Forgotten marriages? Measuring the reliability of retrospective marriage histories

Sophia Chae, University of Pennsylvania

Lack of suitable data may explain the paucity of studies on marital instability in Sub-Saharan Africa. Ideally, prospective panel data would be used to capture changes in marital status over time. Their scarce availability, however, makes it necessary to use retrospective marriage histories collected in cross-sectional surveys. While researchers typically acknowledge the problems associated with marriage histories, such as recall error, it is unclear to what extent they may affect results. The validity of marriage histories would ideally be measured by comparing marriage histories against public records; however, this is not feasible in this context. An alternative solution is to test their reliability by comparing marriage histories of the same respondent from at least two different points in time. In this paper, I will examine the reliability of marriage histories by using data from two waves of the Malawi Longitudinal Study of Families and Health (MLSFH).