sensitivity-analysis-design

$npx mdskill add yogsoth-ai/de-anthropocentric-research-engine/sensitivity-analysis-design

Design sensitivity analyses to test meta-analysis robustness

  • Solves the problem of assessing how sensitive results are to study inclusion or exclusion
  • Uses input data on included studies, outliers, and subgroup variables
  • Applies statistical methods to detect influential studies and subgroup effects
  • Generates a structured sensitivity analysis plan for review and execution
SKILL.md
.github/skills/sensitivity-analysis-designView on GitHub ↗
---
name: sensitivity-analysis-design
description: Design leave-one-out, influence diagnostics, subgroup analyses, and robustness checks
execution: subagent
prompt: ./prompt.md
input: included_studies, potential_outliers, subgroup_variables
used-by: meta-analysis
---

# Sensitivity Analysis Design SOP

Design comprehensive sensitivity analyses to test the robustness of meta-analytic conclusions under different assumptions and exclusion criteria.

## When to Use

- After primary analysis plan is established
- When outliers or influential studies are identified
- When methodological decisions could affect conclusions

## Input

- `included_studies`: List of included studies with characteristics
- `potential_outliers`: Studies flagged as potential outliers or influential cases
- `subgroup_variables`: Variables for pre-specified subgroup analyses

## Output

Complete sensitivity analysis plan specifying each analysis, its rationale, and interpretation framework.
More from yogsoth-ai/de-anthropocentric-research-engine