Fig. 1From: Frameworks for estimating causal effects in observational settings: comparing confounder adjustment and instrumental variablesA directed acyclic graph with one confounder and one IVBack to article page