1. We excluded the small number of sample members younger than age 15 in 1999, so the sample is comparable to data from the National Survey of Family Growth, which does not survey people younger than 15.
2. Our analysis of data from the 2002 National Survey of Family Growth suggests that dropouts have somewhat less positive views of marriage than do teens who have remained in school. Therefore, the attitudes toward marriage of all teens (including dropouts) are somewhat less positive than the attitudes toward marriage of high school students (the population that can be tracked with MTF data). However, it is unclear how these differences in attitudes between dropouts and enrolled teens would affect trends in attitudes observed in MTF data.
3. Our NSFG analysis sample includes 356 African Americans and 369 Hispanics. Other subgroups used in our analysis of NSFG data are substantially larger than these subgroups, with the exception of our subgroups based on family income. The sample sizes for these subgroups are 843 for those with family income below 200 percent of poverty, 622 for those with family incomes from 200 percent to 400 percent of poverty, and 344 for those with family incomes more than 400 percent of poverty.
4. YRBS data from 2007 were not yet available at the time we conducted our analysis.
5. Teens who have dropped out of school are more likely to be sexually active than are teens who have remained in school. Therefore, the levels of sexual activity based on YRBS data reported in Chapter II would be somewhat higher if high school dropouts could be included. It is unclear, however, how the exclusion of dropouts affects the recent trends in teenage sexual activity observed in YRBS data.