This guide has two main purposes:
· To explain how statistics produced in Demographic and Health Surveys (DHS) reports are defined and calculated and serve as a reference document for researchers.
· To provide an overview of the structure and use of DHS datasets.
The first version of this document (Rutstein and Rojas, 2003 [updated 2006]) had been prepared in response to feedback received from DHS data users on the need for a detailed guide to assist them in understanding DHS data sets and reproducing the statistics in DHS reports. This version updates the earlier versions to capture new variable definitions and the expanding array of data.
The first chapter of the guide deals with general topics including DHS data file structure, recode files and variable naming, sampling design, sampling weights and sample stratification, century month codes and other created variables, matching and merging datasets, adjustment factors employed in surveys in which only ever-married women are interviewed, and background variables used in tabulations.
The remaining chapters deal with the definitions of individual statistics presented in DHS reports by chapter.