Sampling
Sampling Procedure
The sampling unit was a cluster of twenty-five households and the clusters were selected in a 3-stage sampling process. As no recent census information or comparable framework exists for southern Sudan the sampling frame had to be created from scratch. The social hierarchy existent in southern Sudan was therefore used, whereby a listing of 'Executive Chiefs' or their equivalents, together with the number of Sub-Chiefs falling under each, was created. Representatives of town populations, fishing communities and large groups of displaced persons were equated as equivalent to an Executive Chief and therefore included in the sampling frame as such. Sample frame information for five of the twenty-eight counties/states could not be obtained prior to the survey, for reasons of security, access, or data availability, and therefore these areas had to be ruled out of scope for the study.
All 25 households within the selected sample segment or cluster were interviewed in turn this approach was adopted to minimise the time wasted moving between households and to avoid potential respondent data registration problems, when adjoining households are not covered;
- A 'household' was defined as those individuals regularly coming together to eat from the same cooking pot;
Response Rate
Out of the 5000 households originally targeted by the survey, a total of 4300 were successfully interviewed. Thus the overall survey response rate was 86 percent (100 x number interviewed /number selected into the sample). It should be noted, however, that where households could be reached by survey staff, the response rate was 100% and therefore the interviewed household shortfall with the target was due to the last minute withdrawal of certain sample locations from the survey for reasons of security and/or accessibility.
