Non-sexual violence in Scotland: report

9.3 Challenges in measuring violent crime

9.3.1 Which is the better source?

It is challenging to measure the full extent of violence using just one source. By combining sources a fuller picture of violent crime in Scotland can be obtained. 

Each source has its own strengths and limitations, see Section 9.3.5. For example, while a broader range of crimes can be found in Police Recorded Crime and Criminal Proceedings datasets, the SCJS estimates that only 39% of violent crimes were reported to the police in 2017-18 (which is consistent with earlier years). Given the under-reporting of violent crime and that police recording practices can often be subject to change, other sources such as victimisation surveys (the SCJS) and Hospital Admissions data can be used to further improve our understanding. 

The SCJS and Hospital Admissions data are however themselves subject to limitations. The SCJS for example can only provide information on those over 16 years old in private households, while Hospital Admissions data are reliant on a person to be admitted and for their injuries to be identified and coded as being due to violence (those receiving care from a GP or in an outpatient setting and those attending emergency departments but not admitted will not be recorded). 

9.3.2 Non-standardised definitions

For each source there is a detailed process used to define the crime. There are however slight variations in how crime terms are used and defined by each source.

In particular, it is important to note that: 

• The SCJS includes actual and attempted minor assault (which is simliar in nature to the what the police would record as a ‘Common assault’), as well as Serious assault and Robbery.
• Police Recorded Crime data does not include Common assault in the category of non-sexual violence (which can also be referred to as Group 1). However when considering trends in non-sexual violence more broadly within this report, Common assault is included in our understanding of violent crime. 
• Hospital Admissions data can only capture violent crime when an injury serious enough to need an emergency admission to hospital has occurred, thereby omitting violence that involved less serious or no injury.

For details on how terms are used in this document, see the Glossary.

9.3.3 Defining and analysing deprivation 

In order to assess deprivation, the geographic location of each crime in the samples was assigned to a datazone. These datazones were then compared to the Scottish Index of Multiple Deprivation (SIMD)^[45]. This comparison was only possible for data from the SCJS, Robbery and Attempted murder & Serious assault. Data from Homicide were not able to be linked to the SIMD in this way.

The SIMD is a tool for identifying concentrations of deprivation in Scotland. SIMD uses information from a range of indicators including income, employment, education, health, access to services, crime and housing to provide a measure for each of the roughly 7,000 datazones in Scotland. Each datazone is provided a rank ranging from 1 (most deprived) to 6,976 (least deprived) and also defined as being either Urban, Rural or a Small town. 

As the sample for both the Robbery and Attempted murder & Serious assault analyses were drawn across two years, two different editions of SIMD were used for the comparison, for example, the SIMD12 was used to compare crimes from 2008-09 and SIMD16 compared to crimes from 2017-18.

9.3.4 Time periods examined

The precise time periods for each source differ, meaning that they are not directly comparable with one another, although broadly each source refers to a similar time period per annum. 

The 2017-18 SCJS figures included in this report cover interviews conducted between April 2017 to May 2018. As the survey asks respondents about incidents experienced in the 12-months prior to the month of the interview crimes included in the survey could have occurred anytime between April 2016 to May 2018. Because of this data from the SCJS is not directly comparable to data from other sources collected over a specific year. 

The self-completion Partner abuse data from the SCJS and the Police Division level figures are from the 2016-17 and 2017-18 sweeps of the survey. The data are merged over two years to provide a larger sample size, with findings therefore covering 2016-17 to 2017-18 (referred to throughout as 2016-18). The majority of the Police Recorded Crime data referred to here, relates to crimes recorded by the police in the financial years of 2008-09 and 2017-18 (i.e. from April to March of that year). Some of these crimes may have taken place prior to the year in which they were recorded but did not come to the attention of the police until a later date.

The Homicide National Statistics data are derived from the total number of Homicide cases, victims, or perpetrators for the three year periods ending in 2008-09 and 2017-18. This method is used to mitigate against large fluctuations that can occur in Homicide data due to the relatively low numbers involved, and allowing trends to be understood and analysed. Any individual Homicide case is included against the year in which the crime that led to the homicide is first recorded by the police, which is not necessarily the year in which the victim dies.

The Hospital Admissions data relates to emergency admissions to hospital (as a result of an assault) during the financial years of 2008-09 and 2017-18.

Criminal Proceedings data refers to court proceedings, as recorded on the Criminal History System (CHS), that were concluded during the financial years of 2008-9 and 2017-18. Data is only received once the sentence details are given. A pending case on the CHS should be updated in a timely manner but there are occasions when slight delays happen, these usually affect High Court disposals as they are the more complex and length trials. 

9.3.5 Strengths and limitations of data sources