analyzing-causal-dag

What problem does it solve?

This Skill helps you identify which variables to adjust for when estimating causal effects from observational data, so you do not confuse confounders, mediators, colliders, or instruments with ordinary controls.

Core Features & Use Cases

• DAG-first analysis: Forces explicit commitment to treatment, outcome, candidate variables, temporal order, and directed edges before estimation begins.
• Adjustment-set selection: Applies the backdoor criterion to choose valid covariates, exclude post-treatment variables, and avoid collider bias.
• Causal scenario handling: Supports observational effect claims, adjustment questions, sign flips after adding controls, mediator-aware direct-versus-total effect decisions, and instrumental-variable edge cases.
• Defensible reporting: Produces the adjustment set, classification table, assumptions, and sensitivity-analysis guidance such as E-values or tipping-point checks.

Quick Start

Use the analyzing-causal-dag skill to help me define the treatment, outcome, candidate variables, temporal order, DAG edges, adjustment set, and sensitivity analysis for this observational study.