3 ms·
Self-reported studies are bunk and relying on them is bunk science as far as I’m concerned. We know that non-response and outright lies are a problem. So you’ve
by vlovich123 6y ago
Self-reported studies are bunk and relying on them is bunk science as far as I’m concerned. We know that non-response and outright lies are a problem. So you’ve got a garbage in garbage out scenario.
There may be places where surveys are scientifically justifiable but I have yet to see such beyond basic preference ones. Even in simpler cases like political surveys there’s lots of problems and they fail us with regularity. Do you really think more complex applications somehow avoid the problems?
- jb775 6y agoRather than asking patients directly, they could at least use their previous addresses to deduce many things, such as water supply source, air pollution, etc (while keeping it anonymized). The govt could even run studies on cancer deaths cross referenced with location data from their tax returns, and their parent's tax returns (to figure out where they lived as a child). I'm just saying I've never heard this common sense approach actually discussed. It just seems like we often ignore the concept of "an ounce of prevention is worth a pound of cure".