I mean, I'm not actually an expert, but from my little corner of the universe, the short form is a) it's logistically difficult and resource-intensive, and b) it's therefore hard to get the amount of funding that would pay for a Big Research Study.

And c) it's hard to figure out how to really do *good* social science type research, I think. But I also have the impression some people do know what they're doing and have techniques that work at least somewhat. But then you come down to the fundamental choice between things that are true/accurate, and things that are easily summarizable. I presume hard science has this problem as well, but there's lots more things in hard science where you can isolate some one little thing and poke at it and say "yup, that's how it works." Can't so much do that when you're studying human behavior, or if you can, it's not clear how much you've actually learned, because you've decontextualized it so much.

Also, I think for a lot of these things people want research to tell them the One Right Answer, and there probably isn't one.