Yet another great article by Gary Taubes, which focuses on how epidemiological studies of red meat have been analyzed:
While it’s a bit of a complex analysis (they find different techniques from multiple studies, then apply random combinations of those to a different study), here’s a nice summation about this part:
By the way, this is why whenever a study says they used some model or performed “adjustments” or otherwise manipulated data, I stop reading. (1) This is where biases come in (eg, “We know red meat/obesity/whatever is bad”); (2) there is no way to know what value to assign anything, age, weight, sex, exercise amount or type, etc; (3) there’s no way to know that something NOT in your model is more important than all the crap that is in your model.
The article contains links to two great videos from a conference sponsored by Swiss Re five years ago. I strongly advise watching the video of John Ioannidis; the second is a combination of remarks and questions. Our own Richard Morris got to ask one!
Are they consciously or subconsciously letting their preconceptions bias their analyses?
This kind of potential group-think is less of a problem in clinical trials, Zeraatkar notes, because the researchers have to specify in advance their overall analytical strategy before launching the trial and gathering any data. Although the researchers doing the clinical trials can still make numerous smaller decisions about how to analyze their data that can influence their results. The history of science, even in the hard sciences like physics, is rife with examples of researchers getting the incorrect results they expect, rather than what eventually turns out to be the right answer
I’ve read so many clinical trials that most definitely have a group-think going on and their only purpose is to provide proof for the result they wanted to see before the trial even began. Preconceived beliefs are so obvious upfront with some of the statements made in the strategies and summaries of the studies. I do believe there was a time where it occurred less in clinical trials, but not in this day and age. I see them saturated with politics or bias. I wonder if we will ever be able to get true science back into all these study databases? Or are they forever compromised and soiled?
Cool article though. Good read. Amazing connection of “multiverse” logic.
A good question. Greed corrupts all. It’s very hard to recover from that.
The Christian critique of capitalism (Locke, Smith, et al.) assumed that the capitalists would exercise restraint, and that a certain amount of wealth would be enough for them. The unrestrained greed of today’s capitalists would horrify those thinkers. It is literally a form of insanity and immaturity; a sane adult is capable of being satisfied and of self-restraint.
