"Can we work this out?": an evaluation of remote collaborative interaction in a mobile shared environment

Dari Trendafilov, Yolanda Vazquez-Alvarez, Saija Lemmelä, Roderick Murray-SmithMobile HCI 2011 · DOI

TL;DR

This paper explores how to enable effective remote collaboration on mobile devices, even when users can't look at their screens. They found that while visual interfaces offer the best performance, spatial audio combined with haptics can still foster a sense of shared awareness and connection in eyes-free scenarios, opening doors for richer mobile interactions.

What Problem Does This Paper Address?

In today's world, people increasingly use mobile devices to stay connected and collaborate while on the go. While many social networking apps offer asynchronous updates, there's a growing need for more fluid, synchronous communication that mimics real-world interactions. This is particularly challenging for mobile environments where users' attention is often divided, making traditional visual or audio-video communication difficult. The paper tackles the problem of designing and evaluating novel interaction methods for mobile collaborative virtual environments (CVEs) that can support shared awareness and engagement, especially in "eyes-free" contexts where visual attention is compromised.

Methods

The researchers developed three prototype systems for mobile collaborative virtual environments, all based on touch input and utilizing a Nokia N900 mobile phone and a custom sensor pack. The core innovation lies in the different feedback modalities tested:

  • Visual-only: A graphical user interface (GUI) displaying a shared calendar, allowing users to visually track each other's cursor movements and identify free slots.
  • Audio-tactile: This "eyes-free" prototype replaced the visual display with a pitch-tone audio cue to indicate the partner's cursor location and movement direction. A chromatic scale was used, with higher pitches for leftward movement and lower pitches for rightward movement. This was complemented by vibro-tactile feedback for available slots and a sharp vibration when users met in the same slot.
  • Spatial Audio-tactile: Similar to the audio-tactile system, but it employed spatial audio (using HRTF filters) to position the sound source in a 3D space around the user, aiming for a more natural representation of location and direction.

Participants performed a collaborative target acquisition task: finding a shared free slot in their calendars. The experiment used a within-subjects design, with participants walking in a "walking lab" environment to simulate mobile conditions.

Datasets

Not discussed in the paper. The study used custom-built prototypes and a specifically designed collaborative task.

Key Insights and Ideas

  • The enduring power of visual cues: Even in a mobile context, visual feedback significantly outperforms eyes-free modalities in terms of perceived task performance and shared awareness. This highlights the fundamental role of vision in collaborative tasks.
  • Eyes-free is challenging, but possible: Despite performance drops, the eyes-free systems (especially spatial audio-tactile) demonstrated the ability to foster joint awareness. This suggests that with careful design, non-visual modalities can support collaboration.
  • Spatial audio's advantage: The spatial audio-tactile system showed a notable advantage over the simpler audio-tactile system in the eyes-free conditions, indicating that a more naturalistic spatial representation of information is beneficial.
  • The importance of "togetherness": While efficiency might be lower in eyes-free modes, the systems still managed to build a sense of "togetherness" and social presence, which is crucial for remote collaborative experiences.
  • Design challenges remain: The study identified usability issues with the input device (small size) and limitations in the audio mapping (pitch-tone range and mapping clarity), suggesting areas for future improvement.

Key Results

  • Perceived Task Performance: The visual condition showed significantly higher perceived task performance (p < 0.001) compared to both eyes-free conditions.
  • Togetherness, Social Presence, Affective Benefits: Participants reported significantly higher levels of togetherness (p < 0.001), social presence (p < 0.005), and affective benefits (p < 0.001) in the visual condition compared to the eyes-free conditions.
  • Eyes-Free Comparison: Within the eyes-free conditions, participants perceived significantly better performance in the Spatial Audio-tactile condition than in the Audio-tactile condition (p = 0.036).
  • Actual Performance: While perceived performance favored the visual condition, actual target acquisition also showed a trend favoring visual, though the difference was not statistically significant.

Limitations and Future Work

  • Short experiment duration: The experiment was relatively short, and the authors acknowledge that a marked learning effect was not observed, suggesting that longer training periods might be necessary to fully realize the potential of these novel interaction methods.
  • Limited training: Participants did not have the opportunity to verbally agree on strategies before the experiment, which made the task more challenging. Future work could explore how pre-defined strategies impact performance.
  • Input device limitations: The small size of the input device and the capacitive sensing resolution presented usability challenges, particularly for precise control.
  • Spatial audio mapping: While spatial audio showed promise, the mapping of pitch-tones and the narrow range were identified as areas for improvement. Further research is needed to explore the full potential of spatial audio in eyes-free mobile CVEs.
  • Longitudinal studies: The authors suggest that longitudinal studies are needed to better understand the long-term impact and learning potential of these interfaces.

Summary updated

BibTeX
@inproceedings{trendafilov2011can,
  author = {Dari Trendafilov and Yolanda Vazquez-Alvarez and Saija Lemmelä and Roderick Murray-Smith},
  title = {{"Can we work this out?": an evaluation of remote collaborative interaction in a mobile shared environment}},
  booktitle = {Mobile HCI},
  year = {2011},
  doi = {10.1145/2037373.2037447},
}
Figure