Trauma-informed reporting on gender-based violence rests on a single principle: the survivor retains control over their own account, before, during and after publication. Everything else — how you approach, how you interview, what you publish, what you do afterwards — follows from that.
This is not a softer standard than conventional journalism. It is a stricter one. It requires more time, more preparation and a willingness to lose stories you cannot obtain safely.
Consent is a process, not a signature
A consent form obtained at the start of an interview tells you almost nothing about whether someone understands what publication will mean for them six months later.
Treat consent as something that is established, revisited and can be withdrawn. In practice:
- Explain distribution concretely. Not "this may be published online" but "this will be on a website anyone can search, including people in your community, and it will stay there."
- Name the risks you can foresee — identification through detail, family reaction, employer reaction, legal exposure — and say plainly that you cannot foresee all of them.
- Re-confirm before publication, once the piece exists and the person can see what they actually agreed to.
- Honour withdrawal, including after publication where removal is technically possible. Agree in advance what you will do if asked.
The Dart Center for Journalism and Trauma publishes detailed guidance on interviewing in these conditions and is the most useful single resource for reporters working this beat.
Interviewing about gender-based violence: design around control
Trauma responses are, in large part, responses to loss of control. An interview that removes control re-enacts the harm. An interview that returns it does not.
The practical adjustments are unglamorous and effective. Let the person choose the location and who else is present. Tell them at the outset that they can stop, skip a question, or end the conversation entirely with no consequence to their access to services. Ask permission before moving to a difficult area rather than sliding into it. Avoid asking for a chronological account of the incident itself unless it is genuinely necessary — narrative reconstruction under pressure is where most re-traumatisation happens.
Ask about aftermath rather than event where you can. What has changed since. What support existed or did not. What they want to happen now. This usually produces better journalism as well as safer interviews, because the aftermath is where the systemic failures live.
Anonymity is a technical problem, not a promise
Changing a name is not anonymisation. Communities identify people through detail: the location of a workplace, the number and ages of children, a distinctive job, the timing of an event, a photograph of hands or a doorway.
Work backwards. Assume a reader who lives in the same neighbourhood. Ask what combination of published details would let that reader identify the person, then remove enough of them that the answer is none. This will sometimes cost you specificity that would have made the piece stronger. Pay it.
Be equally careful with images and audio. Voice is identifying. So are room interiors, uniforms and visible landmarks.
Know what you are referring people to
Do not conduct an interview about violence without knowing, specifically, what support exists locally and whether it functions. A referral to a service that has no staff, charges fees, or requires travel the person cannot afford is not a referral.
Verify before you need it: what exists, what it costs, what hours it operates, whether it is safe for this particular person to be seen entering. UN Women and the World Health Organization both publish frameworks for what adequate response services look like, which is useful for judging whether what exists locally is adequate.
Aftercare, including your own
Two things get skipped almost universally: following up with the person after publication, and attending to the interviewer.
Follow-up need not be elaborate. A message a week after publication, asking whether anything has changed and whether they need anything, closes a loop that would otherwise leave someone exposed and alone with the consequences of a decision you asked them to make.
For the reporter or communications officer: repeated exposure to accounts of violence has a cumulative effect that is well documented and routinely ignored in small organisations. Build in supervision, rotate the beat where possible, and treat this as an operational requirement rather than a personal weakness.
Common questions
Should survivors ever be named?
Only where the person has asked to be named, understands the specific risks, and there is no coercion — including the subtle coercion of being told their story will have more impact. Default to protection.
Is it ethical to pay for interviews?
Payment for testimony creates an incentive to disclose, which compromises consent. Reimbursing genuine costs — transport, childcare, lost earnings — is a different matter and is usually appropriate.
What if the story is in the public interest and the survivor declines?
Then you do not have that story. Public interest justifies intrusion on the powerful, not on the person who was harmed.
The same tension between narrative impact and the dignity of the subject runs through why humanitarian storytelling keeps failing its subjects, and the protocols overlap considerably with building a crisis communications plan that survives contact.