The main question that comes up when I see a study like this is if they were able to take the same hypothesis and replicate it on another dataset in a different locale. For instance, presumably you could run the same study on UK data. Would we see the same results?