Redevelopment of domestic abuse statistics: research update November 2021

The way domestic abuse manifests is constantly changing. To ensure our statistics continue to provide the most accurate information and meet the needs of users, we are undertaking a user engagement, research and testing program to improve the collection of data on domestic abuse.

The questions on domestic abuse asked as part of the self-completion section of the Crime Survey for England and Wales (CSEW) have largely remained the same since they were introduced in April 2004, allowing for a long comparable time series. However, there are a number of issues with the data currently collected, mainly:

• they do not align with the definition of domestic abuse introduced in the Domestic Abuse Act 2021
• they exclude the offence of coercive or controlling behaviour introduced in 2015
• they do not measure the number of incidents or frequency of abuse
• there is a greater user need for data to understand the nature of the abuse

Given these, significant changes to the questions are needed. In line with wider CSEW redevelopment, we are also looking at survey mode to consider alternative options.

This update outlines our recent research, findings, and plans for redeveloping domestic abuse statistics over the coming months. As the work progresses, we will continue to engage with stakeholders to provide updates and seek feedback.

The offence of controlling or coercive behaviour in an intimate or family relationship was introduced as part of the Serious Crime Act 2015. In response, we conducted research to develop survey questions to measure its prevalence. The questions were developed and evaluated through the ONS Domestic Abuse Statistics Steering Group (DASSG), drawing on both the legislative guidance and advice from topic experts. Between April 2017 and March 2019, we included new questions on the Crime Survey for England and Wales (CSEW) on a split sample basis. Our April 2019 publication outlined our research and findings.

Analysis of the new questions concluded that there was uncertainty in whether the measure adequately captured victims of the offence. Working closely with topic experts it was decided that the questions required further development and testing before we could be confident that they were fit for purpose. As such, the questions tested in the split-sample experiment were removed from the survey in April 2019 to allow this work to take place.

In summer 2020, we carried out a user survey to understand more about how users of the CSEW domestic abuse statistics made use of the data. In addition, we were interested in understanding future requirements and priorities for continuing the time series, making improvements to the measurement of domestic abuse and recording the frequency and nature of abuse.

Analysis of the user survey data highlighted a number of issues with the data currently collected and suggested that a wider program of improvements was required.

Our research to develop survey questions to measure controlling or coercive behaviour is ongoing and now forms part of our wider redevelopment work on domestic abuse statistics.

In Autumn 2020, we put out a research tender to take forward the redevelopment of domestic abuse statistics. There were two main parts to the research:

• to explore the issues with the current survey questions and data collected, alongside the user requirements
• to investigate the use of alternative survey modes to ask respondents about their experiences of domestic abuse

In November 2020, we awarded a research contract to a consortium led by the University of Bristol¹, also involving the College of Policing, Women’s Aid and Men's Advice Line. The work took place between December 2020 and the end of June 2021 and comprised of:

• a review and mapping exercise of the current questions asked on domestic abuse in the Crime Survey for England and Wales (CSEW)
• an international comparison of survey questions on domestic abuse
• focus groups and individual interviews with male and female domestic abuse victims and survivors
• analysis of the 2020 user survey data
• a non-response analysis of the current survey questions
• a rapid evidence assessment of the impact of survey mode on sensitive questions
• consultation with a core group of topic experts and stakeholders

Survey content and structure

The research recommended that future data collection should focus on the identification and estimation of different “abuse profiles”. This would complement an improved headline measure, allow for more targeted follow up questions to further understand the nature of the abuse, and better inform policy and practice. For example, it could be possible to differentiate victims of “one-off” or infrequent abuse where the impact felt is reported as limited, from those suffering ongoing, coercive abuse.

Identifying different abuse profiles would allow an alternative use of follow up survey questions as different victim profiles may have different characteristics in terms of sex, the impact of abuse, and reporting and help-seeking.

The research concluded with the following recommendations (in priority order).

The headline prevalence measures of abuse should be revised and improved

Most stakeholders agreed that the existing headline measure was not fit for purpose. There was still a requirement to ask respondents about victimisation both since the age of 16 years and in the past 12 months, by a partner, ex-partner, and family member. Most felt improving the headline measure was more important than continuing the current time series.

The survey should measure coercive control, following robust question development and testing

Coercive or controlling behaviour is the only domestic abuse-specific criminal offence and a robust measure was highlighted by stakeholders to be a priority. It is recognised to be at the core of domestic abuse and cannot be separated from other abusive behaviours. Therefore, it is an integral part in the identification of abuse profiles. However, there remain significant issues relating to measuring coercive control, and there is no agreed measurement instrument internationally.

Measurement of impact

Stakeholders and survivors highlighted the importance of measuring impact as a means of determining severity and harm. Analysing impact alongside forms of abuse is also important in deriving an accurate overall prevalence measure of it. The research recommended that this could be achieved through the identification of different abuse profiles.

Measurement of physical assault should be disaggregated

To identify different abuse profiles, it is important for levels of physical assault to be disaggregated. In particular, strangulation and choking are a highly coercive form of sub-lethal violence, and non-fatal strangulation is now a criminal offence. The severity of violence is also a key means of discriminating abuse likely to come to the attention of support services and statutory agencies.

Further consideration should be given to measuring frequency of physical (and sexual) victimisation

Measuring the frequency of physical assaults is a helpful way to identify different profiles of abuse. However, there are challenges to accurately measuring this within the self-completion part of the CSEW. Survivors expressed reservations about measuring frequency, both in terms of recall and appropriateness, but further development work could determine a viable approach.

Consideration should be given to matching abuse to a specific perpetrator

Although not currently possible, being able to link victims and the type of abuse experienced to a specific perpetrator(s) would enable a more robust interpretation of the data collected on nature of abuse. However, survivors felt that a general question on numbers of perpetrators could be perceived as victim blaming.

Questions relating to whether what the victim reported to the survey was abuse or a crime, attitudes to domestic abuse, and abuse experienced by other people should be dropped from the survey

Survivors felt the attitudinal questions and those relating to abuse experienced by other people were insensitive and victim blaming. Dropping these would provide space for measuring more important issues such as frequency and severity of physical assault and impact.

Survey mode

The following recommendations were proposed regarding the survey measurement of domestic abuse.

Consideration should be given to how and where questions on domestic abuse are best hosted in relation to the wider social survey infrastructure

The context of the survey within which data on domestic abuse is collected has important implications for estimates of prevalence. It is well established that general purpose victimisation surveys underestimate prevalence of domestic abuse in comparison with targeted domestic abuse surveys. The Crime Survey for England and Wales (CSEW) offers many advantages in surveying domestic abuse. However, the framing of the CSEW itself may discourage disclosure of wider symptoms of abuse, for example, because of its legalistic focus on criminal incidents.

New modes of data collection such as online surveys might offer one means of reframing domestic abuse questions within the CSEW. This could make them less directly linked to the wider CSEW focus on victimisation and reporting of criminal incidents. Consideration might also be given to whether alternative official surveys (for example, on health and wellbeing) offer a more suitable platform in the longer term.

Decisions about mode of respondent contact will depend on the preferred mode of data collection

The case for self-completion methods in the survey measurement of domestic abuse is overwhelming, and it is essential that self-completion delivery continues. However, different self-completion modes (for example, online or computer-assisted self-interviewing) may be used with the same questions, for example, to reduce non-response bias by offering respondents multiple ways of completing the survey.

Sequential sampling designs should be considered as most effective in minimising survey non-response errors while balancing costs

Sequential sampling designs may start with a cheaper delivery mode (for example, online) and then offer alternative self-completion delivery mode(s), preferably of the respondent’s choice. Improving data quality by reducing non-response error seems important in view of what are likely to be relatively small mode-based variations in survey response between different self-completion methods.

Consideration should be given to the feasibility of a move to long-term online delivery as the preferred primary data collection mode for domestic abuse data

Online surveys share the same advantages as other self-completion methods in surveying sensitive topics and may offer additional benefits which can reduce total survey error. However, web-surveys can also have important drawbacks including coverage and unit non-response errors, reduced opportunities for interviewer input, and perhaps reduced respondent engagement (for example, distraction behaviours and question skipping). Additionally, there are ethical and safeguarding considerations associated with a move to online data collection including assessing the risks to the physical, emotional and psychological well-being of respondents. Our October 2020 publication highlights findings from qualitative research on the ethics of online data collection relating to sensitive topics.

Research outcomes

From the research work undertaken by the University of Bristol, alongside wider engagement with stakeholders, we have made the following initial decisions to:

• continue to ask respondents about victimisation both since the age of 16 years and in the past 12 months by a partner, ex-partner, and family member
• take forward work to develop an accurate measure of the prevalence of domestic abuse including developing a way to measure controlling or coercive behaviour within the overall prevalence measure
• look to develop a measure of the impact of abuse
• investigate ways to measure the number of incidents and frequency of abuse
• proceed with developing a more complete measure of domestic abuse as a priority over preservation of the time series
• drop questions on attitudes and perceptions of domestic abuse and abuse experienced by others from the self-completion modules of the CSEW from October 2021 when face-to-face data collection resumed
• continue to explore the risks and issues associated with different survey modes

Notes for: Full review of the questions

1. The University of Bristol team were granted funding from the Oak Foundation (grant number OFIL-20-205) which was also used for the research to support the work.

As part of the next stage of the redevelopment we will produce a new set of qualitatively tested survey questions on domestic abuse as well as continuing to further explore the survey mode. In October 2021 we awarded a research contract to a consortium led by the University of Bristol to take this work forward over the next six months. The main aims of this research are:

• to understand how questions should be asked to provide the information users need, with a particular focus on measuring controlling or coercive behaviour, impact of abuse and frequency of abuse
• to develop and qualitatively test survey questions with victims and survivors of domestic abuse and the general public
• to investigate the use of alternative survey modes and further examine the options and associated issues through qualitative research with victims, survivors, the general public, and victim services and support providers

We aim to complete this stage of the research by Spring 2022 and will provide a further update on progress alongside the annual crime publication in Summer 2022.

For further information on the current collection of data on domestic abuse see the Domestic abuse quality and methodology information report.