Statistical Bulletin Crime and Justice Series: Fire Statistics Scotland, 2010-11

3. Changes to data collection and revision of data

3.1 Change in data collection

In April 2009, Scotland started to use the Incident Recording System ( IRS); this is an electronic system which captures data for all incidents that Fire and Rescue Services ( FRS) attend. The main advantage of this system is that all incidents, including non-fire incidents, are recorded fully. Previously the data capture for national statistics meant that fires involving fatal and non-fatal casualties were collected in full but other primary fires were sampled for data entry (re-typing into the database) and weighted to fire and rescue services totals. Secondary fires were only collected as a monthly aggregated total and special services were not included.

The transfer to the new system has managed to maintain the majority of main fire trends even though the data capture system has increased the sub-categories collected. In some cases, the sub-categories cannot be matched between this system ( IRS) and the previous paper collection ( FDR1) and the trend data has been affected. Where this has happened this information has been noted with an asterisk (*) and a footnote is supplied that refers to the annex explaining this change.

3.2 Changes in data source timings

This is the first time a Fire Statistics Scotland publication has contained provisional incident data as the most current data for all the tables. In the previous publication only Tables 1 to 5 provided provisional data, the rest of the publication was based on 2009-10 data. The Department of Communities and Local Government ( DCLG) decision to move to provisional data for it's Fire Statistics Great Britain Series ¹ ( FSGB) was due to the implementation of IRS. " Prior to April 2009, a large proportion of incident records was submitted on papers forms from FRSs across the UK. These were received in batches often after some delay, and they then required data entry. Electronic data capture and quality assurance have greatly reduced the timescales by which data can be submitted, meaning that statistics can be published sooner.

As a result of this, fatality numbers are now likely to be subject to some future revision, whereas before April 2009, sufficient time had elapsed by the time of publication for the records to be considered final. Consequently numbers of fatalities in the FSGB publication are now marked as 'provisional' since updates to these records, can cause noticeable revisions. Revisions to numbers of incidents and non-fatal casualties may also occur, but these are typically negligible at GB level. "

Scottish Government and DCLG are working together in order to maintain comparable data between the different fire statistics published by each nation in Great Britain. In the future this Scottish fire statistics publication will be produced on the most current financial year's provisional dataset. This enables us to provide timely data, with a view to finalising data in future publications. The change between provisional and revised figures is small for Scotland, for instance, the percentage change from the provisional 2009-10 data published in October 2010 to the revised figures in this publication is shown below.

Therefore the benefits of a more timely publication on provisional data outweigh the provisional status of the data. Finalising the dataset can take up to two years, mainly due to investigations into fire casualties.

In the Scottish publication, the provisional data in tables and charts will be annotated with a ^p. Where the data has been revised since last published - as in the case of 2009-10, an ^r will be used to annotate tables and charts. Fatal and non-fatal casualties will be described as provisional throughout the body of this text, but for ease of reading, provisional figures for fires, false alarms and special services will not explicitly be described at each mention.

3.3 New tables

A number of new tables have been introduced this year:

Special services:

False alarms for special services have been included ( Table 4b and 4c). This information has only been available at national level since the introduction of IRS. The data is published to improve information on the types of incidents that FRSs attend.

A national breakdown on the most frequent special services - Road Traffic Collisions and Flooding - has been provided to give more detail on the nature of these incidents.

Gender of casualties:

There have been a number of requests within Scottish Government for the gender of fire casualties. This information will now be published and a ten year trend has been provided ( Tables 9 to 10d).

The influence of alcohol/drugs

IRS has the facility for the FRS to indicate where they believed that a fire may have occurred due to the influence of alcohol/drugs. Tables 19, 19a and 19b have been introduced to provide evidence to underpin community fire safety policies. At the moment, only 2 years worth of data is available and it is therefore too early to identify any trends.