zvifadzomatsenazingoni/Malawi-equity-analysis

# Socio-spatial equity analysis of access to hospitals in Malawi ## SOURCE DATA FILES **1) Dataset of ultra-poverty at the district-level (WQ1 district.csv)**: This file contains the percentage of ultra-poor households in each district. The column variable labels are: **id_1**: District identifier. **name_1**: District name. **WQ1prop**: The percentage of ultra-poor households in each district. These source data are used in ArcGIS to construct the district-level poverty map (Fig. 1a). **2) Dataset of cluster-level wealth (WQ1 data.csv)**: This file contains data on the Wealth Index for each cluster. The column variable labels are: **centroidid**: Cluster location identifier. **wealthscorecont**: The average Wealth Index for each cluster. **HHID**: The number of households in each cluster. **WQ1**: The number of households in each cluster that have a Wealth Index in the lowest quintile (i.e., ultra-poor). **WQ2**: The number of households in each cluster that have a Wealth Index in the second lowest quintile. **WQ3**: The number of households in each cluster that have a Wealth Index in the middle quintile. **WQ4**: The number of households in each cluster that have a Wealth Index in the second highest quintile. **WQ5**: The number of households in each cluster that have a Wealth Index in the highest quintile. **WQ1prop**: The percentage of households in each cluster that have a Wealth Index in the lowest quintile (they are ultra-poor). **povertyWQ4**: The number of households in each cluster that have a Wealth Index that is not in the lowest quintile (i.e., they are not ultra-poor). This is the sum of WQ2, WQ3, WQ4 and WQ5. These source data are used as an input for: (i) conducting Empirical Bayesian Kriging in ArcGIS to generate the poverty map of the percentage of ultra-poor households per km2 (Fig. 1b), and (ii) performing a hot-spot analysis of the number of ultra-poor households in SaTScan (Fig. 1c). **3) Dataset for LISA analysis (GA SecondTe …