Probability Sample Selection Method in Household Surveys When Current Data on Regional Population is Unavailable

Availability of the perfect sampling frame only exists in developed countries, which covers a very small proportion of the world countries. On the other hand, in developing countries lists of the latest population census counts are generally used as the sampling frame for sample surveys. Therefore, in developing countries surveys which are planned for future periods long after the census date, cannot be representative of the related time period if the same census counts are utilized. Instead, population projections and data adjustment methodologies must be used to provide a representative probability selection of the updated population. This article proposes a population projection and adjustment methodology in order to establish the ideal selection probability for household surveys. The method contains the correction on the differences of the sum of strata and aggregated values. Comparative examples are also provided to clarify the proposed methodology.