Abstract
OBJECTIVE: To assess the quality and consistency of reported age patterns of female genital mutilation in self- and proxy-reported survey data. METHODS: We used 10 Demographic and Health Surveys (DHS) from 2005 to 2023 in Senegal. These surveys contained information on female genital mutilation status and age at experiencing this practice for women who reported data on themselves and daughters for whom data were reported by their mothers. We assessed data quality by completeness of information on age at female genital mutilation in a logistic regression analysis. We compared the occurrence of age heaping across DHS and individual survey characteristics such as education, age cohort and completeness of date of birth reporting. We estimated the median age at female genital mutilation of daughters and women to assess the consequences of differences in data quality for the interpretation of survey data on this practice. FINDINGS: Self-reported data were more prone to incomplete reporting of age at female genital mutilation and age heaping than proxy-reported data. These findings held true across individual survey characteristics and different DHS. The estimates for median age at female genital mutilation were susceptible to differences in data quality of age at female genital mutilation of daughters and women. CONCLUSION: Self-reported data on age at female genital mutilation are of lower quality than proxy-reported data. These differences potentially distort trend estimates of age at female genital mutilation. Caution is needed when combining self- and proxy-reported survey data on female genital mutilation.