How does it handle causal assumptions?

The skill provides tools for drawing DAGs and performing sensitivity analysis to evaluate the assumption of sequential ignorability and unmeasured confounding.

Can I analyze multiple mediators with this skill?

Yes, it provides implementation patterns for both parallel and sequential (serial) mediation architectures using the CMAverse framework.

Does it support non-continuous outcomes?

Absolutely; it includes implementations for binary outcomes using logistic regression and survival analysis using Cox proportional hazards models.

Which R packages does this skill support?

It supports the most widely used packages including mediation, medflex, CMAverse, survival, and dagitty for comprehensive causal modeling.

Is the Sobel test included in the workflows?

While the Sobel test is provided for historical context, the skill prioritizes and recommends modern bootstrapping methods for more accurate statistical inference.

Causal Mediation Analysis in R

Name: Causal Mediation Analysis in R
Author: choxos

bychoxos

0•

数据科学与机器学习

Performs advanced causal mediation analysis to decompose total effects into direct and indirect effects using industry-standard R packages.

This skill provides specialized guidance for conducting causal mediation analysis in R, helping researchers and data scientists quantify the mechanisms through which an exposure affects an outcome. It covers a comprehensive range of methodologies—from traditional Baron-Kenny steps to modern natural effect models and regression-based approaches—while offering robust support for complex data types including binary mediators, survival outcomes, and multiple sequential mediators. By incorporating sensitivity analysis and causal diagrams, it ensures that mediation results are both statistically sound and theoretically grounded.

主要功能

01Supports complex architectures like multiple parallel or sequential mediators

02Generates causal Directed Acyclic Graphs (DAGs) to visualize and validate assumptions

03Handles diverse data types including continuous, binary, and survival outcomes

04Conducts sensitivity analysis for unmeasured confounding using medsens and E-values

050 GitHub stars

06Implements mediation workflows using mediation, medflex, and CMAverse packages

使用场景

01Quantifying the indirect effect of a clinical treatment through a specific biological biomarker

02Performing sensitivity checks to determine the robustness of mediation results against hidden bias

03Analyzing how social programs influence long-term employment through intermediate skill acquisition

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add choxos/biostatagent causal-mediation

For use in Claude.ai and ChatGPT

主要功能

01Supports complex architectures like multiple parallel or sequential mediators

02Generates causal Directed Acyclic Graphs (DAGs) to visualize and validate assumptions

03Handles diverse data types including continuous, binary, and survival outcomes

04Conducts sensitivity analysis for unmeasured confounding using medsens and E-values

050 GitHub stars

06Implements mediation workflows using mediation, medflex, and CMAverse packages

使用场景

01Quantifying the indirect effect of a clinical treatment through a specific biological biomarker

02Performing sensitivity checks to determine the robustness of mediation results against hidden bias

03Analyzing how social programs influence long-term employment through intermediate skill acquisition