movement-science-stats

What problem does it solve?

It prevents incorrect or misleading statistical conclusions in movement science and neuroscience by systematically checking common analysis and reporting errors that frequently slip into manuscripts and code reviews.

Core Features & Use Cases

• Automated statistical QA checklist: Reviews analyses and results sections for the 10 most common mistakes (e.g., missing controls, inflated unit of analysis, spurious correlations, underpowered designs, circular analysis, p-hacking, uncorrected multiple comparisons, over-interpreted null results, and correlation/causation confusion).
• Kinematics-aware guidance: Flags issues specific to time-series and bilateral/trial-level data (e.g., why running independent tests per timepoint is risky; when to aggregate to participant level or use mixed models).
• Tool- and workflow context: Supports common scenarios where you are analyzing data, interpreting results, writing results, or reviewing analysis code, aligned with Makin & Orban de Xivry (2019).

Quick Start

Ask the assistant to run the movement-science-stats checklist on your manuscript results paragraph and analysis description, highlighting any of the 10 statistical mistakes and what to fix.