Annex A: Sources of information and data quality
A.1 The annual aggregate CJS return for local authority criminal justice social work services was introduced for 1999-00 and covered social enquiry reports (now criminal justice social work reports), community service orders and probation orders. The content and format of the return has changed over time to reflect new developments and an increasing demand for information, as well as to clarify points of definition in relation to particular data items. Additional items include:
- supervised attendance orders (2000-01, removed (alongside community service orders and probation orders) from 2016-17 onwards)
- throughcare (statutory post release supervision) (2001-02)
- diversion from prosecution (2001-02)
- drug treatment and testing orders (2003-04, removed from 2012-13 onward)
- bail information (2003-04)
- voluntary throughcare (2004-05)
- court services (2004-05)
- throughcare addiction service (2005-06, removed from 2015-16 onward)
- community payback orders (2011-12, removed from 2012-13 onward)
- fiscal work orders (2015-16).
A.2 Data for community payback and drug treatment and testing orders has been collected at unit level for each order since 2012-13. The aim of this change has been to enable analysis of the process and outcomes for individual orders, which is not feasible through a collection of aggregate tables. This allows the scope for looking at how each order progresses and provides more detailed information on outcomes.
A.3 All 32 Scottish local authorities have been able to provide the CPO unit level data in each of the last five years. For some of the tables in the bulletin, local authorities were not able to supply all of the information requested. As a result, these tables include estimates, and such cases are indicated in the footnotes.
A.4 With the introduction of the CPO, information on the legacy orders (community service, probation and supervised attendance orders) has been phased out of the aggregate return. Due to the small numbers involved, recent years’ returns have not collected information on the number of legacy orders commenced and terminated. These numbers are estimated to be fairly small and do not adversely affect trends observed across the years.
A.5 As a result of the change from the legacy orders to community payback orders, it is important to note that comparability between the data for subsequent years may be affected by the type of people most likely to complete a community payback order during the first few years of implementation. Orders which finished during the early years after they were introduced tended to be lower-tariff orders which generally took less time to complete than those which finished in later years.
A.6 The statistics presented in this bulletin reflect information on criminal justice social work activity in the financial year 1 April 2017 to 31 March 2018. Figures are extracted from live information management systems and may differ slightly from those published previously as administrative systems are updated.
A.7 Revisions are flagged up in the publication at the time but not in future publications. The live tables, including earlier data at sub-Scotland level, on the Scottish Government crime and justice statistics website may be revised at any point if required, and revisions are highlighted in the relevant table.
A.8 As a result of information provided by local authorities with their 2017-18 unit returns, some revisions were made to the 2016-17 and (to a lesser extent) earlier years’ data. These revisions were mainly a result of:
a. The inclusion in the 2017-18 data of orders which were in existence before the 2017-18 year but which had erroneously not been included in earlier years’ data returns, and
b. Some orders which were previously advised in the 2016-17 returns as being in existence at the end of that year but which had actually been completed/ terminated before then. Some authorities advised of substantial numbers of CPOs falling into this category, due mainly to orders not being closed off on their IT systems. However, the number of authorities for which this was the case is declining each year and this is expected to improve further in the future.
As a result of these changes, the total for CPO terminations in 2016-17 has been revised upwards by around 400, with CPOs in existence at 31 March 2017 revised downwards by 470.
A.9 For some new records added to the CPO data for 2012-13, 2013-14 and 2014-15, not all of the necessary information was supplied. Some tables therefore contain a "not known" category for the information in question (e.g. Table 11, CPOs by court type).
A.10 In this bulletin, the calculation of successful completion rates for community payback orders and drug treatment & testing orders (Table 2) is different from previous years. Instead of dividing the number of successful completions and early discharges by the total orders terminated, they are now divided by the total less orders which were transferred out of a local authority. As a result, successful completion rates are slightly higher than reported in previous years although the trend over time remains similar.
A.11 Figures in this bulletin on the number of new orders commenced are not collected on the same basis as those published in the criminal proceedings bulletins. This is due to differences in the unit of analysis (cases versus orders) and criminal proceedings data referring to the court rather than the local authority implementing the order.
A.12 The data obtained from local authorities is considered of good quality as they come from recording systems which the local authorities use for case management and for internal monitoring. However, administrative data of this type will be subject to some degree of error which may arise in any large scale recording system. Therefore the data has been quality assured as far as practicable through a series of validation processes before publication.
A.13 The aggregate return includes electronic checks to notify local authorities of inconsistencies within the data. In the case of substantial changes since the previous year, the local authority is asked to confirm the figures are correct. Once data returns from all authorities have been received, further analysis is carried out to detect any major changes to figures over recent time periods.
A.14 The unit level collections for community payback and drug treatment and testing orders include automatic checks built into the data loading system so that errors in information provided are flagged up at an early stage for correction. The data are checked for accuracy by internal statistical administrative staff. Checks are also made to ensure there is consistency between 2017-18 and earlier years’ data.
A.15 Additional datasets at local authority level are available on the Scottish Government website.