liveBook · Manning

Probability of disablement

As explained in Causal AI

The probability of disablement is the conditional probability that an outcome would be absent if a binary cause were absent, conditioned only on the outcome having occurred.

Probability of disablement (PD) is a counterfactual causal measure: among cases where an outcome occurred, it asks whether that outcome would have been absent if the binary cause had been absent. Formally, PD = P(Yₓ₌₀ = 0 | Y = 1). section 8.3.1

Unlike the probability of necessity and sufficiency, PD does not condition on whether the cause occurred. It therefore evaluates disablement across the outcome-positive population rather than only among cases exposed to the cause. section 8.3.1

Promotion and churn

For customer churn, PD asks: among users who churned, what is the chance they would not have churned if they had not received a promotion? section 8.3.1

How Causal AI covers it

How deep it goes
Defines the query formally and contrasts it with necessity, sufficiency, enablement, and necessity-and-sufficiency probabilities.
Examples
Uses customer churn and promotions to illustrate conditioning on observed churn while removing the promotion counterfactually.
What you can do after
Evaluate whether an observed outcome would disappear under removal of a binary cause.
How it connects
Connects counterfactual analysis with probability of necessity, probability of sufficiency, enablement, and causal attribution.
Cover of Causal AI
liveBookmanning.com
Read Causal AI on liveBook Get the book at manning.com

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Causal AI. The section links point to the parts of the book it draws on.