Assessment of the quality of data on age at first sexual intercourse, age at first marriage and age at first birth in the Demographic and Health Surveys

An analysis of the quality of the data on age at 1st union, age at 1st birth, and age at 1st sexual intercourse collected in Demographic and Health Surveys (DHS) from 22 developing countries revealed considerable variation in the accuracy and completeness of the information collected. Accurate data on these parameters is especially important since they determine the total number of years of exposure to the risk of pregnancy and identify when the period of exposure began. Completeness of reporting on data of 1st union was greatest in Latin America and weakest in West Africa, but, at least 75% of women in all 22 surveys provided either an age or date for this event. There was evidence of significant heaping on calendar years and durations since 1st marriage ending in 0 or 5. Also evident was an appreciable amount of forward displacement of the date of 1st union in the oldest age groups in most surveys. Finally, the DHS surveys documented a far higher percentage of women married at ages 15-24 years than other surveys--a discrepancy that is considered to reflect a higher level of coverage in the DHS. Data on age at 1st exhibited similar characteristics. There was evidence in over half the surveys of forward displacement of the date of 1st birth, or omission of early births, among older women, especially those in sub-Saharan Africa. Only 1-2% of respondents refused to provide data on age at 1st sexual intercourse. In the African countries, 12-67% of women experienced intercourse 1 or more years prior to 1st union, indicating that reliance on date of 1st union is insufficient to capture all sexual exposure. It is recommended that data collection procedures be modified slightly to reduce the extent that interviewers calculate the age at 1st intercourse based on the data of 1st marriage.