This technical note will discuss the processes involved in creating a time-series dataset for the Health Survey for England (HSE) from 1998 to 2011. This dataset has been used to explore the changing structure and composition of England’s society over time, and then to consider what implications this has for inequalities in health. The HSE, an annual nationwide household survey which is geographically, socially and demographically representative of the England’s residential population is an ideal data source for analyses such as this. Despite the richness of the HSE a lack of variable consistency between survey years hampers analysis of change over time. To successfully explore change, variables of interest must be either derived or recoded before a new, harmonised dataset can be created for analysis. This report will outline the methods required to extract variables of interest and then harmonise inconsistent variables. Some discussion will also be presented on the main statistical methods used within the final analysis as this influences some of the choices made in data preparation.
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)