# Causal model A causal model represents the mechanisms through which the variables of a system interact and influence one another, rather than merely the correlations between them. Working across metaphysics and statistics, these models use formal notations — structural equation systems and the [[Directed_acyclic_graph|directed acyclic graph]] chief among them — to make the assumed causal relationships explicit and checkable. ## Microsim — queued <p class="wt-pending"><strong>Microsim/diagram layer in the draft queue.</strong> Troy is building the interactive station for this concept by hand; the sourced overview below is complete and citable now.</p> ## Overview Modern causal modelling, developed largely by Judea Pearl, separates three levels of reasoning: association (what is observed), intervention (what happens if something is deliberately changed), and counterfactual (what would have happened under different circumstances). That ladder lets researchers in epidemiology, machine learning, and signal processing draw causal conclusions from observational data — the same conclusions a [[Bayesian_network|Bayesian network]] reaches through explicit conditional probabilities, and that a qualitative [[Causal_loop_diagram|causal loop diagram]] sketches before any equation is fit. **On the spine:** [[Bayesian_network]] · [[Directed_acyclic_graph]] · [[Path_analysis_(statistics)]] · [[Causal_loop_diagram]]. ## Wikipedia : Wikitube **Strict pair:** [Wikipedia](https://en.wikipedia.org/wiki/Causal_model) : [Wikitube](https://en.wikitube.io/wiki/Causal_model) ## Previous hub tags Hubs: `Systems`. Portals: [[PORTAL_Causal_loop_diagram]]. --- *Repopulated 2026-08-05 · text transfer · 0 deletions.*