Fostering reflection: Development and initial evaluation of a mentalization-based GenAI simulator for foster care...

Area of Science:

• Artificial Intelligence in Social Work
• Human-Computer Interaction
• Professional Development

Background:

• Foster care supervisors need advanced reflective functioning (RF) for complex demands.
• Traditional training methods for RF are difficult to scale and access.
• Generative AI (GenAI) offers potential for simulated practice environments.

Purpose of the Study:

• To develop a GenAI-based chatbot simulator for enhancing RF in foster care supervisors.
• To evaluate the initial usability and acceptability of this novel training tool.

Main Methods:

• A mixed-methods design was employed with 46 female foster care supervisors.
• Participants engaged in a single session with a GenAI simulator featuring interactive dialogue and feedback.
• User experience was assessed using quantitative ratings and qualitative thematic analysis.

Main Results:

• The GenAI simulator demonstrated high usability and acceptability.
• Feedback was rated as highly contributory (mean 5.83/7), and willingness to recommend was strong (mean 6.67/7).
• Qualitative data indicated the simulator served as a safe space for reflection, with users desiring both supportive and challenging interactions.

Conclusions:

• The GenAI-based simulator is highly acceptable for RF training among foster care supervisors.
• It may complement traditional training by offering a scalable, accessible, and safe practice environment.
• This technology shows promise for improving supervisory skills in foster care settings.