causalfragility-lab/confoundvis-explorer: confoundvis-explorer v0.1.0 — Interactive Sensitivity Geometry

Read full text Summarise with VIRAT
Authors: causalfragility-lab
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

confoundvis-explorer v0.1.0 First public release of an interactive visualization system for sensitivity analysis to unmeasured confounding. Overview confoundvis-explorer provides a browser-based interactive environment for exploring how unmeasured confounding affects estimated relationships in nonlinear and multilevel settings. The system extends traditional scalar sensitivity metrics (e.g., ITCV) into a spatial and dynamic framework over SES–time surfaces. Key Features Interactive 3D surface visualization of f(SES, t) Local gradient analysis (∂f/∂SES) ITCV-based sensitivity thresholds Real-time slope attenuation and nullification under confounding Visualization of fragile regions across the surface Fragile mass: proportion of the surface vulnerable to confounding Purpose This tool introduces a geometric perspective on sensitivity analysis, enabling researchers to identify where effects are locally fragile rather than relying solely on global summary measures. Live Application https://causalfragility-lab.github.io/confoundvis-explorer/ Related Work This project complements the confoundvis R package by providing an interactive visualization layer for sensitivity geometry and fragility analysis. Notes This is the initial public release (v0.1.0). Future versions will extend functionality and integrate more closely with the confoundvis ecosystem.

Fields of Study & Subjects

Advanced Causal Inference Techniques Data Visualization and Analytics Statistical Methods and Bayesian Inference

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
30 March 2026
Language
EN