Quote:
Originally Posted by ToothSayer
I have a science degree (physics/math), so calling me "ignorant" of science or publication is a stretch. I'm not a working scientist though.
Should you saying that you have a proper subset of the degrees I hold make me more or less inclined to believe you?
Quote:
Most published research claims are false. Note that this applies to psychology/medicine/similar fields (which is what we're talking about), not the hard sciences.
I'll just leave you with the words of the former editor of the British Medical Journal:
Uhhh.... yeah... Here's a good summary from the Scientific American:
http://blogs.scientificamerican.com/...eve-it-or-not/
Quote:
Originally Posted by article
Where does that leave us? Is the global rate of false statistical positives in research closer to 15% or 50% or more? I think Goodman and Greenland make the case that we still don't know.
Reasonable minds accept that error can and do happen. But Ioannidis' approach is a bit strained. Ironically, one can argue that he had "something to gain" by making his results far more severe than the reality. And the article above points out how when actual statisicians looked at his work, they found it to be problematic. (Ioannidis is part of the statistics department "by courtesy" -- and yes,
that's a real thing.)
Your approach also suffers tremendously from an as-yet-unjustified extrapolation and ignoring data that works against you. This is just logically problematic. But your philosophical chops are weak, so such errors can be expected.
Now, had you been more clever, you would have pointed to the vast array of non-peer reviewed journals, which publish all sorts of nonsense, as a way to bolster the failure rate. And if you had done that, you might actually have gotten me. After I posted, I did worry about that, though that one is much harder to meaningfully quantify, and I could argue with you on whether those are considered acceptable publications (they're not). But since you didn't go there, it doesn't matter.
Quote:
I'll expound more on this when I get time to respond to your "Trump supporters are more authoritative" thread, in which you thought it was ridiculous that the result was more than 50% to be false (which is actually a lock). There's a hilarious data point that just came out in that field that will make you look very gullible. People are gullible and don't know how scientific research and publishing works and how incredibly unreliable it is.
That's not to bash science. Science is more reliable than most things. It's just that the best reliability we can come up with for any new field/area of inquiry is incredibly low. The world is full of error and noise and selection bias, and it takes many decades (of research, of engineering improvements) to work through that.
You may continue with both the bad science and the bad philosophy. I have no problem with you doing that.
Edit: Actually, I will be quite interested in seeing you wriggle around on your failure to understand the distinction between causation and correlation. Please do go back to that thread.
Last edited by Aaron W.; 07-03-2016 at 03:18 AM.