2019/05/19 by Issa J Dahabreh, Dahabreh, Issa J., Sebastien J-P. A. Haneuse +11 · 2 citations
Mathematics · #Advanced Causal Inference Techniques #Applications (stat.AP) #FOS: Computer and information sciences #Methodology (stat.ME)
paper · pdf · doi:10.48550/arxiv.1905.07764
openalex publication_date 2019/05/19 · openalex created_date 2022/07/29 · openalex updated_date 2026/07/28
We examine study designs for extending (generalizing or transporting) causal\ninferences from a randomized trial to a target population. Specifically, we\nconsider nested trial designs, where randomized individuals are nested within a\nsample from the target population, and non-nested trial designs, including\ncomposite dataset designs, where a randomized trial is combined with a\nseparately obtained sample of non-randomized individuals from the target\npopulation. We show that the causal quantities that can be identified in each\nstudy design depend on what is known about the probability of sampling\nnon-randomized individuals. For each study design, we examine identification of\npotential outcome means via the g-formula and inverse probability weighting.\nLast, we explore the implications of the sampling properties underlying the\ndesigns for the identification and estimation of the probability of trial\nparticipation.\n