Better Definition and Calculation of Throughput and Effective Parameters for Steering to Account for Subjective Speed-accuracy Tradeoffs

Title of the Paper

Better Definition and Calculation of Throughput and Effective Parameters for Steering to Account for Subjective Speed-accuracy Tradeoffs

Paper Information

• Domain: Human-Computer Interaction (HCI)
• Keywords: trajectory interaction, effective parameters, performance evaluation metrics, throughput, ISO9241-411, user behavior understanding

Research Background and Problem Statement

• Identified Problems or Challenges:

  1. Throughput (TP), as a metric for evaluating input devices and user performance, has been widely applied in pointing tasks but its applicability in steering tasks remains unverified.
  2. Whether existing TP calculation methods based on effective parameters (e.g., effective width We and effective amplitude Ae) can smooth subjective speed-accuracy tradeoffs has not been fully assessed.
  3. The impact of specific path shapes (e.g., linear, circular, sinusoidal paths) on TP is still unclear, especially under varying user speed-accuracy biases.

• Significance of the Research: Investigating the stability of TP across different path shapes and subjective user biases is crucial for improving the accuracy of input device and user performance evaluations, as well as guiding Graphical User Interface (GUI) design.

• Motivation and Related Work:

  • By improving TP calculation methods, the research aims to better reflect user performance in steering tasks and enhance model stability under multi-task biases.
  • Previous studies have demonstrated the predictive capability of effective width for movement time (MT), but the role of effective amplitude in steering tasks and TP stability has not been comprehensively explored.

Solution

• Proposed Methods:

  1. Based on effective parameter theory, the study refines TP definitions and model adaptation for steering tasks.
  2. A general method is proposed to calculate effective width We and effective amplitude Ae, representing path deviation and actual movement amplitude, respectively.
  3. Experiments are conducted to validate the applicability of these methods under different path shapes and speed-accuracy biases.

• Innovations:

  1. Theoretically extends the well-established Fitts’ law and effective parameters to steering tasks, providing the first empirical validation of their adaptability and TP smoothing performance.
  2. Evaluates effective amplitude (Ae) and its role in improving model adaptability for sinusoidal and circular path tasks for the first time.
  3. Proposes TPwe and TPe throughput metrics based on effective width and amplitude in steering tasks, offering fairer benchmarks for comparing input devices.

• Implementation Steps:

  1. Define and formalize effective width We and effective amplitude Ae.
  2. Design three independent path task experiments (linear, circular, sinusoidal paths), each testing different speed-accuracy biases (accurate, normal, fast).
  3. Use linear regression to validate model fit, assess main effects through ANOVA, and introduce throughput stability metrics.

Research Findings

• Specific Results:

  1. Effective Width (We):
     • We effectively smooths the impact of speed-accuracy biases on TP across all path shapes.
  2. Effective Amplitude (Ae):
     • Ae significantly improves model fit and task stability in circular path tasks.
     • However, Ae has limited effect in smoothing subjective biases for other path shapes.
  3. Optimal Path Shape:
     • Sinusoidal paths are recommended as the standard path shape for device comparison experiments due to their lack of occlusion issues and stable speed-position variations.
  4. TP Stability:
     • Compared to traditional TPn, the newly defined TPwe and TPe are better suited for comparing user performance across different biases and path shapes.

• Advantages Over Existing Solutions:

  • The improved effective parameter model significantly enhances TP stability and fairness under multi-task conditions compared to traditional TP definitions.
  • Provides comprehensive evaluations of user performance in complex path tasks (e.g., sinusoidal paths).

• Experimental or Evaluation Results:

  1. All three experiments confirm that TPwe using We smooths subjective speed-accuracy biases.
  2. Introducing Ae in circular paths improves performance prediction for large path tasks.
  3. Sinusoidal paths exhibit the highest overall TP stability and are recommended for input device evaluation experiments.

• Limitations and Future Directions:

  1. This study only validates linear, circular, and sinusoidal tasks; other complex path shapes (e.g., narrowing paths or paths with corners) remain untested.
  2. The impact of path curvature and wavelength on TP and user behavior has not been thoroughly discussed.
  3. Future research should explore performance differences across other touch input devices and user groups (e.g., elderly users).