Figure 1.

Sampling designs for studies extending inferences from a randomized trial to a target population. For detailed descriptions of notation, see the main text. Briefly, X denotes baseline covariates, S the trial participation indicator, and D the indicator for sampling into the study sample. The dashed box around D = 0 denotes the fact that covariate data may not be available from this subset and that in nonnested trial designs the size of the subset may be unknown.