Case Study

Abduction Inspector — IatrogenicMedications.bayes

Enter what you actually measured for a patient. The tool computes each node's structural residual — the individual background factor that the model cannot explain from the observable covariates alone.

Why abduction matters

Population-level causal effects answer "on average, what would happen under intervention?" — but individual-level counterfactuals require something more: the specific, unobserved background factors that made this patient's outcome what it was. Abduction recovers those factors from what was actually observed, anchoring each U node to the value consistent with the facts of the case.

Without abduction, a counterfactual query uses population-average noise — which may be badly wrong for any particular individual. With it, you hold the patient's background fixed and ask what would have happened under a different prescription: the same person, different treatment. This is Rung 3 Step 1 of Pearl's three-step counterfactual procedure — the formal engine behind personalised medicine and individual liability assessment.

above model prediction (U > 0)
below model prediction (U < 0)
no observation — U at prior mean (0)
U node posteriors — structural residuals per outcome node