Supporting mental health self-care discovery through a chatbot

Area of Science:

• Human-Computer Interaction
• Digital Mental Health
• Artificial Intelligence in Healthcare

Background:

• Mental health and wellbeing are crucial, with self-care complementing clinical treatments.
• Conversational agents (chatbots) offer a novel approach to delivering self-care information.
• Understanding user trust in chatbots for sensitive health information is vital.

Purpose of the Study:

• To investigate the use and perceived trust of conversational agents (chatbots) for crowdsourced self-care.
• To compare user trust in a chatbot interface versus a conventional web interface for accessing self-care methods.

Main Methods:

• A between-subjects study with 80 participants was conducted.
• One group used a standalone web interface; the other used the same system integrated into a chatbot interface.
• Quantitative data was collected, supplemented by qualitative analysis.

Main Results:

• User trust was significantly influenced by perceptions of system security and integrity.
• The chatbot interface received lower trust scores compared to the conventional web interface.
• Qualitative data provided insights into the reasons behind trust discrepancies.

Conclusions:

• Security and integrity are critical factors influencing user trust in chatbots for mental health self-care.
• Further research is needed to address trust concerns and optimize chatbot design for mental health applications.
• Chatbot interfaces require careful design to foster user trust in digital mental health interventions.