I split your post to have both topics in their own thread, hope it’s OK!
But in this case wasn’t the data collected first from two cohorts that were already preselected – the T1s and T2s and the healthy controls? Maybe I misread?
Right, this study was indeed retrospective. Was commenting on the point made above suggesting that any study asking about past behavior is retrospective—in some sense, sure, but not usually what is meant by that in terms of research design. As wanted to make that note about prospective designs, since I think it can be easy to assume that those are not correlational, but they still are as well.
yes that’s a good clarification.
I feel like dietary recall studies are a particularly bad type of retrospective study though, because they’re essentially not doing the work to get the good quality data. I mean they’ll ask things like: “in the last six months, would you say you ate a) two servings of blueberries a week b) one serving a week c) less than one a week d) less tan one a month e) never?”
Who remembers those things? And was it blueberry season? And how on Earth can we expect people to accurately recall their food intake when they’re already so prone to things like underestimating portion size and caloric intake…
Very good point! With salt intake it’s even harder to make an accurate estimation.