Fig. 2From: Confounding-adjustment methods for the causal difference in mediansEstimates for the Longitudinal Study Of Australian Children (LSAC) example [12, 13] obtained under each confounding-adjustment method, where \(\hat{\delta }\) \(= \hat{m}_1 - \hat{m}_0\) is the estimated difference in median SDQ scores if all children were exposed to maternal mental health problems compared to if none of them were exposed. Point estimates and their corresponding 95% confidence intervals are presented alongside the figureBack to article page