Summary of
Project 2 Activity Survey
Feiyue Zhang (2025)
Summary of
Project 2 Activity Survey
Feiyue Zhang (2025)
Activity Background
In this activity, participants use a particle system to emit particles and observe the resulting 3D vector motion through three different visualization methods. Their ability to perceive the vector data is then evaluated through a survey to assess the effectiveness of each visualization approach.
Here's the link to the activity. The website is accessible on both PC and VR headsets. However the activity was conducted in a setting where all participants used a Quest 3 headset.
Visualization Methods
Overall preference
Classic mode was the clear favorite: 63.6% chose it as best overall.
Tape and Neon each had 18.2% of users preferring them.
From the user feedback, Classic was preferred for the following reasons:
Strong perception of orientation and motion, thanks to arrows/cones.
Less visual clutter and more intuitive directionality.
Effective for dense or large-scale data.
Survey Metrics Breakdown
1. Vector Length
Classic performed best (avg 4.00).
Tape (3.36) and Neon (3.55) lagged behind.
2. Vector Orientation
Classic again led (avg 4.55), with cones praised for directional clarity.
Tape (3.45) and Neon (3.18) were considered more confusing or cluttered.
3. Vector Density
Classic slightly ahead (4.18), with clear arrows/lines helping density perception.
Tape and Neon tied (3.82), though Tape was noted to be clearer in AR due to visibility.
4. Large-scale Data
Classic was most effective (4.18).
Tape and Neon were equal (3.55), but both suffered from clutter and overdraw when many vectors were present.
Key Insights
Classic mode is most effective across all tested dimensions: length, orientation, density, and scale.
Users value clarity, intuitive direction, and low clutter in viewing 3D motion data, and all these factors were achieved better in Classic.
Tapes' overlapping animations and awkward geometry hindered interpretation.
Neon mode was praised for its appearance but failed at providing enough information compared to other visualization methods.
Classic
Tape
Neon
AR/XR vs. VR
Overall preference
VR was preferred by most (72.7%), due to fewer distractions and better clarity.
AR felt noisy and visually confusing due to real-world background interference.
A few appreciated AR for its "reality mixing", for example when bubbles appeared to interact with ceilings in real-world.
My Thoughts
I was surprised that Classic mode was the better choice across all metrics according to the survey. Because when I worked on the project, I specifically chose those three visualization methods for the reason that they each have their own strength in presenting vector data. For example, Tape should be good at showing continuous change of orientation, and Neon should excel at visualizing density/distribution.
Not denying the survey results, I think there are many valid reasons for choosing Classic over others from the responses. But this indeed made me question the way I set up the experiment.
There was not many instructions on what the participants should do or see during the activity, because I didn't want to lead participants into thinking in a certain way. But this might be where the problem lies as participants are unaware of what to focus when there's too little guidance. My assumption is that some of the feedback were based on a gut feeling. If I were to design the activity again I would put more thoughts into designing instructions to provide a balanced amount of information.
With the recent trend in mixed reality, people praise a lot about passthrough mode and its ability to blend real life environment with virtual objects. But as this experiment suggests, XR isn't always the better "reality".
Participants find passthrough video feed sometimes distracting, especially in the case when the virtual content itself is intense or that the task requires attention. Maybe virtual reality really shines when users do serious work, like for productivity apps, as opposed to the popular notion that "VR = gaming"?