Marital Instability Over the Life Course
Research Design and Methodology
The original target population of the study consisted of all married individuals in households in the contiguous United States with a telephone, both spouses present, and both spouses 55 years of age or less. In 1980, telephone interviewers used random digit dialing to select a sample of households and a second random procedure to determine whether to interview the husband or wife. Seventeen percent of targeted individuals could not be reached after 20 call-backs. Of those individuals contacted, 78% gave complete interviews. The final sample consisted of 2,033 married persons (not couples). When compared with U.S. Census data, the sample was representative with respect to a variety of variables, including age, race, household size, housing tenure, presence of children, and region of the country. The sample was contacted again in 1983, 1988, 1992, 1997, with re-interview rates of 78%, 84%, 89%, and 90%, respectively.
A sample of offspring also was included as part of the 1992 and 1997 waves of data collection. To be eligible, offspring had to (a) be 19 years of age or older at the time of the interview and (b) have resided in the parental household in 1980. Eighty-seven percent of the parents provided names and telephone numbers of children, and we were able to obtain interviews with 88% of these individuals for an overall completion rate of 77%. In households where there was more than one eligible child, a random procedure was used to select the child for inclusion in the study. In households with two or more eligible children, the name and telephone number of a second child also was obtained, and the second child was interviewed when it was not possible to obtain an interview with the first. Five percent of the interviews were with offspring other than the first one selected. Overall, we interviewed 471 offspring in 1992 and an additional 220 offspring who had reached the age of 19 in 1997. We also conducted re-interviews in 1997 with 427 off the offspring first interviewed in 1992.
Wave VI replicated the 1980 national survey of married persons under the age of 55 who had telephones; re-interviewed persons who were part of a sample of married persons first interviewed in 1980 who were subsequently re-interviewed in 1983, 1988, 1992, and 1997; and interviewed a supplementary random sample of people between the ages of 38 and 75 who were married in 1980. A total of 2,960 individuals were interviewed in the cross-sectional survey. From these, 2,000 met the requirements necessary for the replicate sample, 1763 fit the definition for inclusion in the comparison sample, and 802 were included in both groups (the overlap group). These proportions were estimated using the most recent Census age and marital status data, as well as data from the NSFH cross-sectional survey, which contains more detail on marital history.
The sampling procedures for the 2000 cross-section replicated as closely as possible the sampling procedures used in the 1980 survey with changes reflecting improvements since 1980 in telephone interviewing technology and sampling methods. Both cross-sectional components of the survey were generated from a single sample of households that were interviewed with a short screening instrument. The population administered by the screener consisted of all persons living in households with a telephone in the continental United States (excluding Alaska and Hawaii, which were excluded in 1980). A random-digit-dialing sampling frame was used to generate the numbers called.
Last modified: 02/13/06 | Contact Webmaster
