Submitted by the NSFG team
The National Survey of Family Growth (NSFG) collects information on factors that help explain birth rates, such as contraception, infertility, sexual activity, cohabitation, and marriage. The NSFG was first conducted in 1973 and was a periodic survey until the transition to continuous interviewing in 2006. Originally, the NSFG only included a sample of women, but men were added beginning in 2002 and the survey is now a nationally representative sample of women and men ages 15-44. The latest data release includes interviews conducted from 2006-2010 with 12,279 women and 10,403 men. The public use data files are available for download from our website along with more information about the survey including the history, codebooks and User’s Guides for each survey, and key statistics on factors that influence family formation, growth, and dissolution based on the most recent survey.
It is not hard to think about ways that the community where someone lives could influence some of the topics covered in the NSFG. For example, you can probably think about how the characteristics of the larger community might influence who or when a person marries. When researchers want to study how the context or community where someone lives potentially influences their fertility and family life they can use the NSFG contextual data files. The RDC hosts the contextual files for the 1995, 2002, and 2006-2010 NSFGs.
For the 1995, 2002, and 2006-2010 data files, there are two contextual files for respondents; these correspond to the respondent’s address at two points in time, the date of interview and at the time of the most recent Census. The contextual files include variables such as the total population, the proportion of men and women over 15 who have never been married, or the proportion of families with incomes below the poverty level measured at the state, county, tract, block group, and block level.
The codebook for the 2006-2010 contextual file is available online. Browsing through the codebook shows the wide variety of variables available to users of the RDC. There is a wealth of information available at the county level including Hispanic origin and racial composition, employment and earnings, birth rates, education and school enrollment, migration patterns, crime, voting behavior, housing units, rurality/urbanity, marital status, healthcare providers and need, STD rates, and family planning providers.
One NCHS publication that may help researchers get started with using NSFG contextual data is the report, “Community Environment and Women’s Health Outcomes: Contextual Data” by W.D. Mosher, L.P. Deang, and M.D. Bramlett. This report uses the contextual data for the 1995 NSFG to look at how community characteristics are associated with childbearing, marital status, contraceptive use, breast-feeding and other outcomes. It also offers detailed information on understanding the levels of data and conducting a multi-level analysis using SAS.
Another example of the use of the NSFG contextual data, by researchers outside of NCHS, is Magnusson, BM, Sabik, L, Chapman, DA, Masho, SW, Lafata, JE, Bradley, CJ and KL Lapane. In their article “Contraceptive Insurance Mandates and Consistent Contraceptive Use among Privately Insured Women”1 they incorporated contextual factors, and also requested that the RDC link data from the Guttmacher Institute on state-level insurance mandates with the public use data file. They took full advantage of the RDC by not only incorporating the contextual data in their analyses, but also requesting that additional data were added to the files by linking up information at the state level. By using the RDC not only were they able to answer a research question on how insurance mandates at the state level influence individual contraceptive use, they were able to account for other characteristics of the counties (such as education, income, and crime) in their models.
1 Magnusson, BM, Sabik, L, Chapman, DA, Masho, SW, Lafata, JE, Bradley, CJ and KL Lapane. “Contraceptive Insurance Mandates and Consistent Contraceptive Use among Privately Insured Women.” Medical Care 50(7):562-568. 2012.
Weigh In: Have you used NSFG data for your research?