I’d like to use the data (like wall and roof type) on admin2 (subcounty) level for Kenya. How can I do that?
The 2009 Kenya sample available via IPUMS International is a 10% random sample of the entire Kenyan population. Due to the sampling design, the household level sampling weight variable (HHWT) holds the same value for all households in the 2009 Kenyan sample. Therefore, the sampling weights is not necessary when calculating representative statistics, but may be useful when estimating population counts. The second-level administrative boundary is identifiable via the variable GEO2_KE2009. Each household’s wall and roof type are noted in WALL and ROOF, respectively.
To extract this data you can add the variables noted above (along with any other variables you’d like to include) and select the 2009 Kenyan sample from the samples list. Then follow the prompts in your data cart to submit your extract. You have the option to receive your data in a number of different file formats. The default is a fixed-width file with an associated command file to read in the data to Stata, SPSS, SAS, or R. More details about using the IPUMS extract system and downloading data are noted in our online video tutorials.