# Perception validation protocol v1 — freeze notes

Freeze date: 2026-08-31  
Git tag: `perception-validation-v1`  
Enrollment status at freeze: not started

This release freezes the prospective multi-rater study before participant
enrollment or prospective outcome review.

## Frozen design

- same 25 physical specimens as `subjective-pilot-v1`;
- one standardized assembly with labeled physical components;
- one randomized practice pass and two randomized scored sessions;
- whole-number 1–10 Weight and Tactility-sharpness ratings;
- no precompression or alternate assembly condition;
- target 24 completed adult raters and minimum 20 for confirmatory language;
- opaque specimen identities with a private code key independent of the public
  randomization seed;
- declared trial invalidity rules and no repetition of an exposed scored
  trial;
- frozen Weight Index and Tactility Index predictors; and
- participant-level Spearman analyses, participant bootstrap, Holm control,
  and prespecified secondary/sensitivity analyses.

## Operator materials

The release adds a plain-language SOP, verbatim participant script, scored-
session recording sheet, exception checklist, readiness checklist, apparatus
record, amendment log, data templates, and deterministic schedule generator.

The repository freeze does not claim the physical study is ready. The operator
must complete section B and section C of `FREEZE_CHECKLIST.md`, including a
full non-study dry run and the applicable consent/privacy process, before
participant 1.

## Scientific continuity

This protocol release creates no participant result and changes no released
measurement, curve, index formula, reference fleet, Force Curve Bench byte, EC
Parts Builder byte, or canonical-evidence byte. Preprint 2.0 remains the public
methods and exploratory-pilot foundation until prospective results exist.