term [Issue 24] 9. Detailed airport complaint data on number of complainants, number of complaints, and location of complainants [Issues 55, 56] 10. Aircraft noise exposure estimates for each of the past five or ten years (Needed to control for changes in aircraft noise exposure) [Issues 43, 44] 11. Aircraft noise exposure data for the previous days or months (Used to test hypothesis that annoyance reactions are actually directed at a short period.) [Issue 62] 12. Indicators of ambient noise levels, especially traffic noise level (Rough controls for ambient noise might come from r geo-coded information about road type or population density.) [Issue 20] D. Other Elements 13. Estimate outdoor to indoor attenuation of aircraft noise for each residence (A model to estimate attenuation based on respondent-provided information would need to be developed or, if already available, assessed.) [Issue 17] 14. Supplemental sample of aircraft noise complainants (This would ensure sufficient complainants to compare complainants to non-complaining members of the population.) [Issue 57] 15. Questionnaire data from knowledgeable authorities about the political activism of communities [Issue 23] 16. Aircraft ground operation noise predictions [Issue 38] It is unlikely that all 16 of the elements could be included in a final study plan. The first two elements under “A Primary Study Goals” are clearly required. The elements listed under the B and C headings are listed in approximate order of priority under each heading. The elements under heading D are probably H-47

lowest priority. A final decision about the elements should consider both the priority assigned to the study issues and the costs for including each element. References Note: The six-character identifier, following many references is the unique survey identification code assigned in a catalog of residential noise surveys (Bassarab, Sharp, and Robinette 2009). The first three characters are an abbreviation for the country. The following three-digit number is a unique survey number. American National Standards Institute. 2005. Quantities and Procedures for Description and Measurement of Environmental Sound—Part 4: Noise Assessment and Prediction of Long-Term Community Response. ANSI S12.9-Part 4-2005. Amundsen, A. H., Klaeboe R., and Aasvang, G. M. 2011. The Norwegian Facade Insulation Study: The Efficacy of Facade Insulation in Reducing Noise Annoyance Due to Road Traffic. Journal of the Acoustical Society of America, vol. 129, 3, pp. 1381–1389. NOR-559 (NOT IN CATALOG) Avery, G. C. 1982. Comparison of Telephone Complaints and Survey Measures of Noise Annoyance. J.Sound Vib., vol. 82, 2, pp. 215–225. AUL-214 Bassarab, R., Sharp, B., and Robinette, B. 2009. an Updated Catalog of 628 Social Surveys of Residents' Reaction to Environmental Noise (1943-2008). DOT/FAA/AEE/2009-01; DOT-VNTSC-FAA-10-02; Wyle Report WR-09-18. p. 144, Washington, D.C. Borsky, P. N. 1954. Community Aspects of Aircraft Annoyance. NORC Report no. 54. National Opinion Research Center, Chicago. USA-004 Borsky, P. N. 1965. Community Reactions to Sonic Booms in the Oklahoma City Area. NORC Report no. 101. AMRL Report no. AMRL-TR 65-37. AMRL, Wright Patterson Air Force Base. USA-012 Breugelmans, O., van Wiechen, C. M. A. G., van Kempen, E., Heisterkamp, S., and Houthuijs, D. J. M. 2004. Gezondheid En Beleving Van De Omgevingskwaliteit in De Regio Schiphol: 2002 (Health and Quality of Life Near Amsterdam Schiphol Airport: 2002 Interim Report). RIVM rapport 630100001/2004. NET-533 Brink, M., Wirth, K., Schierz, C., Thomann, G., and Bauer, G. 2008. Annoyance Responses to Stable and Changing Aircraft Noise Exposure. J.Acoust.Soc.Am., vol. 124, 5, pp. 2930–2941. SWI-525 SWI-534 Brooker, P. 2008. Finding a Good Aircraft Noise Annoyance Curve. Acoustics Bulletin, vol. 33, 4. Brooker, P. 2009. Do People React More Strongly to Aircraft Noise Today Than in the Past? Applied Acoustics, vol. 70, 5, pp. 747–752. Brown, A. L. and van Kamp, I. 2009. Response to a Change in Transport Noise Exposure: A Review of Evidence of a Change Effect. Journal of the Acoustical Society of America, vol. 125, 5, pp. 3018– 3029. Connor, W. K. and Patterson, H. P. 1972. Community Reaction to Aircraft Noise Around Smaller City Airports. NASA CR-2104. National Aeronautics and Space Administration, Washington, D.C. USA-022 USA-032 USA-044. Dempsey, T. K., Stephens, D. G., Fields, J. M., and Shepherd, K. P. 1983. Residents' Annoyance Responses to Aircraft Noise Events. NASA TP-2121. National Aeronautics and Space Administration, Washington, D.C. USA-219 H-48

Diamond, I. D., Ollerhead, J. B., Bradshaw, S. A., Walker, J. G., and Critchley, J. B. 1988. A Study of Community Disturbance Caused by General and Business Aviation Operations. University of Southampton, England. UKD-324 Diamond, I. D., Stephenson, R., Sheppard, Z., Smith, A. P., Hayward, S., Heatherley, S., Raw, G. J., and Stansfeld, S. A. 2000. Perceptions of Aircraft Noise, Sleep and Health., Dept. Social Statistics, University of Southampton. UKD-424 UKD-482 UKD-489 Directorate of Operational Research and Analysis. 1978. Aircraft Noise and Sleep Disturbance. DORA Communication 7815. Civil Aviation Authority, London. UKD-147 Directorate of Operational Research and Analysis. 1982a. Reaction to Aircraft Noise Around Hamble Airfield. DORA Communication 8105. Civil Aviation Authority, London. UKD-309 Directorate of Operational Research and Analysis. 1982b. Reaction to Aircraft Noise Near General Aviation Airfields. DORA Report 8203. Civil Aviation Authority, London. UKD-243 Ellermeier, W., Eigenstetter, M., and Zimmer, K. 2001. Psychoacoustic Correlates of Individual Noise Sensitivity. J.Acoust.Soc.Am., vol. 109, 4, pp. 1464–1473. Federal Interagency Committee on Noise (FICON). 1992. Federal Agency Review of Selected Airport Noise Analysis Issues. Fidell, S., Mestre, V., Schomer, P. D., Berry, B., Gjestland, T., Vallet, M., and Reid, T. 2011. A First- Principles Model for Estimating the Prevalence of Annoyance With Aircraft Noise Exposure. J.Acoustic.Soc.Am., vol. 130, 2, pp. 791-806. Fidell, S. A., Pearsons, K. S., Silvati, L., and Sneddon, M. 2002. Relationship Between Low-Frequency Aircraft Noise and Annoyance Due to Rattle and Vibration. J.Acoust.Soc.Am., vol. 111, 4, pp. 1743– 1750. USA-429 Fidell, S. A. and Silvati, L. 1991. An Assessment of the Effect of Residential Acoustic Insulation on Prevalence of Annoyance in an Airport Community. J.Acoust.Soc.Am., vol. 89, pp. 244–247. USA-349 Fields, J. M. 1992. Effect of Personal and Situational Variables on Noise Annoyance: With Special Reference to Implications for En Route Noise. NASA CR-189676, FAA FAA-EE-92-03. Federal Aviation Administration, U.S. Department of Transportation, Washington, D.C. Fields, J. M. 1993. Effect of Personal and Situational Variables on Noise Annoyance in Residential Areas. J.Acoust.Soc.Am., vol. 93, pp. 2753–2763. Fields, J. M., de Jong, R. G., Brown, A. L., Flindell, I. H., Gjestland, T., Job, R. F. S., Kurra, S., Lercher, P., Schuemer-Kohrs, A., Vallet, M., and Yano,T. 1997. Guidelines for Reporting Core Information From Community Noise Reaction Surveys. J.Sound Vib., vol. 206, 5, pp. 685–695. Fields, J. M., Ehrlich, G. E., and Zador, P. 2000. Theory and Design Tools for Studies of Reactions to Abrupt Changes in Noise Exposure. NASA CR-2000-210280. National Aeronautics and Space Administration, Washington, D.C. Fields, J. M. 1997. Reactions of Residents to Long-Term Sonic Boom Noise Environments. NASA TM- 20103. National Aeronautics and Space Administration, Washington, D.C. USA-375 Fields, J. M. and Walker, J. G. 1980. Comparing Reactions to Transportation Noises From Different Surveys: a Railway Noise Vs. Aircraft and Road Traffic Comparison. Noise as a Public Health Problem (Third International Congress), ASHA Report 10, pp. 580–587, American Speech-Language- Hearing Association, Rockville, Maryland. UKD-116 H-49

Finke, H. O., Guski, R., and Rohrmann, B. 1980. Betroffenheit Einer Stadt Durch Lärm, Bericht Über Eine Inter-Disziplinäre Untersuchung. (Objective and Subjective Impact on a Town: Report of an Interdisciplinary Investigation). Umweltbundesamt, Berlin. GER-134 Flindell, I. H. and Witter, I. J. 1999. Non-Acoustic Factors in Noise Management at Heathrow Airport. Noise and Health, vol. 3, pp. 27–44. UKD-(NEW) Franssen, E. A. M., Lebret, E., and Staatsen, B. A. M. 1999. Health Impact Assessment Schiphol Airport: Overview of Results Until 1999. RIVM 441520 012. RIVM, Bilthoven, Netherlands. NET-371 García, A., Faus, L. J., and García, A. M. 1993. The Community Response to Aircraft Noise Around Six Spanish Airports. J.Sound Vib., vol. 164, pp. 45–52. SPA-348 Garnsworthy, J. 1977. A Study of Question Order and Wording Experiments (M.Sc. Dissertation), Dept. of Social Statistics, Univ. of Southampton, England. UKD-116 Gjestland, T., Liasjø, K. H., Granøien, I. L. N., and Fields, J. M. 1990. Response to Noise Around Oslo Airport Fornebu. Report No. STF40 A90189. ELAB-RUNIT, Trondheim, Norway. NOR-311 Griffiths, I. D., Langdon, F. J., and Swan, M. A. 1980. Subjective Effects of Traffic Noise Exposure: Reliability and Seasonal Effects. J.Sound Vib., vol. 71, 2, pp. 227–240. UKD-157 Guski, R. 1999. Personal and Social Variables As Co-Determinants of Noise Annoyance. Noise & Health, vol. 3, pp. 45–56. Guski, R. 2004. How to Forecast Community Annoyance in Planning Noisy Facilities. Noise & Health, vol. 6, pp. 59-64. Hall, F. L., Taylor, S. M., and Birnie, S. E. 1980. Spatial Patterns in Community Response to Aircraft Noise Associated With Non-Noise Factors. J.Sound Vib., vol. 71, 3, pp. 361–381. CAN-168 Hawkins, M. M. 1980. An Exploratory Study of Response To Sound (Including Noise) Occurring in Rural Hampshire and Wiltshire. ISVR Contract Report 80/11. Univ. of Southampton, England. UKD-160 Hume, K., Morley, H. E., Sutcliffe, M. J., Smith, G. R., and Thomas, C. 2005. What Do the Location of Noise Complainants and Noise-Contours Tell Us About the Pattern and Level of Disturbance Around Airports? Inter-noise2005, vol. 208, pp. 1259–1266 Hume, K., Terranova, D., and Thomas, C. 2002. Complaints and Annoyance Caused by Aircraft Operations: Temporal Patterns and Individual Bias. Noise & Health, vol. 4, 15, pp. 45-55. International Organization for Standardization. 2003. Acoustics -- Assessment of Noise Annoyance by Means of Social and Socio-Acoustic Surveys. ISO/TS 15666:2003(E). International Standards Organization, Geneva, Switzerland. International Standardization Organization. 2003. Acoustics - Description, Measurement and Assessment of Environmental Noise - Pt. 1: Basic Quantities and Assessment Procedures. International Standardization Organization. ISO 1996-1-2003. Janssen, S. A., Vos, H., Houthuijs, D. J. M., van Kamp, I., Breugelmans, O. R. P., and Miedema, H. M. E. 2010. The Role of Fear in Aircraft Noise Annoyance. Some Interesting Discussion. Inter-noise 2010, vol. 221, p. 6101. NET-371, NET-355 Janssen, S. A., Vos, H., van Kempen, E. E. M. M., Breugelmans, O. R. P., and Miedema, H. M. E. 2011. Trends in Aircraft Noise Annoyance: The Role of Study and Sample Characteristics. Journal of the Acoustical Society of America, vol. 129, 4, pp. 1953–1962. H-50

Job, R. F. S., Hatfield, J., Carter, N. L., Peploe, P., Taylor,R., and Morrell, S. 2001. General Scales of Community Reaction to Noise (Dissatisfaction and Perceived Affectedness) Are More Reliable Than Scales of Annoyance. J.Acoust.Soc.Am., vol. 110, 2, pp. 939–946. AUL-383 Job, R. F. S., Topple, A., Carter, N. L., Peploe, P., Taylor, R., and Morrell, S. 1996a. Public Reactions to Changes in Noise Levels Around Sydney Airport. Inter-noise 96, pp. 2419–2424, Liverpool, England. AUL-383 Kroesen, M. and Bröer, C. 2009. Policy Discourse, People's Internal Frames, and Declared Aircraft Noise Annoyance: An Application of Q-Methodology. Journal of the Acoustical Society of America, vol. 126, 1, pp. 195–207. NET-NEW Kroesen, M., Molin, E. J. E., and van Wee, G. P. 2008. Testing a Theory of Aircraft Noise Annoyance: a Structural Equation Analysis. J.Acoust.Soc.Am., vol. 123, pp. 4250–4260. NET-599 Lam, K. C. and Chan, T. C. 2007. Effects of Information Bias and Frequency of Riding on Annoyance Response to a Newly Operated Railway Line. Inter-noise 2007, vol. 209, pp. 3815–3822, Istanbul, Turkey. HKG-602 Langdon, F. J. 1976b. Noise Nuisance Caused by Road Traffic in Residential Areas: Parts I, II. J.Sound Vib., vol. 47, 2, pp. 243–282. UKD-071 Lazarus, R. S. 1966. Psychological Stress and the Coping Process. McGraw-Hill, New York. Le Masurier, P. 2007. Attitudes to Noise From Aviation Sources in England: Technical Appendices. UKD-604 Liepert, M., Hegner, A., Möhler, U., Schreckenberg, D., Schümer-Kohrs, A., and Schümer, R. 1999. Lärmbelästigung Durch Schienenverkslärm Vor Und Nach Dem Schienenschleifen. Zwischenbericht Zur Hauptstudie Akzeptanzbefragung. (Noise Annoyance From Railway Noise Before and After Rail Grinding—Acceptability Study). Report Nr. 101-707. Möhler & Partners, Munich. GER-479 Lim, C., Kim, J., Hong, J., and Lee, S. 2008. Effect of Background Noise Levels on Community Annoyance From Aircraft Noise. J.Acoust.Soc.Am., vol. 123, 2, pp. 766–771. KOR-554 Maziul, M. 2005. Socio-Organisational Interface Design Between Airport Residents and Airport Management, University of Dortmund. GER-NEW McKennell, A. C. 1963. Aircraft Noise Annoyance Around London (Heathrow) Airport. S.S.337. The Government Social Survey, Central Office of Information, London. UKD-008 Miedema, H. M. E., Fields, J. M., and Vos, H. 2005. Effect of Season and Meteorlogical Conditions on Community Noise Annoyance. J.Acoust.Soc.Am., vol. 117, 5, pp. 2853–2865. Miedema, H. M. E. and Vos, H. 1998. Exposure-Response Relationships for Transportation Noise. J.Acoust.Soc.Am., vol. 104, 6, pp. 3432–3445. AUL-210 CAN-168 FRA-016 FRA-239 NET-240 NOR-311 NOR-328 NOR-366 SWE-035 SWI-053 UKD-024 UKD-242 UKD-238 USA-022 USA-032 USA-044 USA-082 USA-203 USA-204 USA-338 CAN-120 CAN-121 CAN-168 BEL-122 BEL-137 FRA-092 FRA-239 FRA-364 GER-192 GER-372 GER-373 NET-106 NET-240 NET-258 NET-276 NET-361 NET-362 SWE-142 SWE-165 SWI-053 SWI-173 UKD-071 UKD-072 UKD-157 UKD-242 UKD-238 FRA-063 GER-192 NET-153 NET-276 NET-361 SWE-165 SWE-228 SWE-365 UKD-116 Miedema, H. M. E. and Vos, H. 1999. Demographic and Attitudinal Factors That Modify Annoyance From Transportation Noise. J.Acoust.Soc.Am., vol. 105, 6, pp. 3336–3344. H-51

Miedema, H. M. E. and Vos, H. 2003. Noise Sensitivity and Reactions to Noise and Other Environmental Conditions. J.Acoust.Soc.Am., vol. 113, 3, pp. 1492–1504. MIL Research. 1971. Second Survey of Aircraft Noise Annoyance Around London (Heathrow) Airport. HMSO, London. UKD-024 Moran, S. V., Gunn, W. J., and Loeb, M. 1981. Annoyance by Aircraft Noise and Fear of Overflying Aircraft in Relation to Attitudes Toward the Environment and Community. J.Auditory Res., vol. 21, pp. 217–225. USA-129 Morley, H. E. and Hume, K. 2003. Socio-Economic Status of Aircraft Noise Complainers. 8th International Congress on Noise as a Public Health Problem, pp. 274–275 Nguyen, T. L., Yano, T., Nguyenhuy, Q., Nishimura, T., Sato, T., Morihara, T., and Hashimoto, Y. 2011. Dose-Response Relationships for Aircraft Noise Annoyance in Ho Chi Minh City and Hanoi. 10th International Congress on Noise as a Public Health Problem 2011. VNM-NEW Öhrström, E. 1997. Community Reactions to Railway Traffic - Effects of Countermeasures Against Noise and Vibration. Inter-noise 97, pp. 1065–1070, Budapest, Hungary. SWE-413 Patterson, H. P. and Connor, W. K. 1973. Community Responses to Aircraft Noise in Large and Small Cities in the U.S.A. International Congress on Noise as a Public Health Problem, Dubrovnik, Yugoslavia, May 13–18. USEPA 550-9-73-008., pp. 707–720, U.S. Environmental Protection Agency, Washington, D.C. USA-022 USA-032 USA-044 Pedersen, E., van den Berg, F., Bakker, R., and Bouma, J. 2009. Response to Noise From Modern Wind Farms in The Netherlands . Journal of the Acoustical Society of America, vol. 126, 2, pp. 634–643. NET-NEW Perkins, R., Grimwood, C., Stanworth, C., and Thornely-Taylor, R. 2011. Human Response to Vibration in Residential Environments: A Seven Year Journey to Establish Exposure-Response Relationships. 10th International Congress on Noise as a Public Health Problem 2011, pp. 784–786 Robinson, W. S. 1950. Ecological Correlations and the Behavior of Individuals. American Sociological Review, vol. 15, 3, pp. 351–357. Rohrmann, B. 1976. Community Reaction on Non-Commercial and Sporting Aviation. Inter-noise 76, pp. 427–430. GER-114 Schomer, P. D. 1981. Community Reaction to Impulse Noise; Results of the First Survey. CERL Technical Report N-100. Construction Engineering Research Laboratory, Champaign, Illinois. USA-170 Schreckenberg, D. and Schuemer, R. 2010. The Impact of Acoustical, Operational and Non-Auditory Factors on Short-Term Annoyance Due to Aircraft Noise. Inter-noise 2010, vol. 221, pp. 2164–2173. GER-531 Schreckenberg, D., Schümer, R., and Möhler, U. 2001. Railway-Noise Annoyance and 'Misfeasance' Under Conditions of Change. Inter-noise 2001, The Hague, The Netherlands. GER-546 Schultz, T. J. 1978. Synthesis of Social Surveys on Noise Annoyance. J.Acoust.Soc.Am., vol. 64, pp. 377– 405. AUS-014 AUS-093 BEL-122 BEL-137 CAN-121 DEN-075 FRA-016 FRA-019 FRA-041 FRA-063 GER-034 SWE-021 SWE-025 SWE-035 SWI-053 UKD-008 UKD-024 UKD-071 USA-022 USA-032 USA-044 USA-082 USA-102 USA-203 USA-204 Stallen, P. J. 1999. A Theoretical Framework for Environmental Noise Annoyance. Noise & Health, vol. 3, pp. 69–79. H-52

Tracor, Inc. 1970. Public Reactions to Sonic Booms. NASA CR-1665. National Aeronautics and Space Administration, Washington, D.C. USA-023 Tracor, Inc. 1971. Community Reaction to Airport Noise. vol. I, NASA CR-1761; vol. II, NASA CR- 111316. National Aeronautics and Space Administration, Washington, D.C. USA-022 USA-032 United States Environmental Protection Agency. 1974. Information on Levels of Environmental Noise Requisite to Protect Public Health and Welfare With an Adequate Margin of Safety. 550/9-74-004. USEPA, Washington, D.C. Vader, R. 2007. Noise Annoyance Mitigation at Airports by Non-Acoustic Measures: Inventory and Initial Analysis. D/R&D 07/026 version 1.0. MS/R&D/RES,LVNL-ATC the Netherlands, the Netherlands. Van Gerven, P. W. M., Vos, H., Van Boxtel, M. P. J., Janssen, S. A., and Miedema, H. M. E. 2009. Annoyance From Environmental Noise Across the Lifespan. Journal of the Acoustical Society of America, vol. 126, 1, pp. 187–194. NET-371 and 65 other studies van Kamp, I., Job, R. F. S., Hatfield, J., Haines, M., Stellato R. K.; and Stansfeld, S. 2004. The Role of Noise Sensitivity in the Noise–Response Relation: A Comparison of Three International Airport Studies. J.Acoust.Soc.Am., vol. 116, 6, pp. 3471–3479. van Kempen, E. E. M. M. and van Kamp, I. 2005. Annoyance From Air Traffic Noise: Possible Trends in Exposure-Response Relationships. 01/2005 MGO reference 00265-2005. Bilthoven, The Netherlands. van Wiechen, C. M., Franssen, E. M., de Jong, R., and Lebret, E. 2002. Aircraft Noise Exposure From Schiphol Airport : A Relation With Complainants. Noise & Health, vol. 5, 17, pp. 23–34. Vogt, J. and Kastner, M. 2000. The Role of Information Policy in Annoyance Generation and Reduction. Inter-noise 2000, vol. 204, pp. 2233–2236 Vos, J. 2010. On the Relevance of Nonacoustic Factors Influencing the Annoyance Caused by Environmental Sounds - A Literature Study. Inter-noise 2010, vol. 221, pp. 7445–7454 Wehrli, B., Nemecek, J., Turrian, V., Hofmann, R., and Wanner, H. U. 1978. Auswirkungen Des Straßenverkehrslärm in Der Nacht. (Translation Available in: Effects of Street Traffic Noise in the Night. NASA TM-75495). Kampf Dem Lärm, vol. 25, pp. 138–149. SWI-173 Willkie, W. J., Madgwick, F. R., Sweatt, R. A., Frievson, C. A., and Carlton, E. M. 1990. Community Involvement Manual. FAA-EE-90-03. HNTB, Alexandria, Virginia. Wirth, K. and Bröer, C. 2004. More Annoyed by Aircraft Noise Than 30 Years Ago? Some Figures and Interpretations. Inter-noise 2004, Prague, Czech Republic. SWI-525 Woodward, J. M., Briscoe, L. L., and Dunholter,P.: 2009. ACRP Report 15: Aircraft Noise: a Toolkit for Managing Community Expectations. Transportation Research Board of the National Academies, Washngton D.C. Wyle Laboratories. 1971. Community Noise. NTID300.3. U.S. Environmental Agency, Office of Noise Abatement and Control, Washington, D.C. Wyle Research. 2011. Updating and Supplementing The Day-Night Average Sound Level (DNL). WR- 11-04, DOT/FAA/AEE/2011-03. Wyle (Report to Volpe), Arlington, VA. Yamada, I., Kaku, J., Yokota, T., Namba, S., and Ogata, S. 2008. A New Method of Social Survey on Transportation Noise Using the Internet and GIS. The 9th Congress of the International Commission on the Biological Effects of Noise, Foxwoods, CT. JPN-616 Yano, T., Sato, T., and Morihara, T. 2007. Dose-Response Relationships for Road Traffic, Railway and Aircraft Noises in Kyushu and Hokkaido, Japan. Inter-noise 2007, Istanbul, Turkey. JPN-565 H-53