There is no sampling or methodological reason why one could not use women ages 15-19 for calculating national-level estimates of the percentage of women married before age 15. However, the calculation of the percentage of women married before age 18 cannot include any women below age 18, so analysts often choose to calculate that estimate for women age 20-24, especially because there may not be enough women age 18-19 to calculate the indicator only for that age group.

The estimation of the median age at marriage (that is, the age by which half of any cohort has married) cannot be calculated for any age group for which less than 50 percent of the respondents were married for the first time before reaching the beginning of the age group. That is why age 20 is often used as the lowest age for calculating the median age at marriage among women and age 25 is often used as the lowest age for calculating the median age at marriage for men.

The standard DHS tabulation for the percentage of women or men married by exact age 15 starts with the age group 15-19 and the tabulation for the percentage of women or men married by exact age 18 starts with the age group 20-24. The standard age at marriage indicators for UNICEF Multiple Indicator Cluster Surveys (MICS) are the percentage of women age 15-49 years who were first married or in union before age 15 and the percentage of men age 20-49 who were first married or in union before age 18.

Since the DHS and MICS survey data sets are publicly available, data users may choose the age groups for their analysis of age at marriage that are consistent with their own research needs.