Context-Based Interface Prototyping: Understanding the Effect of Prototype Representation on User Feedback

Title of the Paper

Context-Based Interface Prototyping: Understanding the Effect of Prototype Representation on User Feedback

Paper Information

• Subject Area: Human-Computer Interaction (HCI), Autonomous Driving Systems, Virtual Reality and Prototyping
• Keywords: Prototyping, Virtual Reality, User Research, Prototype Representation, Autonomous Vehicles, Human-Machine Interface

Research Background and Problem

• Identified Challenges or Problems:

  • Autonomous vehicles (AVs) require new approaches to design and evaluate human-machine interfaces (e.g., eHMI) for user interaction.
  • Evaluating eHMI in urban environments is complex and costly, often requiring real autonomous prototypes and posing potential risks to participants.
  • Research on how different prototype representations influence user feedback remains insufficiently explored.

• Significance:

  • Autonomous vehicles will play a critical role in urban transportation and service automation, with their success relying on positive evaluations of trust, usability, and user experience.
  • Exploring effective and safe methods to evaluate eHMI design is a core issue in the rapidly advancing field of automation technology.

• Research Motivation and Related Work:

  • Researchers have utilized virtual reality (VR) and video simulation platforms to evaluate eHMI.
  • Previous studies have focused on specific prototype designs and evaluations of interface concepts, as well as comparisons of perceived immersion.
  • This study aims to address the knowledge gap regarding how different prototype representations influence participant feedback.

Solution

• Proposed Approach:

  • Conducted a comparative study of three prototype representations: Real-World Virtual Reality (RW-VR), Computer-Generated Virtual Reality (CG-VR), and Real-World Video (RW-Video).
  • Used a mixed-methods approach to evaluate user trust, user experience (UX), and immersion.

• Innovations:

  1. The first comparison of different simulation methods for evaluating eHMI from a pedestrian perspective.
  2. Provided empirical data to analyze the different aspects users focus on under various prototype representations.
  3. Proposed context-based interface prototyping and experimental evaluation guidelines.

• Implementation Steps and Key Techniques:

  • Constructed autonomous driving scenarios using three simulation technologies.
  • Collected quantitative and qualitative data from 42 participants through questionnaires and semi-structured interviews.
  • Quantitative analysis employed standardized Trust Scale, User Experience Questionnaire (UEQ), and ITC-Sense of Presence Inventory (ITC-SOPI).
  • Qualitative analysis applied thematic analysis to interview data.

Research Findings

• Specific Results:

  • RW-VR and CG-VR scored higher in spatial presence and engagement for immersion.
  • RW-VR performed best in naturalism and visual realism.
  • User experience evaluation results showed no significant differences across the three prototype representations.

• Advantages:

  • RW-VR provided the highest ecological validity, better simulating real-world conditions.
  • CG-VR balanced cost-effectiveness and flexibility for rapid iteration.
  • RW-Video was cost-efficient and suitable for remote, online user experience evaluations.

• Experimental or Evaluation Results:

  • In interviews, most participants preferred the Real-World Virtual Reality (RW-VR) prototype.
  • Qualitative data revealed that virtual characters in CG-VR scenarios could evoke feelings of unnaturalness or distrust.
  • RW-VR demonstrated advantages in assessing the overall feasibility and realism of the system.
  • While RW-Video had lower immersion, it helped users focus on individual interface elements.

• Limitations and Future Directions:

  • Limited by the sample size (42 participants), particularly in the quantitative analysis.
  • The novelty of VR might influence user evaluation results, especially for participants with no prior VR experience.
  • Future research should explore the impact of low-realism or abstract virtual characters on user feedback.
  • Design approaches should balance user interactivity with immersive experiences.

In summary, this paper provides comprehensive insights into context-based interface design and evaluation through a systematic comparison of different prototype representations. It proposes methodological guidelines and offers valuable references for the design and user research of future urban automation technologies.