Skip to main content
. Author manuscript; available in PMC: 2022 Jun 4.
Published in final edited form as: Nat Comput Sci. 2021 Jan 14;1(1):24–32. doi: 10.1038/s43588-020-00005-8

Fig. 1 |. Graphical representations for estimating causality in experimental and observational data.

Fig. 1 |

a, Graphical representation of an experimental study, where there is no link between treatment X and confounder Z as X is randomized. b, Graphical representation of an observational study, where both colliders C and confounders Z can bias causal effect estimates.