AI Tech
Introduction
This document captures the comprehensive discussion on the development and implementation of an AI chatbot designed to assist nursing students in debriefing virtual simulations. The project involves collaboration between a PhD student from Brazil and educators at Queen's University and employs the design thinking methodology to refine the chatbot's functionality.
Background Context
Initial Encounter with AI
The speaker discusses their unexpected involvement in AI technologies, specifically in virtual simulations, highlighting the evolution of these technologies in educational settings.
Collaboration Overview
PhD Student Approach
About a year ago, a PhD student named Agostino from Brazil approached the speaker to collaborate on a simulation project.
Project Focus
Agostino expressed his interests in AI chatbots and showcased a chatbot he developed in Finland, specifically designed to support nursing students experiencing stress. This chatbot provided advice on time management to help reduce stress levels.
Development of the AI Chatbot
Project Concept
The speaker proposed creating a chatbot to assist in debriefing virtual simulations, which Agostino agreed to. This guidance led to a formal project during Agostino's visit to Queen's University, which started in September and concluded shortly before the presentation.
Methodology
Design Thinking Process
Agostino utilized a developmental methodology known as design thinking, which consists of several key steps:
Empathize: Understanding user experiences and needs in the context of debriefing virtual simulations. The focus was on addressing challenges related to debriefing virtual versus in-person simulations.
Define: Reviewing literature to identify known issues and possible solutions to standardize effective debriefing methods for virtual simulations.
Ideate: Brainstorming feasible approaches leveraging expertise from Brazil and Canada to develop the chatbot.
Prototype Development: Creating various iterations of the chatbot, starting from a text-based model evolving into a conversational voice-based format.
Test: Conducting informal pilot testing to gather user feedback on the chatbot's performance.
Empathize Phase
User Research
During the empathize phase, interviews were conducted with simulation experts, including the speaker, colleagues, and learners who experienced simulation debriefings. This aimed to gather insights into the debriefing challenges faced with virtual simulations.
Kansim Involvement
The speaker is involved with Kansim, a nonprofit organization that supports nurse educators in simulation. Kansim offers resources, including virtual simulations that are accessible online, addressing the need for effective debriefs.
Defining the Problem
Healthcare Simulation Standards
According to healthcare simulation standards, all simulations must be debriefed effectively. However, the implementation of debriefing with virtual simulations has been inconsistent, often lacking in quality due to insufficient educator training or oversight.
Current Feedback Mechanism
The existing virtual simulations provide embedded feedback on decisions made during the simulation, yet this feedback merely represents one component of a comprehensive debriefing. This feedback informs students whether their responses are correct or incorrect, coupled with rationale for the choices made.
Importance of Reflection
Reflection is a critical aspect that allows students to evaluate their decision-making process and consider future adjustments. The project aims to improve this reflective process through AI-led guidance.
Ideation and Prototyping
Development Phases
Chatbot Naming
The chatbot was named "DeBry," a combination of "debriefing" and "AI," symbolizing its primary function.
Simulation Selection
The initial project focused on the de-escalation of an angry patient simulation, which was one of the first virtual simulations created by Kansim and remains popular among users.
Prototype Functionality
The prototype required predetermined inputs, scripts for the simulation, and decision points to ensure accurate and helpful responses from DeBry. Critical components included:
Debriefing Frameworks: Two frameworks were chosen: the GAS method and Debriefing with Good Judgment. These structures prioritized a supportive approach to discussion.
Communication Guidelines: Clear behavioral guidelines were essential, ensuring the chatbot’s tone remained empathetic and non-judgmental. DeBry was designed to facilitate discussions rather than evaluate students' performances.
Contextual Knowledge: Specific de-escalation guidelines were integrated to ensure that the chatbot conveyed appropriate responses during debriefing sessions.
Time Management: The session duration was set to encourage effective engagement without being overly time-consuming.
Testing and Feedback
Pilot Testing Insights
Pilot testing involved faculty and student research assistants who assessed the chatbot's communication style and effectiveness in promoting reflection. Participants described DeBry's voice and relational tone as approachable and supportive, emphasizing the importance of these traits during sensitive discussions related to clinical decision-making.
Reflection Prompts
Test participants noted that the prompts used in the chatbot effectively structured guided reflection. They suggested that a 5-10 minute debrief was generally sufficient, as this shorter duration aligns with individualizing the experience in virtual simulations versus group settings.
Future Directions
Usability and Effectiveness Evaluation
A comprehensive usability and effectiveness study will be conducted to assess how well DeBry supports and enhances the debriefing process. Feedback from this study will inform future improvements.
Expansion of Chatbot Functionality
Plans are underway to explore the development of additional chatbots applicable to other virtual simulations, expanding the scope of the AI's utility and availability on the associated educational platforms.
Conclusion
This innovation signifies a promising step towards optimizing debriefing methodologies in nursing education through AI, emphasizing the pursuit of enhanced reflective practices without replacing the essential role of educators. The conversation initiated by this project opens avenues for future dialogue and development in AI-based educational tools.