Well, there are many topics that we do science with, and which have multiple variables that we can't always control for, especially if you're doing survey studi

Well, there are many topics that we do science with, and which have multiple variables that we can't always control for, especially if you're doing survey studies for health, or disease or other macro-level stuff. It seems like the extremes of micro and macro of most systems of knowledge are always subject to a large amount of variance. That said a lot of macro-economics is very hard to replicate, making it of dubious value.