In this webinar, understand how PhD student Benjamin Daniels and his research team at the Harvard T.H. Chan School of Public Health used the SurveyCTO LLM Conversations field plug-in to come up with a more cost-effective way to conduct clinical vignettes.
Clinical vignettes are essentially open-ended interviews and while they are crucial to measuring the competencies of medical professionals, they can also be cost prohibitive—particularly in resource-constrained environments around the world.
Watch this recording to understand how Benjamin commissioned the SurveyCTO team to create this field plug-in, how his team modified it to suit their needs, and see it in action!
Key webinar takeaways
- How you can scale the way you measure healthcare provider competencies while keeping costs low
- How the SurveyCTO LLM Conversations field plug-in was customized for this specific healthcare use case
- A live demonstration of the context-aware chatbot and the respondent experience
How the research team verified that AI conversations met the quality standards of human experts - Ideas for adapting this plug-in for broader applications in education, market research, and beyond.
Time stamps
- 0:15 – Overview of SurveyCTO field plug-ins
- 3:10 – Background on Benjamin’s research project that prompted the need for the LLM Conversations field plug-in
- 7:18 – Why other tools didn’t work for creating a chat-based evaluation tool
- 8:10 – Overview of the LLM Conversations field plug-in
- 16:51 – Live demo of the LLM Conversations field plug-in
- 23:31 – Walkthrough of how to provide the AI with the necessary prompt guardrails
- 32:00 – How Benjamin and his team validated the outputs of the AI-powered chatbot evaluations to ensure their quality
Webinar resources
New to SurveyCTO? Sign up for a free trial today to explore our solution and make use of our field plug-ins for a better survey experience. During the 15-day trial, you’ll have access to the entire platform.
If you have general questions or questions about our field plug-ins, our support team is available 24×7. Go to the Support Center and click on the link to “Submit a support request.” No question is too small!
