In the following, I will outline some of our work on distribution shift.
Explain or predict? Out-of-distribution generalization in replication studies (ongoing work with Ying Jin and Naoki Egami)
Let's say you want to know whether a statistical result generalizes. If you have data from numerous sites, you can use meta-analysis. However, for many scientific results, we have data from only a few sites or studies.
If you have individual-level data, you could use re-weighting methods from the generalizability literature to generalize from one site to another. However, this is often not successful. Below is a plot from ongoing work where we use re-weighting methods to generalize from one experimental site to another. For both entropy balancing and doubly robust approaches, the re-weighting does not move us closer to the target (dashed line), and the coverage of prediction intervals can be low. The data comes from the Pipeline Project (Schweinsberg et al., 2016), where 25 laboratories independently replicated experiments for 10 scientific hypotheses concerning moral judgment.