Feasibility and acceptability of the MentiParent AI chatbot for training parental reflective functioning

Area of Science:

• Psychology
• Artificial Intelligence
• Human-Computer Interaction

Background:

• Traditional interventions for Parental Reflective Functioning (PRF) have limitations.
• There is a need for scalable and accessible tools to support reflective parenting.
• Generative AI offers potential for novel therapeutic and educational applications.

Purpose of the Study:

• To assess the feasibility and acceptability of MentiParent, an AI-driven chatbot designed to enhance PRF.
• To explore the potential of AI in simulating parent-child interactions for reflective parenting support.
• To gather qualitative feedback on user experience and identify areas for improvement.

Main Methods:

• Two pilot studies were conducted involving mental health practitioners and general population parents.
• Participants interacted with MentiParent, simulating conversations with virtual children.
• Usability, relevance, feasibility, and acceptability were evaluated using quantitative scales and qualitative feedback.

Main Results:

• Both practitioners and parents rated MentiParent highly for feasibility and acceptability.
• Qualitative feedback indicated engaging and realistic interactions, with valued feedback provided by the chatbot.
• Identified limitations included authenticity and technical fluidity, alongside emerging ethical considerations.

Conclusions:

• MentiParent demonstrates strong feasibility and acceptability among both professionals and parents.
• The AI chatbot shows potential as a scalable and ethically grounded tool for enhancing reflective parenting skills.
• Further development is warranted to address technical limitations and ensure ethical implementation.