Statistics Verifier

This gives you structured checklists for fact-checking statistical claims and catching common analysis errors. It covers the basics like verifying sample sizes and checking if conclusions match the data, plus deeper stuff like detecting p-hacking, choosing the right significance tests, and spotting when correlation gets passed off as causation. The red flags table is genuinely useful for quickly identifying sketchy claims. It won't replace domain expertise, but it's a solid starting framework when you need to audit someone's numbers or validate research findings. Most helpful when you're dealing with press releases, research summaries, or any situation where someone might be overselling their data.

Install to Claude Code

npx -y skills add travisjneuman/.claude --skill statistics-verifier --agent claude-code