STRATA (V022) defines the pairings or groupings of primary sampling units used in the calculation of sampling errors when using the Taylor series expansion method.
In two-stage sampling, The DHS Program may first group small geographic areas, such as enumeration areas from the last census, into broad strata defined jointly by region and urban versus rural areas within a region. In the second stage of sampling, a subset of geographic areas (sample clusters) are selected as the sites for interviewing from within defined strata.
STRATA is not the same as DOMAIN (V023). (Domains represent statistically representative areas for which valid summary statistics can be calculated for a survey, such as provinces and national urban versus rural combined areas, as described in the survey's final report.) The DHS Program recommends using STRATA along with the variable PSU (V021) to account for the impact of the sample design clustering on the estimates of variance and standard errors.
Users are strongly encouraged to review the FAQ section provided by The DHS Program on specifying stratification and clustering here [URL omitted from DDI.], using Stata or SPSS to account for sample design.