Participant Information Sheet (PIS)
English Draft v0.1
Dialogical AI for Trauma Recovery Pilot Study
Study Title
Dialogical AI for Trauma Recovery: Exploring a Trauma-Informed Conversational AI Prototype
Invitation to Participate
You are invited to take part in a research study exploring the use of a conversational AI system designed to support grounding, emotional regulation, reflection, and wellbeing in people who may have experienced stress or trauma.
Before deciding whether to participate, it is important that you understand:
why the research is being conducted,
what participation involves,
and any possible risks or limitations.
Please read the following information carefully. You are welcome to ask questions at any time before deciding whether to participate.
Participation is entirely voluntary.
What Is the Purpose of the Study?
This study explores whether a trauma-informed conversational AI system may provide a supportive and emotionally manageable interaction experience for users.
The project focuses particularly on:
emotional pacing,
grounding and regulation,
conversational safety,
bilingual interaction (English/Ukrainian),
and how users experience different conversational styles or “personas.”
The chatbot is designed around a “trajectory-first” approach. This means the system attempts to prioritise emotional safety, stabilisation, and appropriate pacing over deep exploration or analysis.
The study is exploratory and experimental. The chatbot is a research prototype and is not intended to replace professional mental health care.
Why Have I Been Invited?
You are being invited because you are:
aged 18 or over,
able to provide informed consent,
and interested in participating in research involving conversational AI and emotional wellbeing.
The study may include participants with experiences of stress, trauma, military service, displacement, or emotionally difficult life experiences, but participants do not need to disclose personal trauma histories in order to participate.
What Will Participation Involve?
If you decide to participate, you will be asked to:
Read this information sheet and complete a consent form.
Complete a short pre-session questionnaire.
Interact with the chatbot for approximately 15–30 minutes.
Complete a short feedback questionnaire after the interaction.
Optionally participate in a short follow-up discussion or survey.
During the interaction, the chatbot may:
ask reflective or supportive questions,
guide short grounding or stress-reduction exercises,
respond in English or Ukrainian,
and adapt its conversational style based on emotional context.
You do not have to answer any question you do not wish to answer.
You may stop the interaction at any time without giving a reason.
Important Information About the Chatbot
The chatbot used in this study:
is an experimental AI research prototype,
is not a therapist, doctor, or crisis service,
does not provide medical diagnosis or clinical treatment,
does not provide a clinical intervention or clinical assessment,
and may occasionally misunderstand or respond imperfectly.
The system is designed to avoid encouraging harmful behaviour or emotionally overwhelming discussions. However, AI systems can sometimes produce responses that feel inappropriate, inaccurate, or emotionally unhelpful.
Participants are encouraged to use their own judgement and to stop the interaction if they become uncomfortable.
Possible Risks or Discomforts
Some participants may find that discussing emotions, stress, or difficult experiences causes temporary discomfort.
Potential risks may include:
emotional upset,
frustration,
misunderstanding by the chatbot,
or feeling emotionally tired after the interaction.
The study is designed to minimise these risks through:
careful pacing,
grounding-oriented interaction,
and clear safety boundaries.
The chatbot is not designed for crisis intervention.
If you are currently in severe emotional distress or require urgent support, you are advised not to participate and instead seek support from a qualified professional or crisis service.
Safeguards and Support
Participants may pause or stop participation at any time.
At the end of the session, participants will receive:
grounding suggestions,
wellbeing resources,
and relevant support information where appropriate.
If the chatbot detects language associated with immediate risk or crisis, it may encourage the participant to seek human support or contact appropriate services.
Confidentiality and Data Handling
The research team will collect and store:
chatbot conversation transcripts,
questionnaire responses,
and limited technical interaction data.
In the current deployed prototype, generated text-to-speech audio is streamed to the browser and is not retained by the application. If you use speech input, audio is processed to produce the text you review and submit; the formal pilot does not use speech-form analytics or retain derived speech features. The submitted text becomes part of the conversation transcript in the same way as typed text.
The browser keeps only an opaque session identifier in a secure cookie. The deployed research session expires after 12 hours. This browser session is separate from the pseudonymous research record: session expiry does not itself delete conversation transcripts or participant-bound records.
You will receive a random 12-digit participant code and a separate access PIN. On first use, you choose your own language and chatbot companion. The code does not assign either choice or place you into a study condition. On a private device, you may choose to remain signed in using a revocable device token. On a shared device, you will be asked for the code and PIN again. Keep both values private.
The research team keeps the issued-code and PIN register separately from the app. The app stores keyed fingerprints rather than the raw values. Private-device tokens are also stored only as keyed hashes and can be revoked. This is pseudonymisation, not anonymisation, and the linked research record remains protected research data.
Data will be anonymised or pseudonymised where possible.
Participants should avoid sharing:
full names,
addresses,
phone numbers,
military identification numbers or callsigns,
unit, rank, deployment-location, or operational details,
or other identifying personal information during chatbot conversations.
The chatbot is designed not to carry those military details into its continuity memory. If entered, however, they may still appear in the research conversation transcript and be handled under the study's access, retention, withdrawal, and deletion procedures.
Research data may be reviewed by members of the research team for analysis purposes.
Anonymised findings or excerpts may later be used in:
academic publications,
presentations,
or future research outputs.
No personally identifying information will be intentionally published.
Voluntary Participation and Withdrawal
Participation is voluntary.
You may:
decline to answer any question,
pause participation,
or withdraw from the study at any time without penalty.
If you withdraw, you may request removal of your data where reasonably possible and where analysis has not already been completed.
What Happens Next?
This pilot study will help researchers evaluate:
usability,
emotional safety,
conversational pacing,
and participant experiences of trauma-informed AI interaction.
The findings may contribute to future research into ethical, culturally sensitive, and trauma-informed AI systems.
Contact Information
Researcher:
Dominic Sweeting
frontierpsychology@outlook.com