06 / NOTICE · INSPECT · SUPPORT
Notice the moment. Understand the context.
Instructor awareness in virtual classrooms — an interpretable support dashboard, ready to inspect rather than merely imagine.
Synthetic session · no live learnersBrowser prototype implemented · Multi-user VR integration and instructor study: planned
Paused · 5× scripted replay · Clock values are scenario seconds, not measured classroom response times. Explicit help stays visible in both interface conditions.
The classroom, at task level
Six anonymous synthetic learners. Inspect the event evidence before deciding to intervene.
EVENT EVIDENCE / L03
Identify the component
2 recorded errors across this replay · No inference about attention, motivation or mastery.
Stage 1 · started
Stage 1 · completed
Stage 2 · started
Stage 2 · error
Stage 2 · error
Support queue
A cue invites inspection. It does not diagnose a struggling learner.
L02Help requested
An explicit request for instructor support.
Cue opened 00:26 · 19s agoL03Repeated errors
2 recorded errors on the current stage. Check the cause; difficulty is not inferred.
Cue opened 00:42 · 3s agoLogging support does not alter the scripted learner events. Later completion in this replay must not be presented as an intervention effect.
Make the cue rules explicit
Change a threshold to inspect its effect. Acknowledgement history resets when the policy changes.
Priority: explicit help → repeated errors → event gap. An event gap may mean reflection or missing telemetry, not inattention. Correct completion clears current-stage error/help cues. No opaque score is used.
What the trace can — and cannot — tell us
- Recorded here
- Scripted event availability, cue policy, scenario time, acknowledgement and intended support action. Export contains no personal learner data.
- Requires a real study
- Missed difficulties, appropriate interventions, instructor workload and independent learner outcomes. Acknowledgement alone is not evidence that support helped.
- Requires integration
- A shared VR classroom event adapter, consent-aware identifiers and explicit help/error reporting. This page makes no calls to VRJ’s private backend.
From a dashboard prototype to a defensible study
A proposed evaluation, not a report of results.
Hold the classroom constant
Counterbalance equivalent teaching sessions with and without analytical support. Keep the task, ordinary communication and help channels consistent. Do not replay the identical scenario as if it were unseen.
Measure more than speed
Predefine observable difficulty episodes and an independent coding rubric. Compare notice latency, missed cues, unnecessary interventions and perceived workload. Assess learner performance separately.
Protect learners
Use consent and appropriate ethics approval for participant evaluation. Collect minimal task events, define retention/access rules, and avoid gaze, movement or biometric attention claims.
Research precedent: Han et al. (2020), instructor-facing learning analytics. Their setting and design differ from this proposed VR study.
Event contract and VRJ integration boundary
The replay uses a minimal event: scenario timestamp, anonymous learner ID, stage index and explicit kind (started, completed, error or help). VRJ’s existing progress contracts provide a foundation for completion/results reporting; help requests, repeated-error histories and instructor notice timestamps need explicit event instrumentation. This prototype is not a reconstruction of the existing VRJ admin dashboard.
{
"at": 26,
"learner": "L02",
"step": 1,
"kind": "help"
}