The tilt cursor

Feng Tian, Xiang Ao, Hongan Wang, Vidya Setlur, Guozhong Dai2007 · DOI

TL;DR

This paper introduces the "tilt cursor," a novel cursor design for pen-based interfaces that dynamically changes its shape to visually represent the pen's 3D orientation. Experiments show this design significantly improves user performance and reduces response latencies in specific drawing tasks on touchpads, enhancing stimulus-response compatibility.

What Problem Does This Paper Address?

Pen-based user interfaces, while intuitive, often face challenges with stimulus-response (S-R) compatibility, especially on devices like touchpads where the input and display are separated. This separation can lead to mismatches between how users intend to move the pen and how the cursor visually responds. While pen input devices can provide rich information like position, pressure, and 3D orientation, most current cursors only use position, ignoring the valuable orientation data. This paper investigates whether leveraging the pen's 3D orientation as a visual cue through a dynamically shaped cursor can enhance user performance and S-R compatibility in pen-based interactions, particularly on touchpads.

Methods

The core innovation is the tilt cursor, which dynamically reshapes itself to reflect the pen's 3D orientation (altitude and azimuth). The cursor's shape is calculated based on these orientation parameters, with its head at the origin and its projection onto the XY-plane forming the visual representation. The paper also introduces a specific marking menu selection task designed to evaluate drawing direction adjustments, differing from traditional marking menus.

Two experiments were conducted using a within-subject design with twelve participants.

  • Experiment 1: A circular menu selection task, a standard benchmark for S-R compatibility.
  • Experiment 2: The newly designed specific marking menu selection task, requiring users to draw a straight stroke from a central item to a destination item, following a center line.

Both experiments compared the tilt cursor against two baselines:

  1. Shape-fixed arrow cursor: The standard GUI cursor with a fixed arrowhead.
  2. Live cursor: A cursor whose arrowhead follows the movement direction.

Performance was measured by movement time and error rate.

Datasets

Not discussed in the paper. The experiments used custom tasks and collected data from participant interactions.

Key Insights and Ideas

  • 3D Orientation Matters for Drawing: The pen's 3D orientation is not just incidental; it can be a crucial visual cue, especially in tasks requiring precise directional control.
  • Touchpad Challenges: The separation of input and display on touchpads exacerbates S-R compatibility issues, making visual feedback even more critical.
  • Dynamic Cursors for Dynamic Tasks: While a static cursor might suffice for simple pointing, a dynamically adapting cursor like the tilt cursor is more beneficial for tasks involving nuanced movement and orientation.
  • Beyond Simple Pointing: The tilt cursor's effectiveness is task-dependent. It significantly aids tasks requiring directional adjustments (like drawing strokes) but shows less benefit in simple selection tasks where position is paramount.

Key Results

  • Experiment 1 (Circular Menu Selection): No significant difference in movement time was found between the tilt cursor and the shape-fixed arrow cursor. While the tilt cursor was slightly faster than the live cursor, the overall effect of cursor type was significant, suggesting that in this task, the orientation cue had a limited impact. Error rates were low and not significantly different across cursors.
  • Experiment 2 (Specific Marking Menu Selection): The tilt cursor significantly outperformed both the shape-fixed arrow cursor and the live cursor in terms of movement time. Participants using the tilt cursor took less time to draw the required strokes. Error rates were low and not significantly different.
  • Response Latencies: Analysis of drawing direction adjustments in Experiment 2 indicated that the tilt cursor led to smaller response latencies, meaning users could adjust their drawing direction more quickly and efficiently to follow the center line.

Limitations and Future Work

  • Touchscreen vs. Touchpad: The authors question whether the results for touchpads generalize to pen devices with touch-sensitive screens.
  • User Holding Habits: They also suggest future research to investigate if different cursor designs influence how users hold their pens.
  • Generalizability: While the tilt cursor showed promise, its effectiveness was task-specific. Further exploration of its utility across a wider range of pen-based interaction tasks would be valuable.

Summary updated

Abstract

In order to improve stimulus-response compatibility of touchpad in pen-based user interface, we present the tilt cursor, i.e. a cursor dynamically reshapes itself to providing the 3D orientation cue of pen. We also present two experiments that evaluate the tilt cursor's performance in circular menu selection and specific marking menu selection tasks. Results show that in a specific marking menu selection task, the tilt cursor significantly outperforms the shape-fixed arrow cursor and the live cursor [4]. In addition, results show that by using the tilt cursor, the response latencies for adjusting drawing directions are smaller than that by using the other two kinds of cursors.
BibTeX
@misc{tian2007tilt,
  author = {Feng Tian and Xiang Ao and Hongan Wang and Vidya Setlur and Guozhong Dai},
  title = {{The tilt cursor}},
  year = {2007},
  doi = {10.1145/1240624.1240675},
}

Loading related content...

Figure