Taille des fratries et taille des familles dans les données d'enquêtes utilisées pour estimer la mortalité

Survey data on sibling survival provide a crucial source of information for estimating adult mortality in countries where vital records are incomplete. This article assesses the quality of these data by comparing sibship sizes reported in Demographic and Health Surveys with women?s mean number of children ever born in the previous generation. This comparison, conducted at aggregate level, suggests that a high proportion of siblings are omitted, since the sibship sizes are 15 % lower, on average, than would be expected on the basis of number of children ever born. Such omissions are more frequent in sub-Saharan Africa than in other developing regions, and their extent increases slightly with the respondents? age. Adult mortality deduced from these data is not necessarily underestimated, however, since omissions appear to mainly concern siblings who died in childhood.