Focus is on transporting findings from a study to a reference population or populations. An incomplete sampling frame or incomplete information on the sampling process play the role of `missing data.' The central issue is whether relations in the sample can be transported to a target population. Success depends on compatibility of causal structures in study and target populations and on appropriate weighting by sample inclusion propensities. Big data and administrative records have the potential to provide relevant information on these and thereby improve transportation. However, these potentials shouldn’t replace proactive collection of information that will improve weighting and increase validity of transported results. Approaches include establishment of field- or subfield-specific population databases, and collection of baseline information that may not be needed to adjust inferences within the population directly studied, but will improve transportation. Importantly, probability-based sampling, including random assignment should be employed whenever possible. Examples are drawn from clinical trials, observational studies and surveys.