Method
How a session runs
- 01PositionA transmitter position is selected by the protected schedule.
- 02CueThe transmitter receives only the instruction needed for that trial.
- 03Look / shamThe receiver remains condition-blinded.
- 04CommitThe receiver commits a continuous bearing.
- 05RecordThe response is recorded before the next trial begins.
Blinding
The receiver is not told trial type or transmitter position. The transmitter sees only a trial identifier and a cue that is either a bearing or SHAM, and never sees the receiver’s reports. The full schedule is not displayed in the room.
“Blinded” here describes study condition, not eye state. Receiver visual-state instructions will be standardized before data collection and are currently under faculty review as part of Phase 1 protocol finalization.
Analysis lock
The confirmatory analysis is frozen in the research source before any human data exists. The primary test and its alpha are read from that lock, not restated here:
- Confirmatory directional primary
- V-test (modified Rayleigh), wrapped error, hypothesised direction 0°, one-sided [protocol/HOTE-RS-01_Analysis_Lock.md § 2.5 V-test (confirmatory directional primary) @ af2d884]
- Alpha
- 0.005 [protocol/HOTE-RS-01_Analysis_Lock.md § 2.5 V-test (confirmatory directional primary) @ af2d884]
- Confirmatory session length
- 200 trials
- Sham trials per session
- 40 trials
No optional stopping. No peeking at confirmatory endpoints during a session. The presence measure is reported as a separate family and is not used as a second gate on the directional test.
Materials controls
A direct-only session cannot distinguish the effect under test from sensory cue leak. The materials contrast — air and flat glass as allowed positive conditions, against mirror, binoculars, live video and recorded playback as pre-specified controls — is required before any materials interpretation.