3 ms·
Whoa, that seems like a very strong claim claim and is contrary to my understanding. I know that instrumental variables can be a powerful approach, but it's not
by pseudonom- 9y ago
Whoa, that seems like a very strong claim claim and is contrary to my understanding. I know that instrumental variables can be a powerful approach, but it's not always easy to find a good instrument. And you can try to control for confounders, but it's hard to convince yourself you've gotten all of them. See, for example, https://arxiv.org/abs/1706.04692 https://arxiv.org/abs/1706.04692 where standard controls were of limited efficacy and it's only when they include an additional 3,700 controls that things start looking up.
Then there are things like https://en.wikipedia.org/wiki/Parameter_identification_problem https://en.wikipedia.org/wiki/Parameter_identification_probl... and whole books like https://en.wikipedia.org/wiki/Causality_(book) https://en.wikipedia.org/wiki/Causality_(book) examining the problem and its' difficulty.
Can you explain what you mean?
- achileas 9y agoEasily accounted for doesn't mean just proving absolute causality - you can account for the issue by either certain statistical tests giving evidence for causation, but more easily you write about associations, what gives you evidence for the reduction of confounds (strengthening your evidence, as they did in this paper), and, of course, papers are written to an audience that understands correlation != causation, and are written as such. People pointing out such an obvious thing that is always accounted for in various ways in research shows their ignorance of the specific article as well as the philosophy of science in general. It's just a reflexive response for some people at this point.