We as humans have a tendency to draw false conclusions all the time (want to emphasize that there's nothing special about this being only a human tendency. Any intelligent animal is prone to draw false conclusions). As Taleb said so clearly-humans have a penchant to relate facts, to see stories, to get fooled by randomness; we see inter-relatedness and causality where none exists, or can even possibly exist!
This is the problem with consciousness, of intelligence. The same intelligence which gave us the ability to light fire and antibiotics, gave us religion and psychology. In one case the cause-effect was demonstrable; the other was largely our imagination, but so strong was our imagination that even if we can't find anything at all in the real world to prove our assertions, we hold on to our assertions anyway.
Consider this article "Are Stress And Multiple Sclerosis Linked? Apparently Not" published in a scientific journal, Neurology.
When you see the summary-you are astonished to see how wild the whole "data taking" and conclusion drawing process must be. I know this firsthand because when I was younger and less experienced in life, I used to draw lots of these false conclusions, in a Department of Psychology. You get carried away by forming your hypothesis, analyzing and collecting data, and finding proof or unproof of whatever you said. The only saviour in this study is that the final conclusion is good-that stress and MS are not linked.
Now use your intelligence a bit. Without any study at all, can you say with good confidence, let's say something like 99% confidence, that there is no correlation between stress and MS? What is stress anyway? In the study we are trusting the self-analysis of the subjects to find out their levels of stress.
I rest my case. If Neurology publishes such useless articles; you can imagine what happens in journals of social sciences-Sociology, Psychology, etc. They are full of all kinds of false conclusions, which are so unlikely to be true, that you might as well dump all these journals and never read them. Here's an example-where a psychology researcher is trying to establish that teaching somebody two languages helps them prevent Alzheimers! This story by the way is one of the most popular NY Times stories at this moment, so you can imagine that thousands of people are actually believing the preposterous logic in this research.
You will come back to say that it does have to start somewhere, that even if 1% chance of it being right, we have to investigate it, etc. etc. but let's step back a bit; instead of throwing random darts let's build slowly upon the good knowledge we have already accumulated, that's real science and real research. Not come up with improbable hypothesis and waste tons of time in trying to prove or disprove them. But when you are an avid researcher in these fields you dont even see this-you are so enamored by your research areas that you keep on testing one hypothesis vs. another, even if to an outsider they may seem completely unlikely to be true right from the start.
Let's evaluate our conclusions and hypothesis first with good sound logic based on first principles of science and mathematics (I admit I stole this phrase from Elon Musk of Tesla/SpaceX, who I am a great fan of); and see if there are holes in them. We will save so much of our time! (I re-iterate that I was a victim of this earlier in my life, especially in Graduate school; but after reading Smith and Taleb, the ability to think about the world logically took a quantum leap).