Triangulating on Possible Futures: Conducting User Studies on Several Futures Instead of Only One

Research Background and Issues

• Identified Problems and Challenges: The author highlights that relying on single case studies to infer future technology use scenarios may lead to limitations in research findings, particularly when predicting the uncertainty and complexity of the future. This approach makes it difficult to derive conclusions with high generalizability and credibility.
• Importance of the Problem: The field of Human-Computer Interaction (HCI) is typically future-oriented. Designing and evaluating the application of new technologies requires accurate predictions of future scenarios to avoid misplanning and overlooking potential risks. Additionally, studying multiple scenarios enhances the overall perception of future possibilities, supporting more universal design and technological development directions.
• Research Motivation and Related Work: Although triangulation has been widely used in social sciences and psychology, its potential in future-oriented HCI research remains underexplored. Previous related studies have largely focused on single future scenarios, neglecting the possibility of triangulation across different future scenarios.

Solution

• Proposed Method: The author introduces a novel research method—user studies involving triangulation across multiple future scenarios. This method constructs two distinctly different yet appropriately overlapping future scenarios to extract findings that are either broadly applicable or scenario-specific through comparative analysis.
• Innovations:
  • The method employs two contrasting possible futures (e.g., "CorpoCapitalist" and "MyData" scenarios) and places users in similar but fundamentally different interaction prototypes for deeper comparisons.
  • Future scenarios are analyzed from multiple dimensions, enabling the study to reveal cross-scenario applicability or scenario-specific insights into a particular design or technology through cross-validation and in-depth understanding of differences.
• Implementation Steps and Techniques:
  • Designing Future Scenarios: The two futures focus on learning-oriented AI under personal data control (MyData scenario) and technology centered on corporate goals (CorpoCapitalist scenario).
  • Creating Prototypes: Two AI user interface prototypes were developed ("Invezterr" representing MyData; "Stockify" representing CorpoCapitalist), incorporating explicit system feedback mechanisms.
  • Conducting User Studies: Experimental designs captured user attitudes, behavioral patterns, and feedback, including task execution, system interaction, and cognitive adaptation.
  • Data Evaluation: Data collection combined quantitative (surveys) and qualitative (interviews and observations) methods, followed by comparative analysis of the two scenarios.

Research Findings

• Specific Findings:
  1. Shared Discoveries: Both scenarios revealed that users hold a positive attitude toward AI usage, viewing AI as an opportunity rather than a threat. Additionally, users developed techniques to avoid bias, although these techniques manifested differently across the two scenarios.
  2. Scenario-Specific Discoveries:
     • In the MyData scenario, users experienced a greater sense of control and viewed AI as a tool for skill development and professional growth.
     • In the CorpoCapitalist scenario, users engaged in gamified interactions, with many believing their contributions significantly influenced broader AI decision-making, albeit with a lower sense of personal control.
• Advantages Over Existing Solutions:
  • Overcomes the limitations of single-scenario studies by revealing shared findings and scenario-specific conclusions through comparative analysis across multiple possible futures.
  • Provides new perspectives for designing and evaluating future technologies, particularly enabling more forward-looking designs in the HCI domain.
• Experimental or Evaluation Results:
  • User satisfaction in the MyData scenario was significantly higher than in the CorpoCapitalist scenario, especially regarding the sense of control and task compatibility.
  • Gamified elements in the CorpoCapitalist scenario inspired a unique user motivation, despite the more passive nature of its tasks.
• Limitations and Future Directions:
  • Limitations: The small sample size makes it difficult to derive widely applicable quantitative conclusions through statistical analysis. Future research needs to address the challenge of distinguishing "overlapping findings" from "unique findings" across scenarios.
  • Future Directions:
    • Explore more types of cross-future triangulation research, such as studies involving different user groups, locations, or social contexts (proposed F3 type).
    • Develop systematic methodologies to extend triangulation applications across three or more future scenarios.
    • Investigate the socio-psychological theoretical background of user motivations and behavioral patterns to provide practical guidance for AI-oriented work design.

Conclusion

This study proposes an innovative method for future-oriented HCI research—triangulating possible futures to better understand the broad applicability and limitations of technology design. The research not only offers profound insights into future technology design but also lays the foundation for constructing more forward-looking research methodologies.