Repeat violent victimisation: evidence review

2. Methods 

This RER followed a systematic process of search and assessment, which involved four broad stages:

1. Evidence search

2. Application of inclusion and exclusion criteria for assessing relevance

3. Quality assessment of studies

4. Synthesis of the body of evidence 

The details of stages 1-3 are described below, with the synthesis of the evidence presented in Section 3.  

Search

In the first instance, the search for studies was carried out by the Scottish Government Library Service using KandE. KandE is an online search engine which covers a range of high quality databases, which are detailed in the below table.   

Table 1: List of Databases Searched

This search was informed by key words submitted by analysts within the JAS division of the Scottish Government. The search phrases included:

• “multiple victimisation” OR “multiple victimization”
• “repeat victimisation” OR “repeat victimization”
• “repeat violent victimisation” OR “repeat violent victimization”
• repeat AND (victim OR victims OR victimisation OR victimization)
• revictimisation OR re-victimisation OR revictimization OR re-victimization
• polyvictimisation OR polyvictimization OR poly-victimisation OR poly-victimization

These search phrases were used to identify studies exploring the extent, prevalence and nature of RVV. However, it is important to note that there are certain types of violence which are inherently repetitive (e.g. domestic violence and racial attacks) which are not explored extensively in this review. Although there is a wealth of literature on types of RVV such as these, given the timescales for this review, and the fact that some types of RVV have distinct causes and consequences to others, it was not practical or feasible to review and quality assess this literature for the purposes of the present RER. Therefore, while this review covers a range of literature on RVV, it should not be regarded as a comprehensive or definitive account of the evidence. Rather, it constitutes a collation of relevant material which could be identified and accessed within a relatively short period of time.

Coverage

The specified time coverage for the search was from the last 5 years and the geographical coverage was international. A series of broader searches were then conducted using Google and Google Scholar, as a sweep of studies that may not have been found in the initial search. In addition, a snowballing technique was employed whereby the references of studies were reviewed for additional evidence. 

The geographical coverage of the secondary literature search was also international. However, although the initial literature search only covered material from the last 5 years, the secondary literature search identified sources dating back to 1990. The central reason for this is that in the 1990s and 2000s there was a proliferation of research on RVV, as interest in the issue grew in research and policymaking circles to varying degrees in the UK and elsewhere (Shaw and Pease, 2000). Excluding research conducted within this period would therefore mean discounting a substantial proportion of the existing evidence on RVV. The implications of including these older studies are discussed below.  

Inclusion and Exclusion Criteria

Using the initial search results, the relevance of the studies to the research questions was assessed. The table below provides a summary of the inclusion and exclusion criteria applied to the selection of the studies. 

Table 2: Inclusion and Exclusion Criteria

Applying these criteria led to an evidence base comprising a wide range of sources, including academic articles, government reports, surveys, evaluations, evidence reviews and books. The studies were based in England and Wales, Scotland, Sweden, the Netherlands and the US. The studies identified are detailed in Appendix 1^[1]. 

Quality Assessment

Each of the studies identified was then quality assessed. This involved identifying the key characteristics of the studies and their limitations, which are summarised in Appendix 1. 

Strength of the Evidence 

The body of evidence identified in this report consists of 43 studies. Many of these use robust and high quality methods. For example, 20 use nationally representative survey data, which are either cross-sectional or longitudinal in nature. There were also at least 6 studies identified which employed qualitative methods, including interviews and focus groups, providing a more in-depth insight into the nature of RVV. The remaining studies predominantly used quantitative methods of data collection and analysis, including police records and surveys.  

However, as well as the limitations highlighted in Appendix 1, the evidence base also suffers from several further shortcomings. Firstly, although some qualitative research was identified, the overwhelming majority of studies use quantitative methods. Given the importance of qualitative research for illuminating social processes and meaning, the lack of qualitative investigation of RVV appears to be a gap in the evidence base. Moreover, most of the research identified was conducted outwith Scotland, with the majority from England and Wales and the US. Indeed, only 3 studies collected primary data in Scotland (Scottish Government, 2019; Shaw and Pease, 2000; Farrell et al., 2005). The findings from many studies, therefore, are not necessarily applicable to Scotland. Finally, much of the research is now dated, with only 18 of the studies identified published in the last 10 years. This has implications for the findings of the review. In particular, some of the conclusions of earlier research may no longer be applicable or relevant in the present day. The findings discussed below should, therefore, be considered in the context of the limitations identified.