Recruitment (for example for adults)

Inclusion criteria. Within each neighborhood, a sample of households should be randomly drawn. One adult between the ages of 20 and 65 from the selected households should be eligible to participate. Samples should be balanced by gender and should approximate the ethnic distribution of the population. Participants must agree to participate in two assessment rounds and provide signed informed consent.

Exclusion criteria. Individuals in group living establishments (e.g., nursing homes, dormitories, military barracks), unable to walk, unwilling to wear the CSA monitor, or unable to complete written surveys in English should be excluded.

Recruitment. The objective of the sampling strategy is to represent adults who reside in SES-matched samples of high and low walkability census tracts in each location. The sampling frame should be constructed in several steps.

The listing of street addresses from a published city directory in conjunction with maps should be used in each location to compile a list of all street addresses in targeted tracts.

Sampling should proceed in several stages. First, an overlay of census tracts should be drawn on a detailed street map of each city. Second, city blocks within each selected tract should be enumerated and listed along with current total populations. Blocks within each tract should be ordered according to population size, and target blocks should be selected proportionately to population size so that the resulting sample is self-weighted at the tract level. Dwelling units listed in the directory for each selected block should be numbered separately for units with and without telephone listings. Telephone and non-telephone listings should be randomly sampled from the blocks using the random number generator in Microsoft Excel. Respondent selection within a household should be implemented by targeting the adult with the most recent birthday, a practice that should produce a sample that reflects the age-gender distribution of adults in the households accurately. Computer listings for sampled dwelling units should then be compiled to facilitate fieldwork and tracking of respondents.

We estimate that approximately 40% of residential listings will include correct, working telephone numbers. The rest will have only street addresses. Using a random number generator in Excel, target samples of telephone and non-telephone residences should be selected for each block within each tract and sent a recruitment packet.

For telephone households, packets with a cover letter and consent form should be mailed to the head of household. Two or three days after expected delivery, trained telephone interviewers should contact each target telephone number, introduce the project, encourage participation, and request to speak to the adult resident with the most recent birthday. We expect that our procedure should result in a sample of adults stratified by gender, and the sampled person should be solicited to participate in the project. Participants should be required to sign an informed consent statement and mail it back in a prepaid envelope. Up to six callbacks should be made to each household. The cover letter and interviewers should highlight the incentive for Time 1 measurement and larger incentive for Time 2 measurement. The cover letter should request the head of household to complete the consent form and mail it back, in case the interviewer is unable to make contact. This procedure should increase recruitment, given that some of the telephone numbers in the Directory should be incorrect. For people who refuse, a telephone interviewer should call and attempt to convert them to participants during an additional callback. If one adult refuses again, we should attempt to recruit another adult of the same gender in the household. When the gender quota for a tract has been met, only participants of the opposite gender should be recruited. A secured computer master list of names, addresses, and telephone numbers of those who agree to participate should be constructed for tracking purposes.

For non-telephone households, a project packet should be mailed containing a letter introducing the project and participant selection process, a consent form, a return envelope, and a brief questionnaire soliciting the telephone number, name, age and gender for each adult resident. Incentives should be highlighted. Up to three follow-up packets should be mailed to each household at intervals of three weeks to enhance total response. Returned questionnaires should be used to select random samples within households, and contact information should be added to the master file of participants. When a signed informed consent statement is received, the participants' contact information should be sent to the measurement staff who should commence measurement.

Other Participants Information