AN ANALYSIS OF SAMPLING ERRORS FOR THE DEMOGRAPHIC AND HEALTH SURVEYS

Sampling errors and design effects from 48 nationally representative s urveys conducted under the Demographic and Health Surveys Program for a large number of variables concerning fertility, family planning, fer tility intentions, child health and mortality etc, are analysed for th e total sample, and for urban-rural domains, sub-national regions and various demographic and socio-economic subclasses, The effect of sampl e weights on sampling errors is examined, as well as the effect of clu stering of children in a sample of women who are the actual respondent s in DHS surveys, Modelling of design effect for geographic domains, f or subclasses that are defined in terms of individual characteristics, and for differences between subclasses is offered. At the country lev el, overall design effect (the ratio of actual to simple random sampli ng standard error) averaged over all variables and countries is around 1.5. Variation among countries is high, but less so than among variab les, Urban-rural and regional differentials in design effects are smal l, and can be attributed to the fact that similar sample designs and c luster sizes were used across those domains within each country. Desig n effects for estimates over other subclasses are smaller, and tend to wards 1.0 for small subclasses and differences, apart from the effect of sample weights which tends to persist undiminished across variables and subclasses.