Researching Trauma-Informed Social Media Without Compromising Participant Safety

The Challenge: Learning Without Increasing Risk

This academic HCI study, conducted and co-authored with Prof. Maria Wolters, examined how social media could respond more responsibly to people experiencing collective trauma. The central research challenge was not only what to learn, but how to learn it without creating an additional record of exposure for the people contributing their experiences.

The guiding question was: How can we collect useful evidence in a trauma-sensitive setting while asking for, moving and retaining as little participant data as possible?

An Exploratory Mixed-Methods Approach

I combined a bilingual survey with semi-structured interviews. The survey provided directional patterns across interaction, emotional and safety needs, while interviews added context about how social-media experiences and coping behaviours were understood by participants.

Open responses and interview material were analysed through iterative thematic coding, with codes revisited across multiple sessions. I treated the synthesis as a set of information-rich design signals rather than statistically representative conclusions.

The themes informed low-fidelity concepts. Because direct testing could have created disproportionate risk, I used critical expert walkthroughs against trauma-informed design and digital-wellbeing principles instead of in-situ evaluation with the original participant group.

Privacy Was Part of the Method, Not a Final Check

I used a harm-minimisation approach informed by HCI research with at-risk populations. Each decision was evaluated not only for research value, but also for the trace it could leave behind.

• Recruitment was network-gated through trusted intermediaries rather than conducted through open public calls.
• Participants selected the end-to-end encrypted communication channel used for research contact.
• No direct identifiers were collected, and no re-identifying data were retained.
• The instruments were limited to information necessary for the research questions.
• Reporting remained aggregated and non-attributable; raw responses and participant quotations are not reproduced here.
• Potential research value did not override safety: higher-risk in-situ validation was deliberately deferred.

These choices do not amount to a claim of zero risk. They document how exposure was actively reduced across recruitment, communication, collection, analysis and publication.

What This Changed in the Design Process

The study treated trauma-informed practice as a system-level HCI constraint rather than a tone-of-voice layer. Safety shaped who could be approached, which channels were acceptable, what was collected, how claims were framed and which forms of validation were responsible.

This moved the design process away from extracting maximum detail and toward restoring agency: clear system states, low-stimulation interactions, user-controlled support and protective choices were explored as connected parts of the experience.

Restraint Is a Research Outcome

The most important learning was methodological: rigorous research does not always mean collecting more or validating every idea immediately. In sensitive contexts, credibility also comes from narrowing claims, preserving uncertainty and leaving questions unanswered when answering them would place a disproportionate burden on participants.