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.

5 Sample and Experiment Design A sample of approximately 2,200 addresses was selected for each of the three airport communities in the Study. These three airports were selected purposively from the international airports in the United States to represent a range of climate conditions and airport operations. Although the three airports in the ACRP Study are not representative of the population of airports in the United States, the purpose of the ACRP Study is to compare results of surveys conducted by telephone with those of surveys conducted by mail. This comparison is done by looking at potential differences between mail and telephone respondents separately within each airport community. HMMH provided Westat with noise exposure contours for all three airports. These contours started at 50 dB for the day-night average sound level (DNL), and ended with the maximum noise exposure in areas off-airport. These contours were used to stratify the population from each airport community, using strata 50-55 dB, 55-60 dB, 60-65 dB, 65-70 dB and 70+ dB. The number of addresses falling in each of those strata was then determined for each airport community. In Airport 2, there were sufficient addresses in all five DNL noise strata to allow sampling in all of them. Airport 1 had very few addresses in the population with noise exposure greater than 65 dB; therefore, for Airport 1, only three DNL noise strata were used: 50-55 dB, 55-60 dB, and 60+ dB. Previous literature indicates that, in general, the percent of the population who are highly annoyed increases with the noise exposure (see, for example, Schultz, 1978; Fidell and Silvati, 2004). To allow comparison of the percent highly annoyed (HA) for the telephone and mail respondents across the range of noise exposures, the sample was divided equally among the DNL noise strata for each city. In Airport 1, 732 addresses were sampled from each of the three DNL noise strata (50-55 dB, 55-60 dB , and 60+ dB); in Airport 2, 440 addresses were sampled from each of the five DNL noise strata (50-55 dB, 55-60 dB, 60-65 dB, 65-70 dB, and 70+ dB). The surveys for Airport 1 and Airport 2 were conducted between March and June, 2013.7 Preliminary returns from Airport 1 and Airport 2 showed a smaller than anticipated sample size of respondents in the higher (65+ dB) DNL noise strata. To obtain more information from households exposed to higher noise levels, in the Airport 3 sample, 384 addresses were selected from the 50-55 DNL noise stratum, 400 addresses from the 55-60 dB DNL noise stratum, 420 addresses from the 60-65 dB DNL noise stratum, and 986 addresses from the 65-70 dB and 70+ dB strata. The survey in Airport 3 was conducted between July and September, 2013. The sample drawn from each airport community and DNL noise stratum was randomly divided into two halves: one-half of the addresses were randomly selected for the telephone interview (Appendix A), and the other half were sent the mail questionnaire, referred to as the Survey of Community Attitudes (Appendix D). The addresses selected for the telephone interview were matched against directory information to find telephone numbers associated with the address. For addresses found to have a matching telephone number, an advance letter was sent introducing the survey, and then the number was called. Addresses for which no matching number was found, or where the matched number was invalid, were randomly divided into two subgroups in the Airport 1 and Airport 2 samples. Half of the nonmatching addresses within each airport community and DNL noise stratum were sent a “long” screener which included questions from the mail survey instrument and a question asking for the telephone number (see Figure 3). The other half were sent a “short” screener that asked only for the telephone number (Appendix B). The telephone numbers returned by respondents to either the long or short versions were 7 Two additional surveys were returned by mail in July, 2013, and were included in the data analysis. 13

called and the telephone survey was administered. Figure 3 provides a schematic summary of the design of the study for Airport 1 and Airport 2. Figure 3 ACRP Study protocol for Airport 1 and Airport 2 As a result of the low response rates in the telephone groups in Airport 1 and Airport 2, the design was modified for Airport 3 as described in Appendix E. For the Airport 3 sample, all nonmatching addresses were sent the “long” screener to procure a telephone number. Figure 4 shows the procedure used in Airport 3. All Addresses in noise stratum for airport Telephone Survey (1/2, randomized). Attempt to find telephone number that matches address. Mail Survey (1/2, randomized). Send Community Environmental Questionnaire. Person with next birthday asked to fill out the questionnaire Valid matching telephone number found. Call number and administer telephone survey. No valid matching telephone number found. Randomly split into two groups. Send “short” screener requesting telephone number and number of occupants (1/2, randomized). Any household member asked to fill out the screener. Call number and administer telephone survey. Send “long” screener requesting telephone number, number of occupants, and questions from the Community Environmental Questionnaire (1/2, randomized). Any household member asked to fill out the screener. Call number and administer telephone survey. 14

Figure 4 ACRP Study protocol for Airport 3 Any household member could fill out the long or short screener that was sent to request a phone number from households with no matching number in the telephone group. For the Survey of Community Attitudes sent to the mail group, the person with the next birthday was asked to fill out the questionnaire. All initial requests, including the advance letter for the telephone survey, were sent with a $2 bill to encourage cooperation. All survey instruments were available only in English. All Addresses in noise stratum for airport Telephone Survey (1/2, randomized). Attempt to find telephone number that matches address. Mail Survey (1/2, randomized). Send Community Environmental Questionnaire. Person with next birthday asked to fill out the questionnaire Valid matching telephone number found. Call number and administer telephone survey. No valid matching telephone number found. Send “long” screener requesting telephone number, number of occupants, and asking questions from the Community Environmental Questionnaire. Any household member asked to fill out the screener. Call number and administer telephone survey. 15