Road casualties: FAQs

The BBC News website's road casualties map, which shows all recorded deaths on Britain's roads between 1999 and 2010, has been enormously popular.

As a result, we have received hundreds of emails asking a range of questions. Here are the answers to some of the most frequent about the map and related features.

Where does the data come from?

Every time a road traffic collision is reported to the police in which someone is either injured or dies, the police record details of the incident on a special form, known as STATS19. The police record many details including when and exactly where the crash took place, in what weather, on what class of road and whether it was close to a junction, a school and son on. All of this data is collated every year. The Department for Transport releases quarterly headline figures based on this data and once a year, a full annual report. The latest report for 2010 is here.

Why do you say the data covers Great Britain when it includes Northern Ireland?

The complete data over the time period - 1999-2010 - is for Great Britain. However, we have also included some partial data for Northern Ireland, for 2004-2009. The Northern Ireland data comes from a different source and was in a different format and so isn't directly comparable. Some of it was also incomplete.

For this reason it is not possible for us to say the data is representative of the UK as a whole. Our headline reflects this and we have added a note beneath the map to explain.

A crash close to where I live is missing or the details are inaccurate. Why?

The data comes from police records dated between 1999 and 2010. The information is recorded by police at the scene of a crash. The data is then transferred onto a computer database. As with any large collection of data, there will be some errors and duplicates in the information and these will appear on our map. In addition, the circumstances may have changed since the information was recorded or more accurate information may have come to light.

With regard to the missing incidents, there are 86 fatal collisions in the police data made available to us that have incomplete information. For this reason, we have been unable to include them in the map.

The BBC apologises for any distress or offence that any inaccurate information may cause.

A crash I am familiar with is not located in the right place on the map. Why?

The data is postcode based, so the points on the map may not be positioned in the exact location of the fatal collision.

The map doesn't seem to show individual crashes for my postcode area. Is there a problem?

If there are too many fatal crashes in your local authority area, the black overview square showing figures for the number of collisions will appear instead of individual crash dots. Zoom in further and you will see the individual incidents and their locations. This will happen with a number of areas on the map where there are a large number of crashes.

Some of the incidents I have looked at say it was snowing at the time, but it was summer. Is there a problem with the data?

There was a glitch in the way the map was being fed with data early on 5 December. This has now been resolved.

Where did you get the 730,000 figure for all road casualties?

This is the government's own official estimate of the total number of people killed or injured in road traffic collisions.

How did you calculate the £15 - £32bn economic cost?

Another figure from government officials. Here's a detailed article on how this number is calculated.

What is the source of the Dangerous Driving slide?

This research was conducted by Loughborough University's Transport Safety Research Centre and the Transport Research Laboratory. The two teams tracked hundreds of crashes over several years in the Nottinghamshire and Thames Valley areas respectively. Researchers attended these incidents and recorded first-hand many features of the collisions for the On-The-Spot research project.

They concluded that there seem to be links between offending and collision-involved road users; links between deprivation and precipitating a road traffic collision; offending among people driving for work; and potential over-representation of offending amongst collision-involved road users, compared to the national data.

They also emphasize that this research is preliminary and relates to collisions within the areas studied. It doesn't necessarily follow that this represents a complete national picture, though having demonstrated a way to link in-depth data on collisions with offence histories, there is the potential for further useful research.

Links to Loughborough's research, and TRL's research can be found here.

What is the difference between a "serious" and "light" injury?

These are defined in this document which details how the STATS19 form is filled out by the police.

Are you more likely to be involved in a collision during rush hour in a city?

Not necessarily. The police data we have simply details all the collisions that have happened over 12 years. However the busiest times on roads, while seemingly dangerous, may not be statistically the most dangerous times. And roads where many collisions take place may not be statistically the most dangerous roads. A quieter country road could be more dangerous than a busy city road if it has a higher number of incidents in ratio to its traffic volume.

Road casualties feedback