feat(self-review,analyze-stats): design/power statistic provenance guard#38
Open
Yoojin-nam wants to merge 1 commit into
Open
feat(self-review,analyze-stats): design/power statistic provenance guard#38Yoojin-nam wants to merge 1 commit into
Yoojin-nam wants to merge 1 commit into
Conversation
…e guard Design and power statistics (minimum detectable effect, a-priori/post-hoc power, required sample size) are computed rather than extracted, so they have no CSV or source-paper row for the existing source-fidelity audit to trace — a common blind spot through which a non-reproducible value can survive multiple review rounds. - self-review: new Phase 2.5a-2 requires every reported MDE / power / sample-size value to be reproduced by committed code (method + inputs stated), with an independent recompute, a cross-manuscript method-consistency check, and an enforced gate. - analyze-stats: rule 9 — design/power statistics must be emitted by the committed script (never hand-computed in a side tool) using one consistent method family. validate_skills.sh passes. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Phase 2.5asource-fidelity audit to trace — a blind spot through which a non-reproducible value can pass multiple review rounds.Phase 2.5a-2— every reported MDE / power / sample-size value must be reproduced by committed code (method + inputs stated); independent recompute; classify as calculation error vs method/provenance drift; cross-manuscript method-consistency check; enforced gate.Motivation
Generalises a real failure mode: a hand-entered minimum detectable effect that no standard method reproduced survived several review rounds because no committed script computed it; separately, correct sample sizes were produced by an exact-t tool while the committed script used a normal approximation (right value, no reproducible provenance). The precedent in the skill is synthetic.
Test plan
bash scripts/validate_skills.sh— ALL CHECKS PASSED (0 failures, 0 meta-doc PII)🤖 Generated with Claude Code