OMB Information Collection Request
Part B
Testing of Sexual Violence Definitions and Recommended Data Elements in Three Different Racial/Ethnic Minority Communities
Centers for Disease Control and Prevention
National Center for Injury Prevention and Control (NCIPC)
Division of Violence Prevention (DVP)
Levator Brown
Project Officer
Phone: (770) 488-1530
Fax: (770) 488-4349
E-mail: [email protected]
Phase 1. In Phase 1, half of the 12 interviews for each targeted minority group will be conducted in either the Chicago or Washington, DC area to allow for cost and personnel efficiencies as well as ease of recruiting from known organizations to the data collection contractor. However, to ensure adequate coverage of local or regional differences, we will conduct the remaining half of the interviews in or near the geographic locations chosen for Phase 2 data collection. We will use the South Houston, Texas area for the Hispanic sample frame, Atlanta, Georgia for the African American sample frame, and Seattle, Washington for the American Indian sample frame.
In each area, a convenience sample drawn from recruitment efforts will be used to conduct the cognitive interviews, including some Spanish-speaking women to test the Spanish translation of the questionnaire. Therefore, the respondent universe can be considered those members of the targeted racial/ethnic groups who are likely to come into contact with the contractor’s recruitment materials during the recruitment process. The recruitment flyer and recruitment and screening script for the cognitive interviews are presented in Appendices 3 and 4, respectively.
Table 7 provides the recruitment sources identified for each of the Phase 1 locations. Note that agencies in each of the five locations have been contacted, but will be asked to participate in Phase 1 only after study approval has been received. Additional source organizations will be identified for the Seattle area. These organizations will have some affiliation with the Urban Indian Health Institute (UIHI), which is assisting with identifying the American Indian sample. Therefore, this list should be considered provisional in nature.
Table 7. Referral Organizations for the Cognitive Test
Chicago, IL |
Washington, DC |
South Houston, TX |
Atlanta, GA |
Seattle, WA |
YWCA Metropolitan Chicago |
DC Rape Crisis Center |
Texas Council On Family Violence |
Gwinnett Sexual Assault Center |
Urban Indian Health Institute |
Rape Victim Advocates |
District of Columbia Coalition Against Domestic Violence |
Texas Association Against Sexual Assault |
Georgia Coalition Against Domestic Violence |
|
Illinois Coalition Against Sexual Assault (ICASA) |
Asian/Pacific Islander Domestic Violence Resource Project |
Women’s Shelter of South Texas |
Grady Rape Crisis Center |
|
Howard Brown Health Center |
Ayuda |
Friendship of Women, Inc. |
DeKalb Rape Crisis, Inc. |
|
Staff in the service delivery organizations chosen for Phase 1 will be briefed on the project, its goals, population of interest, and the role we are asking the organizations to play. Following these briefings, NORC staff will ask agency staff to provide the recruitment flyer to potential respondents. Recruitment flyers also will be posted at these organizations to inform victims directly. Respondents who call the number listed will be administered the recruiting script. Potential Phase 1 respondents will be screened to ensure that the cognitive interviews are conducted with women of the targeted racial/ethnic background. We also will seek to include adult women of various ages and other demographic characteristics such as marital status and education, though we do not intend to assemble a representative sample. The main purpose of the cognitive test is not to gather data about individuals and their experiences but to test the adequacy of our questionnaire for the purposes of the main survey in Phase 2.
Phase 2. For Phase 2, we again propose to study Hispanics in South Houston, Texas, and African Americans in Atlanta, Georgia. Both communities are densely populated by the respective racial/ethnic group, which is crucial in containing the screening costs associated with determining racial/ethnic eligibility. Based on the latest census data, 77.9 percent of the population in South Houston is Hispanic or Latino, and 58.6 percent of the population in Atlanta is African American. Additionally, sampling in these urban communities will be cost effective due to availability of the United States Postal Service (USPS) Address List that can serve as a convenient sampling frame. As a result, traditional field listing procedures will not be necessary. Finally, South Houston is an urban community of manageable size from a cost perspective, and Atlanta is partly chosen because of its proximity to the funding agency.
While there are many choices for Hispanic and African American communities, the selection of an American Indian community is a challenging task. An ideal American Indian community would have (1) high density of the target population, and (2) an adequate USPS address list so that field listing of housing units can be avoided. This is especially problematic on Indian reservations as they tend to small in population, located across large geographical areas, and fairly closed communities. After exploring several different sampling options, we elected to take a sampling approach for the American Indian sample that more closely mirrors the approach we have taken in South Houston and Atlanta. That is, we have engaged the services of the UIHI to assist in drawing the sample. UIHI is affiliated with organizations in 34 cities, which represents 92 counties in 19 States. We propose to select an American Indian sample for Phase 2 from the client lists of approximately three UIHI-affiliated organizations in the Seattle area. The clients in these organizations are predominantly American Indian. The number of organizations is contingent on our ability to identify a sufficient number of American Indian to permit us to complete 200 interviews in this population. Although this sample will not allow us to generalize to the population of all American Indian women, it will generate a sample of eligible persons with rather low cost and it will provide an urban sample, which is consistent with the Hispanic and African-American sample locations.
Below, we provide the Phase 2 sampling plan for each community.
South Houston Sampling Plan
The target population of this community is adult Hispanic women living in households within the city boundary during the data collection period. Sample selection will be carried out in two stages: the selection of housing units and the selection of an eligible adult woman from the households.
We recommend approximately 200 completed interviews per community, which will ensure that a sample proportion estimate is within 7 percent of the population true value. To estimate the initial household sample size, we start with the expected 200 complete interviews and work backwards. We will take into account the following factors: expected rate of occupied housing units, expected screener completion rate, expected racial/ethnic eligibility rate, expected percent of screened households with at least one eligible woman, and main interview completion rate. Based on the various assumed rates, Table 9 shows that we should select a sample of approximately 481 housing units to complete 200 interviews.
The first stage household sample will be selected systematically from the USPS list frame. There are 5,120 city style addresses1 on the USPS list in South Houston, which represents more than 90 percent of all residential addresses in the city. To select a sample of 481 households, we will first geocode all addresses in the ZIP codes that cover South Houston to identify those that are in the target area. Using SAS, we will sort the addresses by census tract, ZIP code, carrier route, and walk sequence and, then, select a systematic sample. This sampling technique ensures that the resulting sample is diverse geographically and represents all neighborhoods in South Houston.
The second stage sample of eligible Hispanic women will be selected from the households. To be eligible for the study, a woman must be 18 years of age or older, live in a sample household, and be of Hispanic heritage. If more than one woman is eligible in a household, the woman with the most recent birthday will be selected.2
Table 9: Sample Size Estimate for South Houston and Atlanta
Racial/Ethnic Group |
Hispanic |
African American |
Community |
South Houston |
Atlanta |
Total HH* in Frame |
5,120 |
185,110 |
Sampling Rate |
9.40% |
0.39% |
Sample Size |
481 |
722 |
% Occupied HH |
93% |
82% |
Total Occupied HH |
447 |
594 |
Screener Completion Rate |
90% |
90% |
Total Screener Completes |
402 |
535 |
% Race/Ethnic Pop |
78% |
59% |
Total Target Race/Ethnicity HH |
313 |
313 |
% with eligible women |
85% |
85% |
Total Eligible Women |
266 |
266 |
Interview Completion Rate |
75% |
75% |
Total Completed Interviews |
200 |
200 |
*HH means number of households.
Atlanta Sampling Plan
The target population of this community is adult African American women living in households within the city boundary during the data collection period. Sample selection will be carried out in two stages: the selection of households or housing units and the selection of an adult woman from eligible households.
We recommend approximately 200 completed interviews per community, which will ensure that a sample proportion estimate is within 7 percent of the population true value. To estimate the initial household sample size, we start with the expected 200 complete interviews and work backwards. We will take into account the following factors: expected rate of occupied housing units, expected screener completion rate, expected racial/ethnic eligibility rate, expected percent of screened households with at least one eligible woman, and main interview completion rate. Based on the various assumed rates, Table 9 shows that we should select a sample of approximately 722 housing units in the Atlanta area to complete 200 interviews with African-American women.
The first stage household sample will be selected systematically from the USPS list frame. There are 185,110 city-style addresses on the USPS list in the city of Atlanta, which represents 91 percent of all residential addresses in the city. To select a sample of 722 households, we will first geocode all addresses in the ZIP codes that cover Atlanta to identify those that are in the target area. Using SAS, we will sort the addresses by census tract, ZIP code, carrier route and walk sequence and then select a systematic sample. This sampling technique ensures that the resulting sample is diverse geographically and represents all neighborhoods in Atlanta.
The second stage sample of eligible African-American women will be selected from the identified households. To be eligible for the study, a woman must be 18 years of age or older, live in a sample household, and be African-American. If more than one woman is eligible in a household, the woman with the most recent birthday will be selected.
Seattle Sampling Plan
The target population of this community is adult American Indian women in the Seattle metropolitan area. While this target population is not as general as those in the other two areas (i.e., we are drawing from specific program lists rather than the Seattle population as a whole), we feel the approach offers a good compromise between generalizability and cost for this study. It also offers a population of American Indian women who are residents of urban areas, which is the case for the Hispanic and African-American women samples. This allows for a degree of comparability across the sample locations. Finally, urban Indians represent a particularly understudied American Indian population.
The American Indian sample will be drawn from program lists of various American Indian programs in the Seattle area. The primary sample source will be the Urban Indian Health Institute (UIHI), which has a number of clinics in the Seattle area. However, the Executive Director of UIHI, Miale Taualii, has contacts with a variety of programs designed for American Indians. Ms. Taualii will secure client lists from other programs to avoid the potential bias that might arise from a sample drawn from a self-selected sample of women who present at the clinics with health issues. In addition, NORC will work with UIHI to demonstrate the comparability, or lack thereof, between the various programs we use to draw potential respondents and the Seattle population as a whole.
As shown in Table 10, we estimate that we will need to select 314 women to complete 200 interviews. The actual selection of these 314 women from a variety of programmatic client rolls will be performed by UIHI staff using specifications developed by NORC. Because the data on these client rolls are quite sensitive, we believe it will be more efficient to have the sample selected by those with access to the data, rather than going through a lengthy process to strip the data of its identifiers and transfer it from UIHI to NORC and back. Ms. Taualii has considerable research, statistical, and methodological skills and understands the need to develop a statistically valid sample. She has worked with this population for some time and has addressed issues associated with sampling; therefore, we are confident that she has the skills necessary to ensure that the sample meets the high standards established by NORC.
This sample differs from those discussed above: it will be a single-stage sample from a list frame. There is no need to select households and then women within households. We will specify that the frame purged of all clients who are not American Indian women. Before systematic selection is performed, the frame will be sorted by age. This implicit stratification will ensure that the resulting sample contains women of a variety of ages.
What remains to be decided concerning this sample is the number of organizations from which we will sample. We hope to keep the number of clinics as small as possible to control costs and the amount of work involved in selecting the sample. We will work with Ms. Taualii and her staff to identify the requisite number of organizations that are able to provide us with the numbers we require to complete 200 interviews.
Table 10: Sample Size Estimate for Seattle
Racial/Ethnic Group |
American Indian |
Community |
Seattle |
Number Selected |
314 |
% locatable. eligible women |
85% |
Total Eligible Women |
267 |
Interview Completion Rate |
75% |
Total Completed Interviews |
200 |
Below, we outline our procedures for the collection of cognitive interview data (Phase 1) and main data collection (Phase 2). While we do not expect Phase 1 to result in significant changes, cognitive testing is designed to identify possible refinements to the instruments and procedures. Should such modifications be required, we would submit the necessary modifications to OMB (IRB).
Phase 1. The tasks for identification of recruitment sources, screening of potential respondents, and conducting the cognitive interviews are overlapping. While we anticipate developing a thorough list of recruitment sources early in July 2007, we will seek continually new opportunities to recruit respondents, particularly if we are in need of respondents who meet certain demographic characteristics such as race, ethnicity, or other key study traits. We also plan to begin conducting cognitive interviews while we are still recruiting respondents. This approach will allow us to tailor our recruiting efforts based on the initial findings of our cognitive interviews. For example, we may determine that one recruiting source consistently provides only those respondents who have been sexually victimized by intimate partners rather than strangers, acquaintances, or co-workers. In this instance, we would broaden our recruiting sources to include other agencies whose clients are more likely to have been victimized by non-intimate partners.
Recruiting Participants
Potential respondents will be screened to ensure that the cognitive interviews are conducted with women of the targeted racial/ethnic background. We also will seek to include adult women of various ages and other demographic characteristics such as race, martial status, age, and education, though we do not intend to assemble a representative sample in the pilot test. As respondents are recruited, we will track their demographic characteristics. If we find that the group is not sufficiently diverse, we will work with the recruitment sources to help us establish the required sample diversity. We will include some Spanish-speaking women to test the Spanish translation of the questionnaire. While the main purpose of Phase 1 is not to gather data about individuals and their experiences but to test the adequacy of our questionnaire for the purposes of the study, it may well be that Spanish-speaking respondents interpret questions differently from English-speaking respondents. We will seek to understand any differences in comprehension of the questions as it relates to the Spanish translation of the instrument. If the results of the cognitive interviews suggest that there are differences in the interpretation of the Spanish and English version of the questions, we will revise the Spanish translation to better suit the intended meaning of the English questions.
We also will work with our recruiting sources to ensure that the majority of women are known to be victims of sexual violence by explaining to agency staff the purpose of the study and the importance of testing the instrument with women who have experienced sexual violence. However, we will not require specific information regarding the details of the respondents’ victimization to determine if the respondent is eligible for cognitive interviewing in Phase 1. If we discover during the interview that a respondent has not been a victim of sexual violence, we will continue to conduct the interview, pay the incentive, and thank the respondent for her time. Indeed, the experience of non-victims is of interest to us as we test the data elements, structure, and wording of the questionnaire. Therefore, we will seek to interview approximately six non-victims. We will monitor the number of non-victims over the course of conducting the cognitive interviews. Should we approach the target number of non-victim interviews, we will modify our recruiting strategy, for example, by focusing on specific sources, by working more closely with those sources, by possibly identifying new sources, etc.
Administration of Cognitive Interviews
The cognitive interviews will be administered in person. Face-to-face interviewing will offer the greatest opportunity to build rapport with the respondent, foster a trusting, candid exchange between interviewer and respondent, and provide a more accurate test of study protocol. Interviews will be conducted in the facilities of the recruitment source by experienced cognitive interviewers. The NORC cognitive interviewers have considerable experience administering cognitive interviews and are well schooled in the purpose of these interviews. The interviewers will receive a 2-day training on the interview questionnaire and the purpose of the study and each question. If the agencies’ facilities do not offer sufficient privacy, the agency requests that interviews take place at another location, or the respondent requests a different location, we will conduct the interviews in either the data collection contractor’s office in Chicago or Washington, DC. Each office offers secure office space with sufficient privacy for this work (office space with doors and sound-proof walls). For the interviews that take place in local data collection areas (i.e., South Texas, Atlanta, and Seattle) as opposed to areas in which the contractor has offices, the contractor will secure use of private office/conference room space in hotel or meeting facilities. If the respondent prefers, we will conduct the interview in her home (though we will not request this location due to concerns about privacy and safety). We will consider conducting some interviews by phone only if we cannot identify a sufficient number of respondents willing to conduct the interview in person. In all facets of our scheduling and administration of the interviews, the contractor will be keenly aware of the imperative to protect the confidentiality and safety of the respondents.
While some cognitive interviewing efforts focus on particular areas of a questionnaire, the contractor will administer all sections of the instrument as appropriate, following skip patterns where indicated in the questionnaire. Systematic and complete testing will allow the contractor to assess the data elements, wording, and structure of the instrument when administered to the targeted minority populations. We have inserted probes after each section of the instrument, as well as at the end.
Throughout the cognitive interviewing effort, we will conduct frequent debriefing sessions with the interviewers to assess the effectiveness of the probes, share insights into the preliminary findings from the interviews conducted to date, and assess the appropriateness of the recruiting techniques in identifying the most suitable respondents. This approach of frequent meetings will permit us to maximize the value of resources allocated for cognitive interviewing work by allowing us to make mid-course corrections in our approach prior to exhausting time and funds allocated for Phase 1.
Phase 2.
Interviewer Staffing and Training. Recruiting will begin by identifying current, experienced interviewers who express interest in the study and are located near the communities of interest. They will be supplemented with newly hired interviewers who will undergo extensive screening and training. Theses newly hired interviewers bring to the project a deep familiarity with the physical setting, local norms, and – critically – language of the targeted communities.
All interviewers will attend two-day in-person project training. Newly hired interviewers preface their project training with a one-day general interviewer training that is non-project specific and covers basic interviewing protocols. Topics to be covered include: Introduction and Purpose of the Study, Study Design and Methodology, Gaining Respondent Cooperation, Working Sensitively with Different Types of Respondents and Gatekeepers, Initial Contact with Respondents, Screening for Eligibility, Conducting Informed Consent, Conducting Interviews, Post-Interview Edit Check, Real-Life Issues That May Arise, Maintaining Confidentiality, and Case Management.
The instruction style takes a variety of forms: pre-training homestudy, video (when appropriate), lecture from project staff, lecture from outside scientific experts whose work focuses on the issues examined by the project, role play with full and small groups, private practice, private appointments with a trainer to review topics one-on-one, and listening to audio tips of expert interviewers. At the conclusion of training, each interviewer completes a certification exercise to demonstrate his/her sensitivity, gaining cooperation, and interviewing skills. Only interviewers exhibiting high proficiency in the assessments will be permitted to begin on the project.
Inherent in this training model is that during the two (or three) day period, trainees meet many different project staff, all of whom are intimately familiar with the project and filled with enthusiasm for it. Interviewers depart training imbued with this enthusiasm and highly motivated to treat their respondents with sensitivity and respect and to collect high quality data.
After in-person training ends, the training program continues in various ways. This may include: a practice case at home that must be reviewed with the supervisor before beginning real assignments; periodic booster trainings that reinforce earlier lessons, additional practice as needed, or advice; field newsletters/interviewer memos – eagerly anticipated materials provided each Friday that communicate accolades, “stories from the field,” and useful lessons and procedures; and ongoing supervision that ensures close contact and continual review of caseloads, strategies, and procedures.
Contacting Respondents and Gaining Cooperation. Approximately one week prior to the initial contact with respondents, an advance letter will be mailed to each household in the sample (see Appendix 7). The letter, on CDC letterhead, will inform the respondent that CDC is conducting a study of women’s health in their area, state the length of the survey, the respondent payment, and the toll-free number that respondents can call to confirm or learn more information about the study. The letter will help set the tone in the respondent’s mind that the study is legitimate and prepare them for the interviewer’s visit. The letter is appropriate to all households – set at a 6th grade reading level, and translated into Spanish as appropriate.
Field interviewers will make the first approach to a sampled household in person, or by telephone if that avenue seems to be preferred by the respondent. They will follow the approved protocol, selecting one adult per sampled household. Interviewers will carry copies of the advance letter and respondent-geared study documents to help explain the study. The most crucial step in the contacting phase is to gain respondent’s cooperation. The interviewer’s friendliness, professionalism, sensitivity, and yes, persistence, are always the factors respondents remember. Field interviewers are further supported by their field managers who strategize how to interpret particular situations and how the interviewer – or perhaps a different interviewer – can connect with and convert individual cases.
Administering the Questionnaire. After securing informed consent, interviewers will administer the questionnaire aloud to the respondent using a hard-copy questionnaire in the language version required by the particular respondent (either English or Spanish). Interviewer objectivity has always been at the heart of research, but when conducting highly sensitive interviews something more than straight objectivity is required. Interviewers will be trained to be fully objective, as well as sensitive to the human issues in such a way as to encourage the respondent toward open, full, and honest answers. Such techniques are taught and modeled in training: asking the questions in a straightforward way, defining graphic (“penetration”), painful, or vocabulary terms (“anal” or “rectum”) with equanimity, making appropriate eye contact with the respondent, using body language that projects the interviewer’s patience to wait for the full answer to emerge – either here or at a later question. These techniques are effective and provide affirming experiences for the interviewer and respondent alike. Because respondents will be asked about abuse and victimization, the protocol will ensure that the questions can be asked and answered in absolute privacy without the possibility of another household member overhearing. We may seek out-of-home locations for the interview, or allow the respondent to self-administer questions so he/she cannot be overheard. We will also employ procedures and resources for reacting to respondents in distress, for example, counseling services for the agitated respondent. It is recommended that all interviews be conducted in person, mainly to ensure that interview is conducted in private and to enhance data quality. However, if a respondent prefers a telephone interview, it is possible to accommodate that mode of data collection.
Editing Interview Data. Immediately after completing the questionnaire, interviewers will review it to ensure clarity of responses and that skip patterns and instructions were followed correctly. This will occur prior to concluding the meeting with each respondent. In this way, responses can be modified. This process will reduce the number of errors and improve overall data quality, and will serve as a confidence booster for interviewers as they acclimate to administering a sensitive instrument. Self editing is an essential component for maintaining high data quality.
Data Entry. To monitor the quality of its operations, NORC’s sub-contractor for data entry DSI will conduct an independent verification of its data entry by re-keying 10 percent of all questionnaires. This verification process allows DSI to evaluate the individual performance of each data entry clerk and will ensure all data are captured consistently and accurately. An additional independent verification will be completed by NORC to further ensure the delivered data meet our high quality standards of less than 1 percent data entry error. DSI will pick up questionnaires from NORC and data will be delivered on a continuous flow basis, resulting in efficient transfer from hardcopy questionnaires to electronic data file. Test files will be provided to NORC prior to data collection and rigorously tested to ensure that the data entry program produces fully compliant data consistent with earlier data sets, and functions properly prior to production implementation.
Quality Assurance of Data. As an additional step, we will conduct call-back validations to approximately ten percent of each interviewer’s completed cases. NORC uses this technique to ensure that each interviewer, in fact, completed each reported interview. Field interviewers operate without direct supervision; the call-back procedure provides a measure of field supervision. Examples of the kinds of questions to be asked include:
Good morning. This is ____________ from the National Opinion Research Center. One of our interviewers was to conduct an interview with you on ______________. Can you tell me, did she conduct the interview with you?
About how long did the interview take?
Do you have any questions about the interview that I might answer?
We want to thank you once again for talking to us. This study is very important, and you’re answers will be helpful to the findings. Thanks again. Goodbye.
The protocol for these calls will be such that respondent privacy and confidentiality are not compromised. A third pillar of quality is ongoing data review, which can identify whether the response pattern is following expectations, or if there is a high number of don’t know or refusal answers to particular questions. Any problems will be explored and corrected as necessary.
In Phase 1, we will screen potential respondents to ensure that the cognitive interviews are conducted with women of the targeted racial/ethnic background. We will maintain active contact with each recruiting source, and address any questions, concerns or problems that are identified as possible barriers to participation. These might include concerns about confidentiality and privacy, safety, and the justification for conducting the study. The contractor will ensure that a sufficient supply of fliers is maintained at each source, and will track the number distributed. If stepping up recruiting efforts at the participating sources does not yield respondents of the number and kind required (a mix of ages, martial status, education, language), additional sources will be identified, and submitted to the CDC and OMB for approval.
In Phase 2, the contractor will employ methods of maximizing response rates that draw upon experience with both asking sensitive questions and interviewing minority populations. These methods include developing and delivering training to the interviewers which focuses on the sensitivity of the survey topic and the particular relevance in the targeted minority populations. Specific training modules will be built into our interviewer training agenda that emphasize these issues (see Appendix 13 for list of training topics). Additional methods include the creation of effective promotional and endorsement materials which encourage participation in the study while ensuring discretion and confidentiality for all study participants. Careful selection of field staff will also ensure that the study protocol will be adhered to. Prior NORC studies that focus on sensitivity include studies of sexual behavior, for example, the recently conducted National Social Life, Health and Aging Project, Arrestee Drug Abuse Monitoring program, and the Former Prisoner Study of Sexual Violence in Prisons. Efforts in minority communities include the Making Connections Study, REACH 2010, and the Woodlawn project. The study will be staffed by interviewers and managers who are knowledgeable and sensitive to the cultural and racial makeup of the respondent populations and the survey topic, including Spanish-English bilingual interviewers. Finally, production rates by area will also be closely monitored throughout the course of the project via reports that will be shared with CDC. Any communities or racial/ethnic groups which demonstrate unacceptably low response rates will be identified for additional refusal aversion and conversion strategies. These strategies may include additional interviewer training, revised outreach materials or changes in field staffing assignments.
The two-phase design of the study will allow refinement of questions and minimization of burden by providing information on the content, functionality, and question wording prior to the start of the main survey. In addition to testing the sexual violence tool, we will test the recruitment materials (see Appendices 3 and 4) to ensure that they work effectively in Phase 2.
Stephanie Eckman., 301-634-9442, [email protected]
NORC, a national organization for research at the University of Chicago, will collect Phase 1 and Phase 2 data for CDC. The following individuals at NORC will conduct the work.
Design of the data collection: Phyllis Newton, Angela DeBello, Stephanie Eckman;
Collect the data: NORC field interviewing staff under the direction of DeBello; and analyze the data: Newton and Eckman.
1 City-style addresses are those that are suitable for in-person interviewing, i.e. they are not P.O. boxes. In our areas, the vast majority of the addresses in the USPS list are city-style.
2 The most recent birthday method is an easily implemented way of performing random respondent selection without conducting a roster of the entire household, which some respondents find intrusive.
| File Type | application/msword |
| Author | fmc7 |
| Last Modified By | Petunia L. Gissendaner |
| File Modified | 2008-04-08 |
| File Created | 2008-03-25 |