Applies to EyeSort 1.0 · Verified 2026-10-07Release notes

Compatibility and Known Limitations

The verified design envelope and analysis limitations of EyeSort 1.0.

Last updated: 2026-10-07

Supported design envelope

EyeSort 1.0 targets synchronized continuous EEGLAB .set data from single-line reading studies using fixed-width text, uniform pixels per character, and a consistent named-region schema. It uses horizontal X position only.

Upstream field names are configurable, so compatibility is determined by the event contract rather than eye-tracker brand. The project developed with common EYE-EEG conventions, but EYE-EEG is not required if equivalent events exist.

Not supported or not performed

  • synchronization or eye-event detection;
  • blink/missing-data identification or repair;
  • EEG preprocessing or ocular-artifact correction;
  • multiline/proportional-font geometry;
  • arbitrary image or pixel-drawn IAs;
  • automatic independent selection of every computed word;
  • automatic trial-level exclusion reports;
  • epoching, averaging, statistical modeling, or temporal-overlap correction.

Coding constraints

CC uses condition modulo 100, while RR and LL are formatted as two-digit indices. Designs with more than 99 regions/labels or condition values sharing the same final two digits risk ambiguity or collisions.

Scientific limitations

Behavioral labels do not prove cognitive states. Stringent criteria reduce and unbalance observations. Overlapping fixation-related neural activity remains a downstream analysis problem. Validate counts across participants and items and avoid causal claims unsupported by the design.