What frameworks does it support for evidence assessment?

It primarily utilizes the GRADE (Grading of Recommendations, Assessment, Development, and Evaluations) system and Cochrane Risk of Bias (ROB) tools.

Is it suitable for all scientific fields?

While designed for general scientific rigor, it is particularly effective for empirical research in medicine, life sciences, social sciences, and data-driven disciplines.

Can it detect statistical errors in research papers?

Yes, it assesses sample sizes, p-value interpretations, effect sizes, and checks for common pitfalls like p-hacking, multiple comparison issues, and regression to the mean.

How does this skill evaluate research quality?

It uses a systematic framework to check for internal/external validity, experimental design appropriateness, control groups, blinding, and measurement reliability.

Scientific Critical Thinking

Name: Scientific Critical Thinking
Author: Zehong-Wang

byZehong-Wang

0•

Data Science & ML

Evaluates the rigor of scientific research by assessing methodology, statistical validity, and potential biases using established frameworks like GRADE and Cochrane.

This skill empowers Claude to act as a sophisticated scientific peer reviewer, providing a systematic approach to analyzing research papers, study protocols, and experimental designs. It leverages industry-standard frameworks like GRADE and Cochrane Risk of Bias (ROB) to identify flaws in internal/external validity, detect common cognitive and statistical biases—such as p-hacking or selection bias—and evaluate the overall strength of evidence. Whether you are conducting a systematic review, assessing the validity of a data-driven claim, or planning new research, this tool ensures your analysis remains objective, logically sound, and grounded in rigorous scientific principles.

Key Features

01Systematic bias detection for cognitive, selection, measurement, and analysis biases

020 GitHub stars

03Evidence quality assessment using the GRADE and Cochrane Risk of Bias frameworks

04Comprehensive methodology critique covering internal, external, and construct validity

05Identification of logical fallacies and common causal errors in scientific arguments

06In-depth statistical evaluation including power analysis, p-value interpretation, and effect sizes

Use Cases

01Conducting systematic literature reviews and meta-analyses to weight the strength of evidence

02Peer reviewing academic manuscripts to identify methodological flaws or reporting gaps

03Verifying the validity of data-driven scientific claims in technical documentation or news

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add zehong-wang/kosmos scientific-critical-thinking

For use in Claude.ai and ChatGPT

Download Skill