If you're wondering where our data comes from, how we create our metric types and how we show data for different areas, please read the text below.
The metric types available for LG Inform and LG Inform Plus are collected from numerous different sources and all datasets are checked before being uploaded. Sources include the Department for Communities and Local Government, Public Health England, the Office for National Statistics and many more. (Ask us if you need a full list.) The values in these datasets are attributed to standard geographies, e.g. LSOAs, wards, local authorities, etc., but aren't always available for all levels of geography.
We try to provide metric values for as many different types of area as possible in LG Inform Plus. For example, some population metrics are published for output areas only, therefore we aggregate that data to show population values for wards. When the shape of the smaller areas does not precisely fit the larger areas, we determine a 'best fit' by one of the following means.
- Published best fit areas – such as the best fit output areas published by ONS for major towns and cities.
- Where ONS has not published up-to-date best fit output areas, but we understand their methodology is to assign according to population centric centroids, we apply the same methodology (as for 2015 civil parishes). The methodology given below is used for older areas (such as best fit LSOAs for wards).
- Areas that cover an entire region or country (such as civil parishes across England) are assigned best fit output areas on the basis of the largest proportion of the output area that falls into one area (e.g. parish). So if an output area is 30% in area A, 30% in area B and 40% in area C, it is assigned to area C.
- Areas created in the Natural Neighbourhoods tool are given best fit “child” areas of all the areas of the child area type whose centroid falls in the new neighbourhood.
As a result of these different methodologies, the values for an area that you see in LG Inform Plus may not be concurrent with internal data or other sources, but please do ask us if you have any queries. We're always looking for ways to improve the tools and welcome your feedback
Some of our metric types are calculated from others published for the area, for example calculating the rate of crime per head of population by dividing crime statistics by the population headcount.
If there’s a dataset that you think would be useful, please don’t hesitate to suggest it to us.