Tailoring Persuasive and Behaviour Change Systems Based on Stages of Change and Motivation

Document Title

Tailoring Persuasive and Behaviour Change Systems Based on Stages of Change and Motivation

Document Information

• Subject Area: Human-Computer Interaction and Behavior Design, specifically the design of change and health promotion systems
• Keywords: Persuasive systems, behavior change systems, stages of change, persuasive strategies, behavior change, motivation, customization, personalization, behavior change theory, health and health risk behaviors, design implications

Research Background and Issues

• Identified Problems or Challenges:

  • Persuasive systems (PS) have been proven effective in motivating behavior change, but most systems adopt a "one-size-fits-all" approach, lacking personalized designs tailored to different user variables (e.g., users' stages of behavior change).
  • Few studies have explored how to adapt persuasive systems based on the "stages of behavior change."
  • The motivational effects and mechanisms of persuasive strategies on different users remain unclear.

• Significance of the Research:

  • Personalized design can enhance the efficiency of persuasive systems, better meet user needs, and increase the likelihood of achieving health behavior goals.

• Research Motivation and Related Work:

  • Using behavior change theories (e.g., Transtheoretical Model (TTM)) can guide the design of stage-specific interventions to support users' behavior change.
  • Current research on persuasive systems mainly focuses on users' demographic characteristics or psychological variables (e.g., personality types), with little attention to the relationship between stages of behavior change and persuasive strategies.

Proposed Solution

• Proposed Solution:

  • Develop a model for designing persuasive systems by integrating the stages of change theory from the Transtheoretical Model with the ARCS motivation model and the Persuasive System Design (PSD) framework.
  • Conduct a large-scale user study (568 participants) to validate the perceived persuasiveness and motivational mechanisms of different persuasive strategies across various stages of behavior change.

• Innovative Contributions:

  • For the first time, combine behavior change theory (TTM) with motivation theory (ARCS) and persuasive system design to propose stage-specific personalized design guidelines.
  • Quantify the perceived effectiveness of five core persuasive strategies through user research and provide qualitative feedback on user motivation.

• Implementation Steps:

  1. Select five commonly used strategies in persuasive systems (self-monitoring, reminders, suggestions, social roles, and praise).
  2. Design high-fidelity prototypes to test these strategies and collect quantitative data on users' perceived persuasiveness and motivational effects.
  3. Use structural equation modeling (PLS-SEM) to analyze the relationships between strategies, motivational dimensions (attention, relevance, confidence, and satisfaction), and stages of behavior change.
  4. Conduct thematic analysis of qualitative comments to support and expand quantitative findings.

Research Findings

• Specific Findings:

  • Different user groups show significantly different preferences for persuasive strategies:
    • Users in the pre-contemplation stage are more sensitive to strategies that raise awareness and consciousness (e.g., self-monitoring).
    • Users in the action stage prefer strategies that help substitute problem behaviors (e.g., suggestions), provide reminders (e.g., reminders), and offer social support roles (e.g., social roles).
  • Praise strategies are relatively universal in boosting users' confidence and timeliness.
  • Significant differences in users' perceptions of persuasive strategies across behavior change stages were identified, leading to data-supported personalized design recommendations (see "Design Guidelines" below).

• Comparison with Existing Solutions and Advantages:

  • Addressing the limitations of "one-size-fits-all" designs, this study proposes a personalized design approach guided by stages of behavior change.
  • Provides a set of theoretically supported, highly targeted design recommendations.

• Experimental or Evaluation Results:

  • Modeling and data analysis indicate that motivational dimensions (particularly "attention," "confidence," and "relevance") significantly influence users' acceptance of strategies at different stages.
  • Both quantitative and qualitative user feedback validated the alignment and applicability of strategies with the motivation model.

• Limitations and Future Directions:

  • This study relies on self-reported user data, which may not fully reflect actual behavior change outcomes.
  • The current research focuses on smoking behavior change interventions; further validation is needed for other health behaviors (e.g., physical activity).
  • Future work includes real-world system testing and cross-domain validation of the findings.

Additional Design Recommendations and Implications

• Broadening User Base Design:

  • Systems should integrate universally effective strategies, such as "suggestion" strategies, to enhance user confidence and increase the likelihood of behavior change.
  • Provide customizable reminder functions, allowing users to adjust reminder frequency based on their needs.

• Stage-Specific Design Guidelines:

  • Pre-contemplation Stage: Focus on self-monitoring strategies to enhance users' awareness and consciousness.
  • Contemplation Stage: Recommend combining self-monitoring, reminders, suggestions, and social roles to help users analyze risks and benefits.
  • Preparation Stage: Emphasize praise and reminder strategies to encourage gradual adoption of healthy behaviors.
  • Action Stage: Use dynamic support (e.g., suggestions, social roles) to consolidate behavior change.
  • Maintenance Stage: Employ continuous reminders and alternative suggestions to help users avoid relapse.

• Social and AI Interaction:

  • In applying social roles, increase the ability to switch between human-machine modes, allowing users to interact with real experts as needed, thereby enhancing credibility and user experience.