The trouble with measuring crime by (on or near a) street

Posted: February 26th, find 2012  Author:   No Comments »

We’ve had a lot of positive feedback from the media coverage we received in the Sunday Times and Sunday Telegraph today. Since we launched UKCrimeStats last April, here we’ve always made clear that the official police data that we use that gives us Easting/Northings which can be mapped to a street are not a precise location of the crime. To date we have around 7 million crime and ASB incidents that are mapped to 536, viagra 000 snap-points which are locations deemed to be on a street near or close to where the incident actually took place. So problem number 1 is accuracy.

Problem number 2 is that not all streets are equal. They vary hugely in length, the number of people living in them and what they have on them that may attract a high transient population or footfall because they are high streets for example. We have no street level population data to deflate for the impact of crime by using a crime rate (recorded crime per 1,000 residents).

Problem number 3  is when police.uk decides unilaterally to change the snap-points without telling the 3rd party developers. This happened last month. Here’s a video we’ve prepared to help you understand the consequences.

We’re all for increasing the accuracy of the snap-points but it does mean that all previous history becomes void – that’s why many of our individual streets like St James Street in Weston Super Mare are showing zero crimes for December 2011. Perhaps it’s because we have kept a history of which snap-points accumulated the most crime that they realised there needed to be more granularity.

Problem number 4 is that we know that sometimes crimes with no location are allocated to a street with a Police Station but we don’t know how many or where the Police Stations are. This has a distortionary affect on the street crime level for that street but we are unable to filter it out.

Problem number 5 is please don’t blame us for just representing the official data. We really don’t understand the ill-thought through comments by a spokesperson for Avon and Somerset Police in the Sunday Times today. We’re always happy to explain  how we calculate our figures to the Police and Press, have done so in the past and will continue to do so. Just wish Avon and Somerset had given us a call first !

Our data is the same as on Police.uk – what we do differently is to aggregate the data so that we can tell who has the most and of what kind or the biggest increase of crime between or over two different time points. That is one of our unique services we provide and a very useful tool for the general public and for the NPIA as well because we find errors in the data that they can’t see and we continue to log on our Forum. However, police.uk have not released the same points of interest that they recently uploaded from the ONS – so we don’t necessarily know if there’s a nightclub nearby although as we use google maps and we also unlike police.uk have a satellite view – you can always zoom in and have a look yourself.

Despite all these caveats, problem number 6 is that everyone still wants to know about streets – it’s one of the most popular sections on UKCrimeStats. So we have created new pages for the new snap points – but they are not actually always on or near a street. So please bear in mind all these points if you want to know which streets have the most Total Crime / Burglary, ASB, Violent Crimes etc. . . .

Article source: http://www.ukcrimestats.com/blog/2012/02/26/the-trouble-with-measuring-crime-by-on-or-near-a-street/