Skip to main content

DAG Adjustment Pocket Deck

Recall the main DAG variable roles used in causal adjustment decisions.

causal-inference5 min read

comparison battlecard Roles CMCI What are the four roles to check before adjustment? Bad control Baseline usage versus post-treatment usage Timing changes the role. Objection Should we control for everything so the model is fair? A stakeholder equates more controls with more rigor. Your line More controls are not automatically fair. We control for pre-treatment common causes and keep mediators or colliders out of the main effect. Do not let predictive model habits dictate causal adjustment. It explains the principle without dismissing the concern. Rule What is a backdoor path? A non-causal path from treatment to outcome through a common…

Read the full lesson

Sign up free: one personalized lesson every day, matched to your role and goals.

← Back to library

Contact us