IFSS_AGEFSTSX: Age at first sexual intercourse (with a man)

View as PDF | Download Stata source code | Download EXCEL File

Summary for Variable Group: Sexual history

This variable group provides information on the respondent's sexual experience including whether she has ever had sex (and if not, why not), century month of first sex, age at first sex, the age of her first sexual partner at first sex, her relationship to him, whether she is still having sex with him, the date she last had sex with him, whether she has had sex after her first period (menarche), and century month of and age at her first sex after menarche. It also includes information on the number of male sexual partners she has had in her lifetime, in the last 12 months, and prior to marriage. Additional variables capture her experience with sex education, including whether she ever talked with her parents about how pregnancy occurs, birth control, or sexually transmitted diseases, as well as whether (and in what grade) she received formal education on how pregnancy occurs, how to say no to sex, and STDs.


Age at first sexual intercourse (with a man)


2002 NSFG
1995 NSFG
1988 NSFG
1982 NSFG

Source Variables

Source Questions

C_6 C-6 1988 NSFG
C2A C-2a 1982 NSFG
C2 C-2 1982 NSFG


IFSS_AGEFSTSX specifies the respondent's age at first sexual intercourse with a male partner.

General Comparability Notes

Summary: This variable is mostly comparable across four surveys, with some differences in whether "with a man" was included in the question's wording and whether respondents reported their answers by date of first sexual intercourse or by their age at first sexual intercourse.

The respondent universe differs across surveys. In the 2002 National Survey of Family Growth, the 1995 NSFG, the 1988 NSFG, and the 1982 NSFG, this variable applies to respondents who reported ever having sexual intercourse. The respondent universe in the 1995 NSFG additionally excludes respondents who refused to report the date of her first sexual intercourse. The respondent universe for the 1982 NSFG excludes respondents who have not reached menarche. These differences in respondent universe are not likely to substantively affect analysis.

The question wording is largely similar across surveys, though the structure differs substantially. In the 2002 NSFG and the 1995 NSFG, respondents were asked: "That very first time that you had sexual intercourse with a man, how old were you?" and "Thinking about the very 1st time in your life that you had sexual intercourse with a man, how old were you?" respectively. In the 1988 NSFG, respondents were asked "When did you have sexual intercourse for the first time--what month and year was that?" Respondents who did not know the date of first intercourse were then asked a follow-up question about their age at first intercourse, but the responses are not sufficiently comparable to harmonize here, and are instead harmonized in IFSS_DKAGEFSTSX. In the 1982 NSFG, respondents were asked: "We are talking with women about the methods of birth control they have used, if any, since they began having sexual intercourse. Thinking back, after your first menstrual period, when did you have sexual intercourse for the first time--what month and year was that?" In the 1982 NSFG, if no answer was provided, respondents were then asked a follow-up question: "How old were you at that time?" We use reported age from the follow-up question when it is available, and compute age from the century month of first intercourse when it is not. While the 1988 and 1982 surveys do not specify "with a man," it is unlikely that this change in wording will substantively affect analysis, given the questions' placement in a grouping of fertility- and contraceptive-related questions.

There are slight differences in response format across surveys. In the 2002 NSFG and the 1995 NSFG, respondents provided an age at which they first had sexual intercourse, which has been coded in the harmonized variable accordingly. In both the 2002 and the 1995 surveys, respondents could also answer "Don't know" or "Refused" or "Not ascertained." In the 1988 NSFG and the 1982 NSFG, respondents' age at first sexual intercourse is computed from IFSS_CMFSTSEX (century month of first sex) and IFSS_CMBIRTH (century month of birth). . The follow-up question about age at first intercourse in 1982 is already incorporated into the harmonized variable IFSS_CMFSTSEX, with a random month imputed within the appropriate 12 month window, so the computation yields values identical to the reported values for those respondents. "Refused" and "Not ascertained" responses are imputed; "Don't know" responses are not imputed.

It is worth noting that in the 1995 NSFG and the 2002 NSFG, it was possible to compute age at first sex from IFSS_CMFSTSEX and IFSS_CMBIRTH for the small subset of respondents who provided century month of but not age at first sex. However, the two methods do not produce consistent ages for the majority of respondents who provided both. As a result, the information provided on age is harmonized in this variable and the information provided on date of first sex is harmonized in IFSS_CMFSTSEX, and the two sets of information are not combined in either for 1995 and 2002. Users who wish to combine them are advised to do so with caution.

IFSS_DKAGEFSX harmonizes responses to don't know follow-up questions from the 2002, 1995 and 1988 NSFGs.

See Also:

Imputation Flag Exists

Imputation Flag Values
-1 Illogical value
0 Not imputed
1 Imputed in original study
2 Logically imputed
3 Month imputed (Original file edit)
4 Age known, month imputed
5 Originally imputed as June, recoded to random month
6 Month imputed
7 Multiple regression imputed: Not ascertained
8 Multiple regression imputed: Refused
9 Multiple regression imputed: Don't know
10 Midpoint of a given range
11 Answer comes from a follow-up question
12 Midpoint of a range given as a follow-up
13 Answer is rounded up
14 Answer is rounded down
15 Reported months and weeks, weeks<=4
16 Reported months=trunc(weeks/4.35)
17 Reported months=weeks & weeks>4
18 Reported months!=trunc(weeks/4.35), months!=weeks, weeks>4
19 Months coded NA/RF/DK
20 Weeks coded NA/RF/DK
21 Impossible to determine which method used last

Return to variable list

Inter-university Consortium for Political and Social Research       Inter-university Consortium for Political and Social Research       Population Studies Center
© 2013 The Regents of the University of Michigan