About the MAST Mapping Dataset
Introduction
The MAST Mapping online tool powered by eSpatial contains over a hundred different measures of road safety performance at Highway Authority and District level. In order to understand more about what measures we chose and how they were calculated please read on.
Key Outcome Indicators
In May 2011, the Government produced its ‘Strategic Framework for Road Safety’, setting out its key themes for road safety, including a Road Safety Action Plan. The document explicitly states that the Government does not “believe that over-arching national targets or central diktat that constrains local ambitions and priorities are now the most effective way of improving road safety”. Instead they plan to move “to a more sophisticated method of monitoring progress through a Road Safety Outcomes Framework. This should help local authorities to assess and prioritise their action and show the impact of central Government measures.”
Key indicators have been proposed within the Strategic Framework, which will allow local organisations and citizens to monitor progress towards improving road safety. Progress will be reported annually in Reported Road Casualties Great Britain. The indicators refer to casualty and collision rates across a range of road user groups and involve using a range of data sources.
Road Safety Analysis has produced 116 thematic maps which allow MAST users to easily compare baseline data across a number of key indicators for local and highways authorities. This article explains how they were calculated and what the maps show.
Thematic Maps
All of the thematic maps are based on 2006-2010 Stats 19 data, provided by the Department for Transport (and taken from the MAST database). Whilst there are no specific targets within the Strategic Framework, performance will be measured by the Government against this baseline period. Each indicator is shown by local authority and by highways authority and is assigned a specific code (KOI1 to KOI9). They are also calculated by the following:
- Casualties by crash location (green maps)
- Casualties by residency (blue maps)
- Crashes by crash location (orange maps)
The ‘residency’ indicators (because they are based on populations) are not just shown as total values, but also as indexes and annual rates. It means that the indicators are displayed in 116 different ways and allow all permutations of analysis.
For all of the maps, the darker the colour the higher the risk.
In order to be sure that the correct map has been selected, index tables have been created so users can be sure of the description; measure; severity; and measured area of each indicator. We reccommend you keep a copy of this table handy when analysing the maps.
The maps are stored on 6 different layers (casualties by crash location highways level; casualties by crash location local authority level; casualties by residency highways level; casualties by residency local authority level; crashes by crash location highways level; and crashes by crash location local authority level) so use the index tables to determine which layer to look at.
In order to get the most out of these maps, we have produced a guide to using eSpatial.
What the maps show and how they were calculated
Casualties by residency
This set of indicators is, for us, the most interesting and useful. Instead of measuring stretches of tarmac, we have mapped the home postcodes of casualties in order to show areas of particularly high or low resident risk. To put the indicators fully in context, we have mapped this set by three measures: total numbers of resident casualties per authority; annual average rate of resident casualties by population per authority; and assigned an index to each authority to show how over- or under-represented each ‘casualty by population’ rate is.
For this set, we have added an additional six indicators to those in the Strategic Framework. These extra indicators show all severities for indicators KOI5 to KOI8 and include two new indicators (KOI9), which show child casualties (both KSI and all severiy). The child casualty rates and indexes are calculated using 0-15 population data.
In order to produce residency indicators as accurately as possible, unknown postcodes have had to be dealt with. Postcodes are not always recorded and some areas have better reporting practices than others. If all unknown postcodes were removed from the analysis then areas with low reporting rates are likely to have low resident risk (because a large percentage of these casualties are likely to have been local and would not have been counted). We therefore created ‘correction factors’ to allocate each of the unknown postcodes to an authority area. This is done by determining for each police force the percentage of casualties who live in each authority area and were injured in that police area. The unknown postcodes for each police force were then redistributed to authority areas, based on these percentages, under the assumption that the casualties with unknown postcodes are likely to be similarly distributed across the authorities as the known residency casualties.
Once we’d calculated the absolute numbers of casualties who lived in each authority area, we wanted to put the figures in context. If we just looked at absolute numbers of casualties then areas with high population density would often appear to have disproportionately high casualty figures. The following three maps show young driver casualties who live in each local authority area and shows how using population data can put the absolute numbers into context.
Absolute Numbers
The first map shows the absolute number of young driver casualties from each local authority area. A great deal of London is the darkest shade of blue, as is Wiltshire, Central Bedfordshire and Buckinghamshire. In Wiltshire, there were 1270 young driver casualties compared with 285 in Maldon, Essex. Based on this information, it would seem that Wiltshire has a significantly larger problem with young driver casualties than Maldon.
Annual Rates
The next map shows the annual rates of young driver casualties for each local authority (calculated by dividing the annual number of casualties by the population of the area) and produces a ‘one-in-x’ rate for each area. So, taking population into account, one-in-1793 Wiltshire resident young drivers are casualties each year compared to one-in-1108 for Maldon in Essex (and so has a lower risk).
Indexes
The last residency map shows the indexes for each local authority area. As with the annual rates, a very different picture appears to that of the absolute numbers. The indexes are like those applied to Mosaic analysis: an index of 100 would indicate that the target group in question is behaving exactly as we would expect. Any index over 100 indicates an over-representation of the target group compared to the base population and any index under 100 indicates an under-representation. In this case, Wiltshire has an index of 95, suggesting that Wiltshire young drivers are appearing pretty much as often as we would expect within the casualty statistics, given the population size and compared to young driver casualties everywhere. They are 5% under-represented within the young driver casualty figures. In comparison, Maldon, Essex has a young driver casualty index of 154, showing that these young drivers are 54% over-represented within the figures, given the area’s population and compared to the rest of the country.
All three measures have their place but hopefully these maps have shown how including population figures change the story; put the statistics into context; and will help you choose which maps to select.
Casualties by location
This set of indicator maps show a traditional way of measuring casualty reduction performance. Authorities tend to be judged on the number of casualties who are injured on their stretches of road, regardless of where they come from. As population figures are not relevant with casualties by location, only absolute numbers are shown. We have also only produced 8 key outcome indicators in this set (for each authority level), based on the Government’s Strategic Framework.
The next map shows young driver casualties but this time it is mapped according to the area in which they crashed. It also shows KSI casualties and not all severity of injury, unlike the previous maps. It shows quite different areas of concentration to the residency maps. Whilst the severity levels could account for some of the difference between this map and the absolute numbers of ‘Casualties by Residency’ one, it is also likely to be indicating areas where casualties are not local. An authority that is very dark green in the ‘Casualties by Location’ map but lighter blue in the ‘Casualties by Residency’ one is likely to be in the situation where local people are not over-represented in the casualty figures but instead the casualties come from elsewhere (as tourists or by passing through the area, for example). This information is invaluable for targeting road safety campaigns – you need to know where your audience comes from before you can develop an effective intervention to engage with them.
Crashes by location
The last set of indicator maps show the number of crashes which occurred within an authority area. This too is a traditional way of measuring road safety performance. As casualties and residency are not part of this set of indicators, the only measures are absolute numbers. As with the ‘Casualties by Location’, we have also only produced 8 key outcome indicators in this set (for each authority level).
The orange map shows the number of crashes which involved a KSI young driver casualty by local authority. It produces similar results to the ‘Casualties by Location’ map.
Using the Maps
We hope these maps will provide a starting point for setting local road safety priorities by putting the indicators into context. It should allow local road safety professionals and decision-makers to determine the extent of their casualty and collision issues compared with other authorities and to choose which road user groups should be focused upon at the local and regional level.
