Responsive Design, Weighting, andVariance Estimation in the 2006-2010 National Survey of Family Growth.

Objective-The National Survey of Family Growth (NSFG) collects data on pregnancy, childbearing, men's and women's health, and parenting from a national sample of men and women aged 15-44 in the United States. The 2006-2010 NSFG design was a significant departure from the previous periodic design, used in 1973-2002. This report shows fieldwork results and weighting, imputation, and variance estimation procedures. The report should be useful to users of the 2006-2010 public-use data file and to survey methodologists wishing to learn how the NSFG was conducted. Methods-NSFG's new design is based on an independent national probability sample of men and women aged 15-44. The University of Michigan's Institute for Social Research conducted fieldwork under a contract with the National Center for Health Statistics. Professional female interviewers conducted in-person, face-to-face interviews using laptop computers. A responsive design approach was used in planning and managing the fieldwork for NSFG to control costs and reduce nonresponse bias. Results-The 2006-2010 NSFG is based on 22,682 completed interviews-10,403 interviews with men and 12,279 with women. Interviews with men lasted an average of 52 minutes, and for women, 71 minutes. Weighted response rates were 75% among men, 78% among women, and 77% overall. Analysis of NSFG data requires the use of sampling weights and estimation of sampling errors that account for the complex sample design and estimation features of the survey. Sampling weights are provided on the data files. The rate of missing data in the survey is generally low.

Responsive survey design, demographic data collection, and models of demographic behavior.

To address declining response rates and rising data-collection costs, survey methodologists have devised new techniques for using process data ("paradata") to address nonresponse by altering the survey design dynamically during data collection. We investigate the substantive consequences of responsive survey design-tools that use paradata to improve the representative qualities of surveys and control costs. By improving representation of reluctant respondents, responsive design can change our understanding of the topic being studied. Using the National Survey of Family Growth Cycle 6, we illustrate how responsive survey design can shape both demographic estimates and models of demographic behaviors based on survey data. By juxtaposing measures from regular and responsive data collection phases, we document how special efforts to interview reluctant respondents may affect demographic estimates. Results demonstrate the potential of responsive survey design to change the quality of demographic research based on survey data.

The 2006-2010 National Survey of Family Growth: sample design and analysis of a continuous survey.

OBJECTIVE: The National Survey of Family Growth (NSFG) collects data on pregnancy, childbearing, men's and women's health, and parenting from a national sample of women and men 15-44 years of age in the United States. This report describes the sample design for the NSFG's new continuous design and the effects of that design on weighting and variance estimation procedures. A working knowledge of this information is important for researchers who wish to use the data. Two data files are being released--the first covering 2.5 years (30 months) of data collection and the second after all data have been collected. This report is being released with the first data file. A later report in this Series will include specific results of the weighting, imputation, and variance estimation. METHODS: The NSFG's new design is based on an independent, national probability sample of women and men 15-44 years of age. Fieldwork was carried out by the University of Michigan's Institute for Social Research (ISR) under a contract with the National Center for Health Statistics (NCHS). In-person, face-to-face interviews were conducted by professional female interviewers using laptop computers. RESULTS: Analysis of NSFG data requires the use of sampling weights and estimation of sampling errors that account for the complex sample design and estimation features of the survey. Sampling weights are provided on the data files. The rate of missing data in the survey is generally low. However, missing data were imputed for about 600 key variables (called "recodes") that are used for most analyses of the survey. Imputation was accomplished using a multiple regression procedure with software called IVEware, available from the University of Michigan website.

Experimental Studies of Disclosure Risk, Disclosure Harm, Topic Sensitivity, and Survey Participation.

This article extends earlier work (Couper et al. 2008) that explores how survey topic and risk of identity and attribute disclosure, along with mention of possible harms resulting from such disclosure, affect survey participation. The first study uses web-based vignettes to examine respondents' expressed willingness to participate in the hypothetical surveys described, whereas the second study uses a mail survey to examine actual participation. Results are consistent with the earlier experiments. In general, we find that under normal survey conditions, specific information about the risk of identity or attribute disclosure influences neither respondents' expressed willingness to participate in a hypothetical survey nor their actual participation in a real survey. However, when the possible harm resulting from disclosure is made explicit, the effect on response becomes significant. In addition, sensitivity of the survey topic is a consistent and strong predictor of both expressed willingness to participate and actual participation.

Planning and development of the continuous National Survey of Family Growth.

OBJECTIVES: This report describes how the continuous National Survey of Family Growth (NSFG) (begun in 2006) was designed, planned, and implemented. The NSFG is a continuous national survey of men and women 15-44 years of age designed to provide national estimates of factors affecting pregnancy and birth rates; men's and women's health; and parenting. METHODS: The survey used in-person, face-to-face interviews conducted by trained female interviewers. One person per household was interviewed from a national area probability sample. The data collection used computer-assisted personal interviewing (CAPI). Separate questionnaires were used for male and female respondents. The last section of the questionnaires used a self-administered technique called audio computer-assisted self-interviewing or ACASI. Each data collection period lasted 12 weeks-10 weeks for "Phase 1," the main data collection protocol, and 2 weeks for "Phase 2," an intensive attempt to locate and interview nonrespondents. RESULTS: Each year, about 5,000 persons were interviewed in about 33 areas, called primary sampling units (PSUs). Over a 4-year period, 110 PSUs will be used. This report gives an overview of the procedures used in the conduct of the continuous NSFG. A later report will describe response rates and other results of the data collection, but the early fieldwork has gone well.

Risk of Disclosure, Perceptions of Risk, and Concerns about Privacy and Confidentiality as Factors in Survey Participation.

This article reports on a web-based vignette experiment investigating how likely subjects would be to participate in surveys varying in topic sensitivity and risk of disclosure. A total of 3,672 participants each responded to a series of eight vignettes, along with a variety of background questions, concerns about confidentiality, trust in various institutions, and the like.Vignettes were randomly assigned to respondents, such that each respondent was exposed to four levels of disclosure risk for each level of topic sensitivity (high versus low). Half the sample was assigned to receive a confidentiality statement for all eight vignettes, while the other half received no mention of confidentiality in the vignettes. The order of presentation of vignettes was randomized for each respondent.Respondents were also asked for their subjective perceptions of risk, harm, and social as well as personal benefits for one of the eight vignettes. Adding these questions permits us to examine how objective risk information presented by the researcher relates to the subjective perception of risk by the participant, and to assess the importance of both for their willingness to participate in the surveys described.Under conditions resembling those of real surveys, objective risk information does not affect willingness to participate. On the other hand, topic sensitivity does have such effects, as do general attitudes toward privacy and survey organizations as well as subjective perceptions of risk, harm, and benefits. We discuss the limitations and implications of these findings.

National Survey of Family Growth, Cycle 6: sample design, weighting, imputation, and variance estimation.

OBJECTIVES: Cycle 6 of the National Survey of Family Growth (NSFG) was conducted by the National Center for Health Statistics in 2002 and early 2003. This report describes how the sample was designed, shows response rates for various subgroups of men and women, describes how the sample weights were computed to make national estimates possible, shows how missing data were imputed for a limited set of key variables, and describes the proper ways to estimate sampling errors from the NSFG. The report includes both nontechnical summaries for readers who need only general information and more technical detail for readers who need an in-depth understanding of these topics. METHODS: The NSFG Cycle 6 was based on an independent, national probability sample of men and women 15-44 years of age. It was the first NSFG ever to include a national sample of men 15-44 as well as a sample of women. Fieldwork was carried out by the University of Michigan's Institute for Social Research (ISR) under a contract with NCHS. In-person, face-to-face interviews were conducted by professional female interviewers using laptop computers. In all, 12,571 women and men-7,643 females and 4,928 males-were interviewed, the largest NSFG ever done. RESULTS: Analysis of NSFG Cycle 6 data requires the use of sampling weights and estimation of sampling errors that accounts for the complex sample design and estimation features of the survey. Examples of how to use several available software packages that incorporate complex design features in estimation, such as SAS, SUDAAN, and STATA, are presented.

Plan and operation of Cycle 6 of the National Survey of Family Growth.

OBJECTIVES: This report describes how Cycle 6 of the National Survey of Family Growth (NSFG) was designed, planned, and implemented. The NSFG is a national survey of women and men 15-44 years of age designed to provide national estimates of factors affecting pregnancy and birth rates; men's and women's health; and parenting. Cycle 6, conducted in 2002, was the first time the NSFG included a sample of males. METHODS: The survey used in-person, face-to-face interviews conducted by trained female interviewers. One person per household was interviewed from a national area probability sample in about 120 sample areas, with oversamples of teenagers, African Americans, and Hispanics. The data collection used computer-assisted personal interviewing (CAPI). Separate questionnaires were used for female and male respondents. The last section of the questionnaires used a technique called audio computer-assisted self-interviewing (ACASI). In order to control costs and nonresponse errors, survey managers statistically analyzed results from interviewers' visits to sampled households each day, and used those results to allocate interviewer labor and other resources more efficiently. This management improved response rates and made the sample more representative. RESULTS: Over 12,500 interviews were completed, about 7,600 with females and about 4,900 with males. The response rate was about 80 percent for females and about 78 percent for males. The survey procedures were adapted during the fieldwork to achieve the desired response rates and to control costs.