Primary data will be collected by means of a household-level survey designed with the participation of the humanitarian clusters in Somalia. Cluster leads are asked to outline information gaps and the type of data required to inform their strategic plans. Key indicators are developed by REACH with the substantive input of participating partners, and subsequently validated by the clusters. REACH will draft the household survey tool through an iterative consultation process with cluster partners and OCHA and is aligned, as much as possible, with the Joint Inter-Sectoral Analysis Framework (JIAF) which will serve as a common and structured method for assessing the severity of needs across different clusters.
The assessment will use stratified cluster sampling at the district level using settlements as the clusters and households as the unit of measurement. For some districts, 2-stage stratified random sampling will be used instead of stratified cluster sampling for large urban centres, if it proves to be more efficient and logistically feasible for data collection. The sample will be stratified by population group, disaggregated by non-displaced communities, and IDP settlements; the sample will be further stratified by district to ensure coverage and comparison across the entire country (with the exception of inaccessible areas). In the case of cluster sampling, the minimum cluster size will be set to 6 households. The sample size will be adjusted for the design effect and will enable generalisation of the results to each of the two population strata in each district, with a 90% confidence level and a 10% margin of error.