Survey Methodology and Screening The BPHC requires data collection to follow a random-digit-dial (RDD) method utilizing a computerassisted telephone interview (CATI) system. All Boston households with landline telephones are to be included in a landline frame. In addition, a substantial portion of interviews will be collected in a cell phone frame of non-institutionalized Boston caregiver adults ages 18 and older who use a cellphone. We expect the proposed sampling plan will utilize a disproportionate stratified random sampling method in which high density stratum will be sampled at a higher rate than the medium density stratum to increase the efficiency of the data collection process. The proposed sampling plans should be described in detail and should include plans to periodically evaluate the sampling strategy and adjust to improve response rates while maintaining proportional geographical representation (please see Small-Area Sampling section below). We would like cost estimates for two landline/cell-phone combination options. These are overlapping dual frames in which some percentages of interviews are collected within the landline phone frame, and the remainder are collectedwithin the cell-phone frame. In this mix, the cell-phone users are not necessarily cellphone only users. Please see the cost estimate summary table below for more information on the proportion of landline and cellphone in the sample options. Survey methodology should be conducted according to standard survey methods, with exceptions noted here. The survey partial-complete point (the point after which partially completed interviews will be accepted toward the finalsample size count) is set at a minimum80% completion of the survey instrument. Partials may constitute a maximum of 5% of the sample total for each of the frames. Additionally, two residency screener questions are to be included in the respondent selection process; these are city/town of residence and zip code of residence (Boston zip codesfor purposes of this survey include and are limited to: 02108, 02109, 02110, 02111, 02113, 02114, 02115, 02116, 02118, 02119, 02120, 02121, 02122, 02124, 02125, 02126, 02127, 02128, 02129, 02130, 02131, 02132, 02134, 02135, 02136, 02199, 02210, 02215, 02163, and 02467.) Respondents who do not reside in Boston (i.e., do not meet the residency screener questions) will not be eligible for participation. The acquisition of the landline and cell-phone sampling frames is the responsibility of the applicant and should be included in the proposal. Responses to this RFP should clearly outline their understanding of the appropriate survey methodology, and further should specify methodology assumptions, such as net incidence, coverage, and response rate. In addition to the two residency screeners, there will be a third screener question to ascertain a child residing in the household. Besides, the child health survey has two parts: a) the selection questionnaire; b) topical questionnaire. The selection questionnaire collects information about the presence of all children within the household, child demographic information, and basic questions about each child’s health. One child will be selected from the completed first part, and one of three age-based topical questionnaires is provided to the household based on the sampled child’s age and health status. Small-Area Sampling In similar population-based health surveys, we have utilized cell phone sampling methods that are more commonly applied for efficient sampling in small-area research (county and city level). This included a cell sample design based on telephone exchanges that have their rate centers (the midpoint of the rate area) in the vicinity of the Boston area. Applicants should indicate their understanding and experience with cell phone sampling in geographically small areas; and propose a methodology for cell phone sampling in the city of Boston. Respondent Oversample In order to achieve a survey sample which is more representative of specific Boston community groups, the BPHC may conduct an oversample in specific zip codes. Please provide your oversampling methodology targeting under-represented groups. In addition, we would also like to know more about your experience with oversampling.
