BCI Report Research previewPeople
21
each recorded once on each display
Context transfer · P300 on a screen and in VR
Screen to VR
21 people used the same P300 interface on a PC screen and in a virtual-reality headset. Calibrated on one display and tested on the other, a simple logistic regression reached 66.0% balanced accuracy, where chance is 50%. That is how well a calibration carries over. It is not how much the change costs: this protocol did not also test each person on the display they were calibrated on.
66.0%Balanced accuracy after the display change, 21 people
Above chance after the change — with no same-display score to compare
Interval 61.6%–70.2%; AUROC 0.732. Each person was calibrated on one display and tested on the other in both directions, and the two directions were averaged within the person. Without a same-display score for the same people, this number cannot say whether the display change lowered performance, or by how much.
Two fixed baselines
Calibrated on one display, tested on the other
Target against non-target epochs, 5 non-targets to every target, so accuracy would reward always answering “non-target”; balanced accuracy and AUROC do not. Primary timing.
| Baseline | Balanced accuracy | AUROC |
|---|---|---|
| Mean-window logistic regression | 66.0%61.6%–70.2%66.0% balanced accuracy | 0.7320.671–0.785 |
| Spatiotemporal shrinkage LDA | 63.6%59.6%–67.7%63.6% balanced accuracy | 0.7130.652–0.771 |
Prespecified sensitivity
The timing correction is part of the result
The headset and the screen report stimulus onset with different delays. The primary analysis shifts each epoch by the source-reported offset — 19 samples on the PC and 60 in VR, at 512 Hz. The sensitivity analysis leaves epochs at the recorded tag. Neither scheme was chosen from test scores.
| Baseline | Onset corrected (primary) | At the recorded tag |
|---|---|---|
| Mean-window logistic regression | 66.0%61.6%–70.2% | 63.8%61.0%–66.5% |
| Spatiotemporal shrinkage LDA | 63.6%59.6%–67.7% | 57.2%55.1%–59.4% |
Spatiotemporal shrinkage LDA depends on timing most: 63.6% with the correction, 57.2% without. A mean offset is not the same as correct timing, but a result reported without saying which timing it used is incomplete.
What the scores are computed on
Twenty-one people, one recording on each display
Every label comes from the source; none was interpolated. One malformed row at the very end of one recording was discarded; it followed every possible epoch.
Labelled events
30,240
1 target to 5 non-targets
Epochs kept, primary timing
30,101
139 rejected at 500 µV peak to peak
Methods & limits
What this comparison can and cannot support
A within-person, offline transfer test with two simple baselines. It narrows one question; it does not rank displays or headsets.
No same-display reference
Nobody was tested on the display they were calibrated on in this protocol, so the cost of changing display is not measured here — only the level reached after the change.
Within one person, not across people or days
Each person supplies both their calibration and their test data, on the same day. Nothing here speaks to a model that works for a new person or a later session.
Two fixed baselines, not a ceiling
Two simple CPU classifiers with no tuning. No foundation model and no fine-tuned model was run in this batch.
A mean offset, not per-event timing
Correcting the average onset delay does not remove event-to-event jitter, which a headset can add.
Cattan PC/VR P300 · a BCI experiment in virtual reality and on a personal computer
Grégoire Cattan, Anton Andreev, Pedro L. C. Rodrigues and Marco Congedo · Dataset of an EEG-based BCI experiment in Virtual Reality and on a Personal Computer, Zenodo, doi:10.5281/zenodo.2605205; documentation arXiv:1903.11297.
All participants gave written informed consent covering the experimental process, the data management procedures and the right to withdraw at any moment; the study was approved by the ethics committee of the University of Grenoble Alpes (CERNI). Only cohort aggregates appear here.
Data source: reviewed aggregate JSON · schema bci-report-context-update-v1 · generated 2026-09-27.