The 2011 Census Link project to process and link the 2011 Census data to the Longitudinal Study (LS) database is nearing completion. The different stages of the project are detailed below.
The 2011 Census records with an LS date of birth have been traced against the patient register held by the NHS. This enables information such as birth, death and cancer information to be linked to the traced LS members. The results of the 2011 tracing exercise show that 99 % of records were traced. This figure is comparable with previous decades and is one indicator of the high level of quality of the data the LS holds.
Inform Census Quality Assurance Process
The traced LS data for 2011 has also been used with LS data from the existing research database to carry out analysis to improve our understanding of under and over count in the 2011 Census. This analysis was one of the sources of evidence used to establish the need for, and the level of, an adjustment to the 2011 Census figures prior to their publication
Build the database
The 2011 Census Link beta test database has been built. This involved:
identifying the primary and secondary census records for members with multiple enumerations;
renaming and recoding of census variables to ensure they are consistent where possible with previous decades;
deriving LS derived variables such as distance of migration over the inter-censal period; and
splitting the data into member, non-member, multiple member, multiple non-member and X-file tables (disclosive information like day and month of birth are held in X-files which are not directly accessible to users) to match the existing database structure.
Our existing data dictionary has been updated to include the 2011 Census information. The data dictionary application we use is currently being redeveloped so that we can publish it on the web and place it in the Virtual Microdata Laboratory for easier access.
The alpha test was completed successfully at the end of April 2013. It included the following quality checks to determine whether the database was fit for purpose.
Initial checks on the integrity of the census and life events data extracts.
Census to census linkage rates.
Life events linkage rates.
The beta test is the final testing phase of the Census Link project. The beta test is a chance for experienced researchers to use the LS database for research, before we launch the LS database to the wider research community. An event was held to introduce the data available in the new research database and explain how the beta process would work. Researchers were then invited to submit research proposals following the guidance set out on the beta test web page.
Ten research projects were selected to test the beta test database which began on 7th May 2013. Details of these projects are in the beta test project document (70.7 Kb Pdf) .
The beta test period is due for completion by November 2013 and will be followed by the launch of the 2011 Census Link research database.