Statistical Code Review

What problem does it solve?

This Skill addresses the critical need for maintaining high standards in statistical analysis code, ensuring reproducibility, and catching potential errors before they impact research outcomes.

Core Features & Use Cases

• Reproducibility Check: Verifies that analyses can be rerun from raw data to final output by checking for proper seeding, data path handling, and session information capture.
• Model Specification Audit: Reviews the correctness and justification of model choices, including outcome types, degrees of freedom, confounder inclusion, and DLNM parameterization.
• Data Pipeline Review: Examines data manipulation steps for correctness, missing data handling, temporal alignment, and outlier treatment.
• Results Integrity: Ensures that results are reported comprehensively with confidence intervals and sensitivity analyses.
• Use Case: When submitting a pull request for a new analysis script, use this Skill to perform a thorough review against established best practices, ensuring the code is robust, reproducible, and scientifically sound.

Quick Start

Review the attached R script for reproducibility and adherence to statistical best practices.