Skip to content

docs: run-2 pre-registration Amendment 1 - #12

Merged
Mattral merged 1 commit into
mainfrom
docs/run2-prereg-amendment-1
Oct 8, 2026
Merged

Mattral merged 1 commit into
mainfrom
docs/run2-prereg-amendment-1

Conversation

@Mattral

@Mattral Mattral commented Oct 8, 2026

Copy link
Copy Markdown
Owner

Amendment 1 to docs/v1_run2_preregistration.md, dated 2026-10-08, made before any run-2 data exists. Original text stays in git history (c153902).

  • §8.4: a Rule B failure with ρ_gap ≥ 0.80 reads "underpowered to settle it", not "no added value". Threshold fixed now.
  • §13: Rule B power table from scripts/power_rule_b.py (Gaussian copula, 300 sims per cell, 10,000-resample paired bootstrap, seed 0), with assumptions stated. It broadly reproduces an independent simulation; ranges are wider because r(gap, base) spans 0.2–0.8.
  • §13: limitation that the 24 conditions are 4 factors × 6 ordered severities, not exchangeable draws, so every CI is too narrow.
  • §12: the analysis script is tested on synthetic data and on run 1's footage in blind mode only.

🤖 Generated with Claude Code

Made before any run-2 data exists. Adds a Rule B power table from
scripts/power_rule_b.py (Gaussian copula, 300 sims/cell, seed 0), fixes
that a Rule B failure with rho_gap >= 0.80 reads "underpowered to settle
it", adds the non-exchangeable-conditions limitation, and pre-registers
blind-mode testing of the analysis script on real footage.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@Mattral
Mattral merged commit 6c44167 into main Oct 8, 2026
3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant