Causal-Audit: A Framework for Risk Assessment of Assumption Violations in Time-Series Causal Discovery

TL;DR

Causal-Audit is a framework that assesses the risk of assumption violations in time-series causal discovery, guiding method selection and improving reliability.

Contribution

It formalizes assumption validation as calibrated risk assessment, providing diagnostics, risk scores, and decision policies for more trustworthy causal inference.

Findings

Risk scores are well-calibrated with AUROC > 0.95.

62% reduction in false positives among recommended datasets.

78% abstention on severe violation cases.

Abstract

Time-series causal discovery methods rely on assumptions such as stationarity, regular sampling, and bounded temporal dependence. When these assumptions are violated, structure learning can produce confident but misleading causal graphs without warning. We introduce Causal-Audit, a framework that formalizes assumption validation as calibrated risk assessment. The framework computes effect-size diagnostics across five assumption families (stationarity, irregularity, persistence, nonlinearity, and confounding proxies), aggregates them into four calibrated risk scores with uncertainty intervals, and applies an abstention-aware decision policy that recommends methods (e.g., PCMCI+, VAR-based Granger causality) only when evidence supports reliable inference. The semi-automatic diagnostic stage can also be used independently for structured assumption auditing in individual studies. Evaluation on a synthetic atlas of 500 data-generating processes (DGPs) spanning 10 violation families demonstrates well-calibrated risk scores (AUROC > 0.95), a 62% false positive reduction among recommended datasets, and 78% abstention on severe-violation cases. On 21 external evaluations from TimeGraph (18 categories) and CausalTime (3 domains), recommend-or-abstain decisions are consistent with benchmark specifications in all cases. An open-source implementation of our framework is available.