> afaik a study can consist of one single survey

It definitely can (and they had interviews too), although there's a lot of limitations with their methodology they don't address in the paper.