National Fertility Survey, 1965, 1970, and 1975
National Fertility Survey, 1965
Basic Information
- Author: Norman B. Ryder and Charles F. Westoff
- Conducted by: National Analysts, Inc., Philadelphia
- Funded by: National Institute of Child Health and Human Development.
- Data Prepared by: Center for Demography and Ecology, University of Wisconsin, under the direction of Larry Bumpass
- Universe: Currently married women, born after July 1, 1910, living with their husbands, able to participate in an English language interview. (Maximum age at interview is 55.) The sample for women aged 45-55 is half the size of that for the younger women, while black women were double-sampled. The number of black women interviewed was larger than anticipated; see record weights, below.
- Date of Survey: Midpoint November 1965
- Coverage: Coterminous United States (48 states).
- Sample Size: 5617 completed interviews.
- Record Weights: To adjust for differential sampling
rates, the following weights must be used:
- White and other women < age 45: 1.0
- White and other women aged 45-55: 2.0
- Black women < age 45: 0.365863
- Black women aged 45-55: 0.6070
- Access: Public
Additional Information
Sample
6397 eligible women were selected for interview, of whom 5617 were actually interviewed, yielding a response rate of 88%. 8% of the eligible women refused to be interviewed; the remainder were either repeatedly not at home or not available for other reasons.
Characteristics of the sampled women were compared with those of women in the same age groups from the March, 1965, and March, 1966, Current Population Surveys. Among black women aged 14-44, there was a considerable discrepancy between the proportion married and living with husband in the sample (12.9 %) and the CPS surveys (9.4 %). The proportion of black women under age 25 was also higher in the sample than in the CPS survey. The sample also somewhat underrepresents women with no children and may overrepresent women in the higher education categories, although this comparison can only be estimated.
Field Work
The interviews were conducted by a staff of trained women from National Analysts, Inc. Two days of additional training were devoted to this questionnaire. The average length of interview was one hour and fifteen minutes.
Data Collected
Women were asked questions about fertility and contraception, including contraceptive use and pregnancy histories, opinions on childbearing and childrearing, desired family size, future childbearing intentions and expectation of further children. Questions about coital frequency at the time of interview were asked. Marital history and some labor force participation history were recorded. Background information such as education, income, religion, social characteristics, and place of residence was also collected. The live birth history contains questions about length of breastfeeding and survival staus of the child. The pregnancy history contains questions about methods of contraception used and childbearing intentions at the time of conception.
Other References
Detailed results of the survey may be found in Reproduction in the United States, 1965, by Norman B. Ryder and Charles F. Westoff (Princeton, NJ, Princeton University Press, 1971), from which this abstract was prepared.
Structure of the Data
The data are in a rectangular file, with one record for each respondent. Pregnancy histories and live birth histories are embedded in the record in repeating fields, with "inapplicable" codes assigned to events that didn't occur. These codes vary from variable to variable and must be checked in the codebook. The record weights are not coded in the data and must be defined based on race and age. (See Record Weights above under Basic Information).
Location of Data and Documentation
You can click here for a file list.
National Fertility Survey, 1970
Basic Information
- Author: Charles F. Westoff and Norman B. Ryder
- Conducted by: Institute for Survey Research, Temple University
- Funded by: National Institute of Child Health and Human Development.
- Universe: Ever-married women, born after July 1, 1925. For black women, a sampling ratio double that for white and other women was used.
- Date of Survey: Midpoint December 31, 1970.
- Coverage: Coterminous United States (48 states).
- Sample Size: 6752 completed interviews.
- Record Weights: To adjust for differential sampling
rates, the following weights must be used:
- White and other women: 1.0
- Currently married black women: 0.579
- Formerly-married black women: 0.432
- Access: Public
Additional Information
Sample
The sample listed 38,839 households; for 6% of these, a listing of household members could not be obtained due to refusal or inability to find someone at home. In the households for which member listings were obtained, 7970 eligible women were identified, of whom 6752 completed interviews. This is a response rate of 85% among identified eligible women. Considering the number of women expected if listings had been obtained for all households, the response rate at the household level is estimated to be 80%.
Based on comparisons with the 1970 census, the sample appears to have too few women living in central cities. However, this is at least partly due to reclassification of central cities in the 1970 census. The deficit extends to all racial, marital status and age groups. Age distributions match the Census very closely, but the NFS shows more women in the higher education categories than the Census does. The proportion of ever-married women who are currently married is lower in the NFS than in the Census. The NFS also shows fewer women with no children than the Census does.
Field Work
Trained women on the staff of the Institute for Survey Research conducted the interviews, after receiving two days of additional training. Followup letters and telephone calls to respondents were used to validate the interviews. The average length of the interview was a little over an hour.
Data Collected
Women were asked questions about fertility and contraception, including contraceptive use and pregnancy histories, opinions on childbearing and childrearing, desire for more children, future childbearing intentions, expectation of more children, etc. Questions about coital frequency at the time of interview were asked. The respondent's marital history and employment history were recorded. Background information such as education, income, religion, social characteristics, and place of residence was also collected, as well as attitudes to abortion, sterilization, and the status and role of women. The live birth history contains questions about length of breastfeeding and survival status of the child. The pregnancy history contains questions about methods of contraception used and childbearing intentions at the time of conception.
Other References
Detailed results of the survey may be found in The Contraceptive Revolution, by Charles F. Westoff and Norman B. Ryder. (Princeton, NJ, Princeton University Press, 1971), from which this abstract was prepared.
Structure of the Data
The data are in a rectangular file, with one record for each respondent. Pregnancy histories and live birth histories are imbedded in the record in repeating fields, with "inapplicable" codes assigned to events that didn't occur. The record weights are not coded in the data and must be defined based on race and marital status. (See Record Weights above under Basic Information.
Location of Data and Documentation
You can click here for a file list.
National Fertility Survey, 1975
Basic Information
- Author: Charles F. Westoff and Norman B. Ryder
- Conducted by: Institute for Survey Research, Temple University
- Funded by: National Institute of Child Health and Human Development.
- Data Prepared by:Lois Paul, Office of Population Research, Princeton University
- Universe: Reinterview of women interviewed in the 1970 NFS who met the following
criteria:
- White
- Continuously married, in first marriage
- Married less than 25 years
- Married before age 25
- Husband married only once
- Date of Survey: Late 1975 and early 1976.
- Coverage: Coterminous United States (48 states).
- Sample Size: 2361 completed reinterviews; 1042 new interviews.
- Record Weights: None.
- Access: Public
Additional Information
Sample
Of the 6752 women in the 1970 NFS, 3247 were chosen for reinterview. Of these, 420 proved to be ineligible in the interval, because of marital disruption or incorrect data in the original interview. Of the remaining 2827, 2361 completed the reinterview. 353 women refused to be reinterviewed and the rest could not be located or were otherwise unavailable. The refusal rate varied among subgroups of the population. Catholic women, women who expected no more children, women with longer marriage durations, and women in the lower education categories were more likely to refuse reinterview. Thus the sample cannot be considered representative of the original target population.
Field Work
The interviews were conducted by trained staff members of the Institute for Survey Research.
Data Collected
Women were asked questions about fertility and contraception, including contraceptive use and pregnancy and abortion histories, opinions on childbearing and childrearing, and future childbearing intentions. A five-year calendar of contraceptive use is included. Questions about coital frequency at the time of interview were asked. Background information such as education, income, religion, social characteristics, and place of residence was also collected, as well as attitudes about abortion, sex preferences, and the status and role of women.
The live birth history contains questions about breastfeeding. The pregnancy history contains questions about methods of contraception used and childbearing intentions at the time of conception.
Other References
- N.B. Ryder, "Consistency of Reporting Fertility Planning Status", Studies in Family Planning, Vol. 10(4), April 1979.
- C.F. Westoff and E.L. Jones, "Patterns of Aggregate and Individual Changes in Contraceptive Practice", Vital and Health Statistics, Series 3, No. 17, June 1979.
Structure of the Data
The data were originally on punched cards with 31 cards ("decks") per woman, representing 20 card types. These decks have since been combined in a single record, 2480 columns long, for each woman. However, the codebook has not been changed to reflect the present record structure. The position of a variable within the new record must be calculated based on the deck and column specifications in the codebook. (Position = (deck-1)*80 + column). The codebook also shows dates as years and months, while on the tape they have been converted to "century months". Please refer to the codebook for complete information on date coding.
Some questions were asked only of reinterviewed women, and some were asked only of women being interviewed for the first time. The codebook indicates which questions are inapplicable for each category. Women who were reinterviewed have case numbers in the range 0001 to 6999. Presumably, these match the case numbers in the 1970 NFS.
Location of Data and Documentation
You can click here for a file list.
