anticipated birth and a cherished child. To emphasize this critical distinction, this report does not refer to unwanted births, for example, but rather to births resulting from unwanted pregnancies or conceptions; similar precision is used with the term mistimed.
The distinctions that the NSFG draws between intended and unintended and between mistimed and unwanted carry important implications. For example, unintended pregnancies are far more likely to end in abortion than intended pregnancies, and births resulting from unwanted conceptions appear to carry particular risks for both mother and child. In Chapter 3, these consequences of unintended pregnancy are discussed in detail.
It is also important to note that knowledge about pregnancy intentions derived from the NSFG comes entirely from women. Bumpass (1994) has reported data about partners' preferences, as provided by female respondents in the 1988 National Survey of Families and Households (see also Williams, 1994), noting that in more than one-fourth of the cases in which a woman described a birth as resulting from an unintended pregnancy, she reported that her spouse or partner had either wanted the birth at that time or was indifferent to the timing. Such disagreement may be a factor in the occurrence of pregnancies that women do not intend; this issue of partner interaction is taken up in more detail in Chapter 6.
It is also apparent that the NSFG survey questions, and similar ones used in other surveys, are often not able to measure the complicated mix of feelings that can surround pregnancy, as described by recent ethnographic research in particular (Musick, 1993). It is not uncommon to learn from a pregnant woman, for example, that she both did not "intend" to become pregnant but also was not using contraception, or that perhaps she wanted to be pregnant but was less enthusiastic about having a child (Luker, 1975). In Chapter 6, these complexities are explored in detail.
Other caveats with regard to the NSFG measures of pregnancy intention have been raised. For example, the questions about intendedness are retrospective, asking about all pregnancies in the preceding 5 years, which may make the answers offered subject to distortion and recall error. The questions do not distinguish how many months or years pregnancy timing was off, nor do they reflect the intensity of the woman's feelings about the timing of conception. In commenting on data sources other than the NSFG, Chapter 3 and Appendix G also address these generic problems in measuring the intention status of a given pregnancy, noting in particular that the difference between an intended and an unintended conception is often more complicated than these terms imply. Well aware of these nuances, the 1995 NSFG cycle will use an elaborate approach to establishing the intendedness of a given conception (see Appendix C), recognizing that the concept of intended versus unintended is more a continuum than an either/or matter. Consistent with the new approach being taken by the NSFG,