Sudan-El Fasher Durable Solutions Profiling Exercise

Methodology & limitations: Sample-based Household Survey: A sample of 3,000 households was collected using a stratified cluster sampling approach. The sample was divided into four strata: IDPs in Abu Shouk camp, IDPs in El Salam camp, non-displaced people living in peri-urban El Fasher and non-displaced inhabitants of urban El Fasher. The four strata were divided into clusters of similar population size, based on a grid developed on a map of the areas.Clusters were selected from each stratum with a uniform probability of selection. The sampling was conducted in this manner, because there was no reliable population data available which would have made sampling probabilities to size possible. All the households in the selected enumeration areas were listed, while 12 were selected for interviews in each cluster in simple random draws. The listing exercise resulted in a significantly lower number of clusters than was planned.To compensate for the lower number of clusters, some of the listed clusters were selected for oversampling.To draw the sample, the listing data was restricted to IDP households in IDP clusters and non-displaced households in the El Fasher clusters. The World Bank calculated the sample and applied the weights to the collected data. IOM collected the data during the enumeration and the households survey with the support of community representatives

Limitations: (1) Sampling weights were calculated and applied to each enumeration area depending on its size. The resulting weights could not be tested against population data due to the lack of up-to-date and reliable information. However, for the camps, the population sizes suggested by the weights are close to the latest population estimates by the International Organization of Migration (IOM); (2) The sample in the two El Fasher strata (urban and peri-urban) only included non-displaced households. Therefore, it should be kept in mind that the results for these two strata do not represent the situation for all residents living in the peri-urban and urban neighbourhoods, given that displaced households in these strata were not included in the sample. Specifically, the enumeration indicated that approximately 28% of peri-urban and 21% of urban households are IDPs. The comparative analysis thus focuses on the differences between the population groups by displacement status and cannot be used to compare the areas; (3) The household survey questionnaire administered was very comprehensive and posed challenges in the course of the data collection linked to time consuming interview processes. Enumerators reported survey fatigue amongst respondents.At times, this can be assumed to have impacted the quality of answers provided; (4) The urban analysis looking at the access to services and land captured primarily distance to services but was not able to capture adequately the capacity and quality of the service provision. The ‘reachability maps’ created on the distance to services should be treated carefully, as they are not taking into account local topographic features and context specific circumstances. A deeper reachability analysis should, in any case, be validated on the ground