Sampling
Sampling Procedure
The Sample selected for the 2009 National baseline Household Survey (NBHS) was based on a stratified two stage Sampled Design. The Sampling frame was based on 2008 Sudan Census preliminary count of Household by Enumeration Areas (EAs) and the Census Cartography. The Primary Sample Units (PSUs) was EAs which were Census operational segements indentified on maps, with an average of 184 households in Urban and 136 Household in Rural areas.
For the NBHS, the Census EAs were stratified by State, Urban and Rural Areas. At the second sampling stage, households were selected from the listing in each sampled EA. The Sample Size was determined for obtaining reliable estimates for key survey indicators at State level, and for Urban and Rural domains at the National level.
A sample of 44 EAs was selected at the first sampling stage for each of the Ten States of Southern Sudan, and at the second stage, 12 households were selected from the listing of each sampled EA. Therefore 528 households per State were selected which total to a sample size of 5280 households for Southern Sudan.
Given the above, only 15.2% of the households in Southern Sudan were classified as Urban, a higher first stage sampling rates was used for the Urban Stratum of each State in order to improve the precision of Uban estimates at the National level
Deviations from Sample Design
During the Survey, Derivation from Sample Design occured in Western Equatoria and Jonglei State. These were caused by insecurity in these States. In Central Equatoria, one EA was demolished which force the the Survey team to replace that EA.
Response Rate
The response rate for this Study 100 percent
Weighting
The sample was stratified by region. For reporting at State level results, sample weights were used. Variable hhweight was used to weight records at the household level.
