DSGBV Working Group - Meeting 3

Minutes of Meeting

1. Introduction 

• Donal Kelly (Chair, CSO) opened the meeting, welcomed members, and thanked them for their time. He expressed appreciation for the suggestions provided at the previous meeting, while noting that due to the scale of the questionnaire, it may not be possible to incorporate every proposal. He also emphasised that the CSO must be careful with the structure of the survey to ensure it remains aligned with its design requirements and continues to collect all data mandated by Eurostat. He advised that following today’s meeting, a final report will be prepared and circulated to the group. He also requested the group’s participation in a liaison subgroup, which will be established to review the evolving questionnaire and provide further support, particularly regarding the language used throughout the survey.
• The minutes of the previous meeting were approved.
• Jessica Coyne (CSO) highlighted an error in the minutes from the first meeting. She noted the minutes incorrectly stated that ‘data provided to the CSO will be fully anonymised’ instead of ‘pseudo anonymised’ which reflects the standard practice for all data collected by the CSO. This amendment was agreed by the group and the minutes of the first meeting will be updated accordingly.

2. Review of points raised at the previous meeting

Types of behaviours and omissions –

• The CSO thanked the group for all their previous suggestions around wording and abusive behaviours, which have proven very valuable in considering the questionnaire design. These include additional behaviour types such as ‘Denying access to healthcare’ and ‘Abuse of pets’ under the relevant themes.

Language consideration –

• The CSO noted there was a substantial amount of work to be done on the language used in the survey and that previous suggestions from the group have been advanced with the Questionnaire Design team. It was noted that there will be continual engagement with the liaison group for advice on language and terminology through the questionnaire design phase and that regular feedback will be sought throughout the process. The language of the questionnaire may also change following the results of the focus group, cognitive and pilot testing.

Location DSGBV was encountered –

• The CSO advised they are exploring the possibility of including a location question in the ‘Sexual harassment’, ‘Non-partner’, and ‘Stalking’ sections of the questionnaire to see if this can be captured and queried whether it would be valuable to capture it in the ‘Former/Previous Partner’ section. The difficulty in capturing significant detail from just one question was discussed. The group again emphasised the value in collecting this detail for both partners and non-partners, highlighting that it could help to shape policies or interventions depending on the results, for example in workplace settings.

Measuring impact including psychological and functional impacts –

• Again, the CSO thanked the group for the valuable suggestions previously made and will broaden the types of psychological impact captured under the relevant themes. It was also agreed to review the possibility of separating the mental health themes across questions and to consider distinguishing between short‑term and long‑term psychological consequences.

2. The use of tech-facilitated abuse in relation to DSGBV -

• The CSO explained that, following further discussions and investigation, incorporating tech‑facilitated abuse throughout the questionnaire, as previously suggested by the group, creates a technical issue for returning data to Eurostat and would compromise the integrity of Eurostat’s output variables. To address this, the CSO proposed including tech‑facilitated abuse as a separate section, following the same structure as the existing sections. They highlighted that this approach would allow for the inclusion of additional questions and a wider range of behaviours. However, they also noted that some questions may still be more appropriately placed under other pillars.
• Further considerations in the context of tech-facilitated abuse and violence noted by the group were –
  • Whether someone took an image without consent
  • Situations where a victim may be coerced into sending images or content
  • The need to capture continuous harassment through messages
  • Potential pattern indicator question, such as- ‘Is the content still online?’, ‘It was removed after X months’
  • The use of AI image generation

3. Domestic perpetrators (outside of partner/former partner)

• There was a broad discussion about the inclusion of more domestic perpetrators in the survey, beyond those already included, such as ‘landlord’. The group noted that a landlord could also be a family member and some concerns were raised about whether this issue could be effectively measured by the survey. As discussed in the previous meeting the CSO confirmed they are looking at included behaviours incited by former partners.
• Additional points noted -
  • Important to capture the relationship between the victim and perpetrator
  • Merit to including ‘Sibling’ and ‘Adult child to parent’ abuse
  • Possible question – ‘Based on experience/impact has it impacted your access to housing and security?’
  • Consider rewording ‘Did you witness…’ to ‘Were you present when…’ recognising that someone can be impacted without directly witnessing abuse
• The CSO agreed to look further into the current list of relationships with all categorisations to be reviewed with the Questionnaire Design team.

4. Pattern Indicator

• The CSO noted that a significant body of work was required to review suggestions from the previous meeting regarding the language used to capture frequency, using terms such as ‘repeatedly’ or ‘continuously’ rather than using specific intervals but confirmed that they are open to revising the classifications once they can establish that any changes can be appropriately mapped to Eurostat classifications.
• The group also discussed the challenges in measuring or defining patterns of tech‑facilitated abuse. Despite these complexities, there was agreement that, given the rapidly evolving nature of this area, it is important to view this survey as a foundational step in building a more comprehensive evidence base over time.

5. Additional Observations

• The group discussed the limitations of the current sampling frame, particularly the cohorts that are excluded when surveys are conducted only within private households. The CSO noted that this issue had also been highlighted and discussed during the SVS, and the group suggested that the rationale and explanation for these limitations should be made more transparent to the public. It was also proposed that this area could be considered for further work in the future. The CSO agreed with these points and thanked the group for their ideas and feedback.
• The group discussed whether impact‑related questions would only be collected when a behaviour had been experienced and emphasised the importance of having psychological well-being information for those with no experiences to establish a baseline for comparison. They noted that such data would provide crucial metrics to inform national strategies. In response, the CSO highlighted the already significant response burden on those who have no experiences of violence or abuse (around 100 questions). However, they agreed to explore the suggestion further.
• The group asked whether the results would indicate if one form of abuse had led to another, and whether patterns in the evolution of violence would be measured. The CSO explained that this would be very complex to determine and that, while the results will show the different types of abuse experienced from a current or former partner, they will not illustrate how these behaviours developed or changed over time. The CSO highlighted that more complex analysis could potentially be carried out after the initial publication using Research Microdata Files.

6. Next Steps

• A report compiling the group’s suggestions and outlining the extent to which these suggestions will be incorporated into the final survey design will be prepared and circulated in the coming weeks. The CSO reiterated their intention to incorporate as many of the group’s suggestions as appropriate, provided that the integrity of the Eurostat data is not affected and the burden on respondents is managed.
• The CSO presented an overview of the planned delivery timeline, outlining key milestones from pre‑testing in Q3 2026 through to the publication of national results in Q3/Q4 2028.
• The CSO highlighted the extensive work still required on the questionnaire. Once the questionnaire is finalised, it will undergo comprehensive cognitive testing, and the CSO requested the group’s support with this phase when the time comes.

7. Closing

• Donal Kelly (CSO) thanked the members of the working group once again for their engagement and valuable contributions throughout each meeting. He noted that once the questionnaire has progressed further through the design phase, it will be easier to have more informed and focused discussions on several of the points raised by the group. He again encouraged members to participate in the liaison group, where these discussions will take place. 
• This was the final meeting of the group, which is now formally concluded. No further meetings will take place.

8. Summary / Actions