Data quality assessment of birth history data in Turkey demographic and health surveys

Yadigar Coskun, Hacettepe University

Surveys, carry their problems with themselves depending on their sample design, methodology, questionnaire design, wording of the questions, interviewers’ training period and content etc. A researcher can face with sampling and non-sampling errors. Assessment of the data quality of the birth history and data used for direct estimation of infant and child mortality is very important. DHS is the programme carried out only in developing countries and most of the developing countries have known to be famous with having high fertility and infant and child mortality rates. Therefore the quality of the data of these issues are more important than any other sections in terms of having reliable estimations which is valuable for policy makers to monitor the future programmes. With this study, the data qualities of 1993, 1998 and 2003 Turkey Demographic and Health Survey data sets are going to be discussed on focusing on the quality of data using estimation on demographic rates, impacts of the quality of the data on demographic rates and mechanisms that deal with the data quality. The quality of the data used for the direct estimation of infant and child mortality (IMR and CMR) and fertility is the primary concern of this study. Main focus is on the individual questionnaire and especially on the maternity history section. In DHS surveys, a birth history section on ever-married woman questionnaires is used by which information about all the live births of the woman were purposed to be obtained.