Are Estimated Age Data from Sample Surveys of Poor Quality? Interrogating Interviewer-Respondent Conversations

Emmanuel Souza, University of Malawi

Due to incomplete civil registration records, many low- and middle-income countries often rely on sample surveys to obtain current demographic information. However, due to lower literacy levels than in developed countries, the retrospective data obtained from such surveys may be of questionable quality. For instance, the accuracy of age-related information heavily relies on respondents’ accurate knowledge and recollection of such information. This paper employs a mixed-method approach to appraise age-related information obtained from a mobile phone survey, examining its quality and investigating the process through which such data is obtained. The findings indicate that age-related data obtained through respondent or enumerator estimates involves a significant amount of guesswork and they are very erratic. Specifically, estimated age data are marked by substantial noise, leading to notable discrepancies in summary statistics derived from such data when compared to correctly reported age data. The implications of these findings are discussed.