Unhealthy Comparisons to Promote Healthy Behavior? Exploring the Impact of Social Comparison Strategies in Personal Informatics.

Research Background and Problem

• Identified Issues or Challenges

  1. Social comparison may lead to negative emotions (e.g., inferiority, disappointment), reduced self-confidence, and even trigger rumination and overtraining behaviors.
  2. In health and fitness applications, social comparison features (e.g., leaderboards, rankings, data sharing) are often used to motivate users but may negatively impact users' mental health.
  3. Current research on social comparison strategies primarily focuses on their effects on motivation and behavior change, with insufficient attention to users' emotional responses and mental health.
  4. It remains unclear how different implementations of social comparison affect users' emotional states and their understanding of health and fitness data.

• Why This Problem is Important
  Social comparison is pervasive in personal informatics systems, and understanding the impact of different social comparison strategies is key to ensuring users' long-term mental health. Without optimized design, these features may exacerbate users' psychological conditions and undermine the long-term effectiveness of these systems.

• Research Motivation and Related Work
  By reviewing social comparison theory and existing comparison strategies in health applications, the authors identified the potential of these strategies for motivation and reflection while highlighting their possible harms. Although prior studies have explored similar topics, there is still a lack of granular understanding of how social comparison interacts with different data types (e.g., step counts and body fat percentage) and user roles.

Proposed Solution

• Proposed Solution
  The authors employed a three-part mixed-methods study:

  1. Application Evaluation: Analyzing health and fitness applications on the market to summarize their social comparison mechanisms.
  2. Online Scenario Survey: Testing users' emotional responses to different social comparison strategies (e.g., leaderboards, data distribution charts, sharing features).
  3. Semi-structured Interviews: Exploring users' long-term perceptions of social comparison features and their associated mental health effects.

• Innovations

  1. Proposed a multidimensional analytical framework encompassing comparison strategies, comparison content (e.g., body fat and step counts), and comparison targets (better, similar, or worse).
  2. Refined the understanding of the sensitivity of health data in social comparison contexts (e.g., differences in emotional responses to step counts versus body fat).
  3. Proposed specific design guidelines to improve social comparison features in health applications, prioritizing users' mental health.

• Implementation Steps

  1. Preliminary Study: Application Evaluation
     • Analyzed social comparison strategies, including data sharing, leaderboards, and distributions, in 42 top health and fitness applications.
  2. Scenario Survey Design and Experimentation
     • Created virtual user scenarios simulating different comparison strategies, content, and targets, and collected users' emotional response data using scales such as PANAS.
  3. Interviews and Experience Exploration
     • Conducted in-depth user interviews to understand the long-term psychological and behavioral impacts of social comparison and the complex mechanisms of different implementations on mental states.

Research Findings

• Specific Findings

  1. Application Review: Identified four main social comparison strategies in fitness and health applications (e.g., leaderboards, average distributions, data visualization/sharing) and their specific implementations.
  2. Scenario Survey Findings:
     • Ranking Strategies (e.g., leaderboards) enhanced motivation in step count comparisons but significantly increased negative emotions (e.g., shame and frustration) in sensitive data contexts such as body fat.
     • Body fat percentage was more likely to trigger negative emotions than step counts, especially when users performed below average.
     • Regarding "comparison targets," comparing with "slightly worse-performing" individuals boosted positive emotions, while comparing with "better-performing" individuals significantly increased negative emotions.
  3. Interview Findings:
     • Social comparison may lead to long-term psychological impacts (e.g., anxiety, stress) and behavioral effects (e.g., injury or overtraining).
     • Users prefer comparisons with familiar and similar individuals, which can reduce stress and enhance motivation, whereas comparisons with strangers or significantly better-performing individuals often result in dissatisfaction or frustration.

• Advantages Over Existing Solutions
  Compared to studies focusing solely on motivation, this research provides a more comprehensive perspective by integrating users' emotional responses, social comparison mechanisms, and long-term health impacts. By combining quantitative methods (scenario surveys) and qualitative methods (interviews), it lays the foundation for designing more human-centered and health-oriented personal informatics systems.

• Experimental or Evaluation Results

  1. While leaderboards can enhance competitiveness and pride, they have significant negative effects on vulnerable users (e.g., those with higher body fat) and should be appropriately limited.
  2. Sensitive data like body fat should avoid highly competitive strategies and instead use less pressuring visualization methods (e.g., distribution charts).
  3. Enhancing users' self-reflection awareness and providing control over social comparison features can reduce potential threats to mental health.

• Limitations and Future Directions

  1. Individual Differences: This study did not deeply control for user traits (e.g., motivation, personality characteristics) that may influence emotional responses. Future research could incorporate personalized factors to design more granular experiments.
  2. Data Dimensions: Only two metrics (step counts and body fat percentage) were compared. Future studies should analyze the applicability of comparisons across a broader range of sensitive and non-sensitive data.
  3. Long-term Validation: Using long-term research tools (e.g., diary studies) could further reveal users' adaptability and behavioral trajectories under prolonged social comparison conditions.

Through this study, the authors provide design references for health and fitness application developers to promote healthy behaviors and mitigate psychological harm among users.