TL;DR
This study investigates the usability of the Microsoft HoloLens 2 for elementary school children, finding it to be a "good" system overall. While children adapted well to various interaction modes, the direct "tap" interaction proved most efficient, though "voice command" was preferred. The use of HMD-AR positively impacted children's emotions, suggesting its potential for educational applications.
What Problem Does This Paper Address?
Augmented Reality (AR) holds significant promise for education, offering immersive and interactive learning experiences. Head-Mounted Display AR (HMD-AR) devices, in particular, allow for a seamless blend of the real and virtual worlds, potentially enhancing cognitive engagement. However, a key challenge for HMD-AR in educational settings, especially with younger children, is effective "human activity recognition" (HAR) – the technology that interprets user actions like speech and gestures. Previous research suggested that HAR in earlier HMDs might not be mature enough for young children due to their unique physical characteristics (e.g., interpupillary distance, hand size). This paper addresses this gap by evaluating the usability of the latest HMD-AR device, the HoloLens 2, with elementary school children, examining its interaction modes, and assessing its impact on their learning-related emotions. The goal is to provide an empirical basis for developing educational HMD-AR applications for this age group.
Methods
The study employed a within-subjects design with 47 elementary school children (2nd to 6th grade). Participants engaged with a standardized tutorial on the HoloLens 2, focusing on three AR interaction modes: "tap" (direct touch), "air-tap" (finger and thumb gesture at a distance), and "voice command" (gaze and voice input).
The researchers collected data on:
- System Usability: Assessed using a modified System Usability Scale (SUS) adapted for children.
- Interaction Efficiency: Measured by the "mean number of attempts" and "mean time" required to complete a task (selecting a virtual gemstone) for each interaction mode.
- Interaction Preference: Determined by asking children to rank their "most favorite" and "least favorite" interaction modes and provide reasons.
- Activity-Related Achievement Emotions: Evaluated using pre- and post-tutorial questionnaires measuring "enjoyment," "boredom," and "frustration."
Datasets
Not discussed in the paper. The study utilized a standardized tutorial pre-installed on the Microsoft HoloLens 2 device for data collection.
Key Insights and Ideas
- HoloLens 2 is Usable for Young Children: Despite not being designed specifically for children, the HoloLens 2 proved to be a "good" system for elementary school students, with no reported simulator sickness or issues with visual perception.
- "Tap" is King for Efficiency: The direct "tap" interaction mode was the most efficient, requiring the fewest attempts and the least time. This suggests that for immediate, hands-on tasks, direct physical interaction is superior.
- Voice Command Preferred, Despite Lower Efficiency: Interestingly, children preferred the "voice command" interaction mode, even though it was less efficient than "tap." This highlights that user preference isn't solely driven by objective efficiency, but also by perceived ease of use and personal comfort.
- Positive Emotional Impact: The use of HMD-AR significantly increased children's enjoyment and decreased their boredom and frustration. This emotional boost is crucial for motivation and interest in learning.
- Children's Physicality Matters: A technical issue where AR objects moved away from the child's view when they approached them suggests that the device's spatial positioning is calibrated for adults. This points to a need for technical adaptations for younger users.
Key Results
- Overall Usability: The median system usability score was 80, indicating "good" usability.
- Interaction Efficiency:
- Attempts: "Air-tap" required significantly more attempts than "tap" and "voice command."
- Time: "Tap" was the fastest, followed by "voice command," with "air-tap" being the slowest.
- Interaction Preference:
- Most Favorite: "Voice command" (44%), followed by "tap" and "air-tap" (both 28%).
- Least Favorite: "Air-tap" (58%), often cited as difficult.
- Reasons for Preference: Simplicity was a key factor for "tap" and "voice command," while "fun" was cited for "air-tap."
- Emotional Changes:
- Enjoyment increased significantly (p < 0.001).
- Boredom decreased significantly (p < 0.001).
- Frustration decreased significantly (p < 0.001).
Limitations and Future Work
The authors acknowledge several limitations:
- Sample Bias: The voluntary participation in a laboratory setting might have attracted children with a pre-existing interest in AR, potentially limiting the generalizability of findings to the broader population.
- Non-Randomized Tutorial Order: The fixed order of interaction modes ("tap," "air-tap," "voice command") could have introduced carry-over effects. However, the distinct nature of the modes might have mitigated this.
- Lack of Learning Content: The study focused on interaction usability rather than actual learning outcomes, meaning direct transfer to educational scenarios requires further validation with specific learning content.
Future work should:
- Validate findings with larger, more diverse samples and in actual classroom settings.
- Investigate the long-term effects of HMD-AR on learning and emotions.
- Explore affective processes in learning with HMD-AR, as emotions can have complex impacts on cognitive engagement.
- Address technical adaptations for children, particularly regarding interpupillary distance and object positioning.
- Assess the usability of interaction modes for more complex AR actions (e.g., object rotation, resizing).
- Compare the effectiveness of different instructional strategies when using HMD-AR.
Summary updated