Saint Paul, MN
1. Which do we prefer: intellectual fairness or economic fairness?
2. Can we have too much fear?
3. What is meant by "calculate" in the Book of Revelations?
4. Can you rationally select a belief system?
5. Could there be alternatives to markets as the basis of American civilization?
6. Are there merit goods?
And tonight's question is:
How pervasive is the confirmation bias, or does anyone ever seek the truth? -- asked by Amr.
Amr: confirmation bias: we tend to confirm our biases. Bias free is possible but one must be careful: one must look for evidence opposed to your hunch in order to best justify your opinion.
Jon: science is just empirical. Science does not concern itself with pure reality. It "settles" for only what can be measured empirically.
Amr: philosopher Karl Popper said we don't start with data we start with a theory. A deductive theory, not an inductive theory.
Jon: what are the telltale signs of a good, sound truth? When do you trust a scientist's assertions?
Amr: if they show supporting evidence only, I mistrust their truth/conclusion.
John: are you saying you don't have a "causal mechanism?"
Steve: both deductive and inductive are used in science. As to bias, isn't that why it takes many researchers to do individual theses?
Jim: in common human endeavor, consensus decides. We settle for it.2. we must distinguish between correlations that are 100% true to "that means this causes this". Separate the soft areas from the hard ("snapshots" vs. time-long constants)
Dick: Climate change demonstrates an inferential conclusion
Amr: science would say carbon causes global warming, then one must try to disprove it if they wish to be most convincing.
Shannon: truth has more than one definition, more than one use. And causation is possibly unknown, even though a cause is believed to be found. There's always an assumption that truth is free from the individual but yet it can't be understood by anyone. For example, the elephant being described by 5 or 6 blind people, or the movie Rashomon
John: test method: randomly distribute possible causal factors. Looking for one thing and finding something else unsought but even more important/striking. "multiple comparison problem" (only occurs when there is no hypothesis to begin with) Much of the current economic debate is about confirmation bias. The truth of our present economic world is far too subtle and complicated to be understood, least of all by the political actors out there claiming certainty.
Steve: Shannon said truth can't be separated from the person. I counter by saying certain things like physics truths can be replicated among different individuals. We're limited to where we are, so Newtonian physics works perfectly in one context, Einstein's special/general relativity work in another, Bohr's Quantum Mechanics in yet another -- context context context
Shannon: truth isn't subjective, it's very difficult to know
Jon: Jim's here has been wonderfully honest with us in the recent past when stating that when he encounters factual data that contradicts his core beliefs he REJECTS the data. He rejects that bit of reality. That's a good demonstration of confirmation bias, eh? I personally think Jim's approach is nearly universal to the human condition, that we all have beliefs we are willing to deny reality over. I also believe we are equally capable of letting go of those beliefs in the exact moment we intend to experience the truth.
Jim: core beliefs should do that. They are supposed to stand against even reality.
Amr: phobias are said to demonstrate irrationality. But assuming people with anxiety are irrational is destructive because what may reduce their anxieties may itself be irrational (ie compassion). Irrational is in the eye of the beholder.
Shannon: of the ways to test a theory. Some ways are unethical.
Jim: Some things can not be proven. But my "core" is inviolate.
An excellent conversation all. Glad to meet Rob and Jean, hope to see you again.