There are occasions where an organisation has an information strategy issue, but the issue does not justify the expense of a
dedicated research project. MGL has used a shared cost model to allow a number of organisations to club together in order
to fund a piece of research that none could justify individually, but which provides excellent value for money and results
when carried out collectively.
JUG-T went through two phases to help organisations that use address referenced data, or generate addresses, to
understand how reliably address referenced service data can be used to produce area based statistics.
In the first phase address data for a large part of Northern England, stretching from the east side of Warrington to
the west side of Sheffield, was tested against administrati ve geographies, statistical geographies and ‘block’ geographies.
It was demonstrated that the present set of address products and associated look-up tables, was error prone, but that many
of the errors were due to systematic differences between data sets; such as the use of alternative veriosn of the street name.
In the second phase of the project data for the City of Manchester was examined in order to identify the spatial and temporal
shortcomings of current address bases. Have a look at MGL's
Addressing Interactive Testbed (AddressingIT) for more information.
Some of the methods and results produced in the two phases of JUG-T were then used to resolve anomalies
identified in the results of the 2001 Census and to carry out
robust sample based and full audits of major national address data sets.
Clients for JUG-T included: Manchester City Council, Ordnance Survey, Office for National Statistics, Office of the Deputy Prime
Minister, Landmark Information Group, Royal Mail, Hopewiser, ESRI and the Valuation Office Agency.