We are testing a new beta website for gov.scot go to new site

2009/10 Scottish Crime and Justice Survey: Technical Report

Listen

1 Background

1.1 Introduction to the Scottish Crime and Justice Survey

The Scottish Crime and Justice Survey ( SCJS) is a survey of public experiences and perceptions of crime in Scotland. The survey involves interviews with around 16,000 adults (aged 16 or over) who live in private residential addresses in Scotland annually, with a continuous fieldwork period from April through to March of the following year. The 2009/10 survey is the second sweep of the SCJS, with the first being conducted in 2008/09.

Fieldwork for the 2009/10 sweep of the survey began on 1 st April 2009 and finished on the 31 st of March 2010, with a target of 1,333 interviews being conducted each calendar month. The survey is conducted face-to-face in-home and is administered by specially trained professional interviewers using Computer Assisted Personal Interviewing ( CAPI). The survey also contains a short self-completion section on sensitive topics which respondents answer using Computer Assisted Self Interviewing ( CASI).

There have been nine previous surveys of victimisation in Scotland, beginning with the 1982 and 1988 sweeps of the British Crime Survey ( BCS) co-ordinated by the Home Office. 1BCS coverage in Scotland was limited to south of the Caledonian Canal. The first independent Scotland-only survey was commissioned by the Scottish Office in 1993 under the title of the Scottish Crime Survey ( SCS) and was followed by repeated sweeps in 1996, 2000 and 2003. 2 In 2004, following an external review, the survey underwent both a name change, under the title of the Scottish Crime and Victimisation Survey ( SCVS), and a major methodological change, with a move away from in-home face-to-face interviewing to telephone interviewing. However, the 2006 survey returned to face-to-face interviewing after it was shown that the robustness of the data produced by the 2004 telephone survey could not be substantiated. 3

The first sweep of SCJS was conducted in 2008/09. 4 The 2009/10 sweep retains the design of the 2008/09 survey with an amendment to the sample design and changes to the modular sections of the questionnaire (Box 1). 5

Box 1: Changes between the 2008/09 and 2009/10 surveys

Sampling:

The 2009/10 survey included a change to the sample design with further stratification by Local Authority ( LA) which led to a re-distribution of interviews within Police Authority Area ( PFA).

Questionnaire:

The modular design of the SCJS means that it is possible to rotate modules and sections on and off of the questionnaire, allowing the survey to explore new areas of interest year on year. As a result, the bulk of the major changes to the questionnaire were in the modular sections:

Main questionnaire:

Full sample modules:

  • Local community: section ADDED.

Quarter sample modules:

  • Module A:
  • Workplace violence: section REMOVED.
  • Module B:
  • Insulted / pestered / intimidated: section REMOVED;
  • Police: section ADDED.
  • Module D:
  • Civil justice: section REMOVED (but retained in quarter sample module C);
  • Crime scenario: section ADDED.

Self-completion questionnaire:

  • Sexual victimisation: reference period for incidents of serious sexual assault extended to cover last incident experienced since the age of 16 - see section 3.6.4).

The SCJS is primarily a victimisation survey, in which respondents are asked about:

  • Property crime experienced by the household as a whole ( e.g. housebreaking);
  • Personal crime ( e.g. theft from a person) which the respondents themselves have experienced.

The statistics produced from victimisation surveys provide a picture of the level of crime in the area covered. Respondents are asked directly about their experience of crime, irrespective of whether or not they reported these incidents to the police (police recorded crime). 6 These surveys provide a record of peoples' experiences of crime which is unaffected by variations in reporting behaviour of victims or changes in police practices of recording crime. However, the SCJS and police recorded crime statistics should be seen as a complementary series, which together provide a more complete picture of crime than could be obtained from either series alone.

The survey uses a victim form questionnaire to collect extensive details about the nature of each incident that respondents report (up to five), such as when and where it occurred and details about the offenders and other relevant information. This allows classification and hence counts of crimes in Scotland.

The SCJS collects information on incidents occurring in the previous 12 calendar months before the month in which the interview takes place. This time period is referred to as the survey reference period. Continuous fieldwork throughout the financial year means that the survey reference period varies depending on the month in which the interview took place, although the reference period covers an equal length of time (12 calendar months) for each respondent. More detail on the implications of this design is provided in section 5.1.

The SCJS only collects data on incidents occurring in Scotland in the reference period - incidents which happened in England and Wales are recorded in the BCS, and incidents which happen abroad are not covered by the survey (termed non-valid incidents). Incidents which meet these criteria and which are identified as crimes within the scope of the survey (see section 7.1) are used to produce the 'all SCJS crime' statistics which are published in the 2009/10 SCJS Main Findings report. 7

The remit of the SCJS is much wider, however, than a simple victimisation survey. The survey collects socio-demographic information from respondents which allow a picture to be built up about the nature of crime in Scotland and the risks of victimisation among subgroups of the population. It also collects information on a number of sensitive issues, including the prevalence of illicit drug taking, sexual victimisation and partner abuse (collected via the self-completion element of the questionnaire).

The SCJS has a quarter-sample modular design. All respondents (16,036) were asked the main questionnaire, and four randomly assigned groups of c.4,000 respondents were asked one of the four modules. This design allows a broader range of other topics to be covered.

All respondents were asked to complete the self-completion questionnaire though they had the option to refuse (13,418 respondents accepted the questionnaire - 84%).

Chapter 3 provides more detail on the coverage of the questionnaire.

The SCJS is designed to achieve a minimum effective sample size of 1,000 interviews in the eight Police Force Areas ( PFAs), allowing robust analysis at this level. The survey also provides analyses for a number of performance targets for the public sector in Scotland, at a national and a local level. Specifically, it is the main source used by the Scottish Government to measure progress against two of its National Indicators (used to monitor National Outcomes). 8

Despite changes in the design of crime surveys in Scotland over time (Box 2), the wording of the questions that are asked to elicit experiences of victimisation have generally been held constant. Care must be taken, however, when comparing different surveys, both those conducted in Scotland and other UK surveys, 9 and analysts should be careful to read the relevant technical documentation to ensure that like-on-like comparisons are being made. 10

Box 2: Variation in design of past Scottish crime surveys

The previous versions of the crime surveys in Scotland have varied in different ways, for example:

  • Data collection technique - from paper to telephone to Computer Assisted Personal Interviewing ( CAPI);
  • Sample size - from around 5,000 interviews in most previous surveys to 16,000 interviews for the SCJS;
  • Sample design - from excluding Highlands and Islands to covering the Highlands and major islands, and from representative data at only a national level to producing robust estimates at individual PFAs.

1.2 Outputs from the SCJS 2009/10

The data arising from the SCJS 2009/10 are reported by TNS- BMRB Scotland, who conduct the survey, working with the Scottish Government. These reports include the 2009/10 SCJS Main Findings report, as well as three supplementary reports on the subjects of partner abuse, sexual victimisation and illicit drug use. A separate User Guide is also available for data users.

Copies of the reports and other SCJS related Scottish Government publications are available from the Scottish Government survey website: http://www.scotland.gov.uk/Topics/Statistics/Browse/Crime-Justice/crime-and-justice-survey.

The survey data and supporting technical documentation, providing information and guidance for users of the survey data, are published on the UK Data Archive at the University of Essex. 11 There are three separate data files for the SCJS 2009/10:

  • Respondent file ( RF) (containing data from the main questionnaire and full and quarter sample modules);
  • Victim form file ( VFF);
  • Self-completion file ( SCF).

The SCJS is accredited as National Statistics by the UK Statistics Authority. This accreditation ensures that the survey data is produced to the highest professional standards and free from political interference. 12

1.3 Structure of the Technical Report

The survey sample design is set out in Chapter 2. Chapter 3 provides a summary of the content and development of the questionnaire, while Chapter 4 examines fieldwork procedures and response rates, and Chapter 5 the details and practicalities of the interview itself. Chapter 6 provides information on data processing, including the offence coding process and checking of data. Chapter 7 looks at the offence codes, survey statistics and crime groups used. Chapter 8 outlines the design, calculation and application of the weighting required for analyses of the data. Chapter 9 looks at the data outputs, including the structure of the SCJSSPSS data files and conventions used in them. Chapter 10 provides information on statistical significance and confidence intervals for the data and Chapter 11 presents guidance for comparing the SCJS data with other sources of data about crime.

Annexes referred to in this report are published as a separate document: 2009/10 SCJS Technical Report Annexes. 13

This report documents how the SCJS was designed, the way in which it was conducted and the how the survey data are produced, and should be read whenever using data from the survey. It is worth emphasising that the SCJS, in common with most victimisation surveys, is a complex study with data organised at different levels (households, individuals, and incidents) and has a number of sub-samples contained within it, including the modular and self-completion samples. Accordingly, data users should read this report before analysing the data to ensure that it is interpreted in a valid manner.

The dataset does not contain any indicators which allow identification of a particular person or address, and information which would allow this is not released to the Scottish Government or any other organisation by TNS- BMRB. Substantial emphasis is given in the course of the interview to assure respondents that the information they provide will be held in confidence.