Welcome!
Background
Large Language Models (LLMs), like ChatGPT, are generative AI tools that mimic human conversation abilities. These models present key opportunities to advance training and education in suicide prevention and mental health applications.
We aim to develop a language model capable of role-playing as an individual in suicidal distress, for deployment with suicide prevention gatekeeper trainees. Such a model would enable trainees to practice engaging with those in distress, building valuable skills and confidence.
In this specific study we have developed a text based interface & model (similar to ChatGPT) capable of crisis conversation simulation. That is, the model is trained to simulate an individual in potential suicidal distress, to simulate a situation where a trained gatekeeper may be needed. In this study we aim to test the capability of this novel approach, and gather valuable feedback from area experts, such as yourself.
Participation Requirements
If you choose to participate, we will ask your name, age, sex, and information relating to your gatekeeper training experience. You will then be granted access to the virtual simulation where we will ask you to act in the role of a trainee gatekeeper to interact with the simulated individual in distress via text interface. There are three separate scenarios, each lasting a maximum of 5 minutes. After each scenario we will ask your thoughts on the performance of the model, and at the conclusion of all three scenarios we will ask for your thoughts on the wider aspects of the training in general. At the conclusion you will also have the opportunity to provide free form comments if you wish.
We expect your total time spent on this study to be between 20 and 30 minutes. Once you've finished, you will be asked to enter your details to receive your gift voucher. If you have any questions, please contact the researcher listed on the participant information sheet.
It is recommended you complete this evaluation on a desktop/laptop computer as the presentation of the simulation interface and evaluation components are not optimised for mobile devices. If you wish to continue, please do so now.
Thank you for your participation.
By clicking "I agree" below:
- I understand I am being asked to provide consent to participate in this research study;
- I have read the Participant Information Sheet, or it has been provided to me in a language that I understand;
- I consent to the use of information collected about me for this research, as well as for future studies that extend from or are closely related to this project, or those within the same general field of research;
- I understand that, if necessary, I can ask questions, and the research team will respond to my questions;
- I freely agree to participate in this research study as described and understand that I am free to withdraw at any time during the study and withdrawal will not affect my relationship with any of the named organisations and/or research team members;
- I understand that I can download a copy of this consent from from this link.