A Suite of Mobile Conversational Agents for Daily Stress Management (Popbots): Mixed Methods Exploratory Study

Area of Science:

• Mental Health Technology
• Human-Computer Interaction
• Digital Health

Background:

• A significant gap exists in providing stress management support during primary care visits, despite the high prevalence of psychological stress.
• Advances in natural language processing have spurred interest in mental health chatbots for public health.
• Developing effective conversational agents for stress mitigation faces challenges in complexity and cost.

Purpose of the Study:

• To develop and evaluate a mobile suite of automated shallow chatbots, termed Popbots, for stress management.
• To explore Popbots as a novel type of chatbot complementing human support in stress management.
• To assess the feasibility and effectiveness of a multi-chatbot system for daily stress mitigation.

Main Methods:

• An initial Wizard of Oz study (N=14) assessed the feasibility of multiple chatbots.
• A web-based study (N=47) evaluated the prototype implementation, assigning participants to chatbots based on cognitive or behavioral interventions.
• Stress levels were measured using self-reported psychometric evaluations, including daily surveys and the Patient Health Questionnaire-4.

Main Results:

• 66% of participants completed the study, viewing chatbot conversations as helpful or neutral.
• Users reported increasingly positive sentiment towards using chatbots for proactive stress management.
• Higher engagement with the chatbots (median conversations or more) was linked to a significant decrease in depression symptoms (P=.01).

Conclusions:

• Suites of shallow chatbots offer potential benefits for both users and developers in stress management.
• The study provides design guidelines and identifies challenges for multi-chatbot systems in conversational interventions.
• Future research directions include AI-enabled recommendation algorithms and authoring systems for chatbot development.