Can this skill help with outlier management?

Yes, it features specialized exclusion analysis to identify if specific outliers or high risk-of-bias studies are disproportionately influencing your overall analytical results.

How does this skill improve my data science workflow?

It automates the rigorous process of cross-checking results against various assumptions, reducing manual testing time and ensuring your conclusions are scientifically robust and unbiased.

What specific capabilities does it provide for Claude?

It provides tools for exclusion analysis, multiple imputation for missing data, alternative measurement scoring, and a standardized framework for reporting uncertainty and confidence levels.

When should I use this Claude Code skill?

Use it when you need to validate meta-analysis results, test for outlier influence, handle missing datasets, or verify if your findings hold up when switching between parametric and non-parametric tests.

What does the Sensitivity Analysis skill do?

This skill enables Claude to evaluate the stability of research findings by testing how conclusions shift under different analytical assumptions, data conditions, and statistical models.

Sensitivity Analysis

Name: Sensitivity Analysis
Author: astoreyai

byastoreyai

•

データサイエンスとML

Evaluates the robustness of research findings by testing how conclusions change under varying analytical assumptions and data conditions.

Ensures the integrity of scientific and data-driven research by systematically testing the stability of results against different methodological choices. By facilitating exclusion analyses, alternative statistical modeling, and missing data simulations, this skill helps identify whether findings are truly robust or merely artifacts of specific decisions. It is an essential tool for automated research workflows, providing clear guidance on interpreting consistent results versus those that require caution due to sensitivity.

主な機能

01Exclusion analysis for outliers and high risk-of-bias studies

022 GitHub stars

03Measurement sensitivity via alternative scoring and outcome definitions

04Comparison of parametric and non-parametric statistical tests

05Standardized interpretation framework for reporting confidence and uncertainty

06Missing data handling including multiple imputation and best/worst-case scenarios

ユースケース

01Testing if results remain significant when accounting for missing data via imputation

02Determining if research conclusions are sensitive to specific statistical transformations

03Validating meta-analysis findings by performing one-study-removed analyses

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add astoreyai/ai_scientist sensitivity-analysis

For use in Claude.ai and ChatGPT

Download Skill