Re: Guest Post - "Evidence-Based"?


Let's all post our correlation coefficients!

Posted by: SP | Link to this comment | 03-26-16 8:58 AM
horizontal rule

Gelman has a solution?

Posted by: nosflow | Link to this comment | 03-26-16 9:01 AM
horizontal rule

Is the solution "blame societal problems on universities instead of the financial sector"?

Posted by: Moby Hick | Link to this comment | 03-26-16 9:06 AM
horizontal rule

I remember dsquared having strong opinions about evidence-based medicine in the past, but I don't remember what they were. Does anybody remember what his massive argument with LB about pregnancy and delivery was about? I could just look it up.

On the other hand, I agree with the contention that, in the cases in which the evidence appears to contradict (or point in a significantly different direction) from one's prior sense of, "how we should be doing things." That the bar for changing public policy should be high.

One of the issues that he mentions is whether "evidence-based" policy-making is contrasted with ideological policies (or, "anecdote and political prejudice " in his post), and I think that's a fair question, but I also don't think it cuts only one way. I think there are plenty of examples of people making the argument that social services can be cut back because, thanks to better data collection it's possible to achieve the same outcomes with less resources, and that is both an appeal to "evidence-based" policy and an ideological position.

Posted by: NickS | Link to this comment | 03-26-16 1:51 PM
horizontal rule

What dsquared says about about Gelman is this.

There's also a lot of methodological overlap with education research and even development economics. Based on my reading of key papers like Andrew Gelman's Garden of Forking Paths (PDF), any area of research which uses the treatment/response model of controlled experimentation from medical science but which doesn't require its methods for data selection and analysis to be registered ahead of time, is likely to be producing fragile results. Moreover, in many areas the academic literature is so compromised by unreproducible results and publication bias that even large metastudies are not going to solve the problem.

I'm not quite sure what the solution is supposed to be, but that implies that it involves, "requir[ing] methods for data selection and analysis to be registered ahead of time."

Posted by: NickS | Link to this comment | 03-26-16 1:54 PM
horizontal rule

Sorta topical

Posted by: J, Robot | Link to this comment | 03-27-16 10:46 AM
horizontal rule

I was at a meeting of telecommunications industry muck-a-mucks, at which it was emphasized that regulators should only engage in evidence-based policy making, because everything else discourages investment. Their vision of evidence-based policy making was closely aligned with "no regulatory policy whatsoever."

Posted by: Jack Warner | Link to this comment | 03-27-16 2:26 PM
horizontal rule

At least in medicine, failure to aggressively adopt "evidence-based" clinical practices has become a major source of liability, at least for large medical institutions. My only-barely-informed understanding of the issue is that this probably led to a few genuine practice improvements that are now fairly widespread but at this point it mostly causes a huge amount of wasted effort and routine reshuffling of clinical procedures (in ways that are likely to increase actual human error rates). But "there's a published study that shows [x] reduces error rates!" is basically irresistible evidence for a jury.

Posted by: urple | Link to this comment | 03-27-16 6:19 PM
horizontal rule

I still hope they keep with the hand washing.

Posted by: Moby Hick | Link to this comment | 03-27-16 6:44 PM
horizontal rule