EX-ITTEN-007 — Peripheral chromatic interface cues

Question and falsifier

What cue size and contrast are needed for alerts under divided attention and realistic motion? The hue-fragility prediction fails if hue-only detection remains invariant with eccentricity after retinal size and luminance are controlled.

Work completed

A proxy crossed eccentricity 0/5/10/20/30°, size .1/.25/.5/1/2°, and four chroma levels (100 conditions). It estimated the minimum tested size passing an arbitrary sensitivity criterion.

Pilot results

At the lowest modeled chroma, the minimum was .25° foveally, .5° at 5°, 2° at 10°, and no tested size passed at 20° or 30°. At 20°, only the two highest chroma levels passed, both at 1°; no condition passed at 30°. These are model predictions, not thresholds.

Gaze-contingent experiment ready to run

• Apparatus: calibrated display, chinrest, high-rate eye tracker, gaze-contingent presentation, and controlled ambient light.
• Factors: eccentricity, angular size, hue direction, luminance contrast, chroma, duration, motion, cue redundancy, and central-task load.
• Use cone-contrast coordinates and verify rendered output; include protan/deutan observers and older adults after the initial normal-trichromat pilot.
• Outcomes: detection sensitivity (d′), criterion, localization error, response time, fixation break, and central-task cost.
• Staircases estimate 75%-correct thresholds; trials with broken fixation are repeated, not silently included.

Current position / continuation

Stage: condition grid and analysis target complete; no gaze data collected. Next action: implement gaze-contingent stimulus timing and validate end-to-end latency before preregistering the 12-participant threshold pilot. Data: ../experiments/data/ex-itten-007.json.