The three follies

[nc-1701]

Recently I've been trying to apply 3 checks to things that I think, read, and hear. My results so far have somewhat mind opening but much more misanthropic. The idea is this, a great amount of what people accept as truth is built on one of these mistakes. Avoid them and you can "take the red pill" and think more accurately.

1) Confirmation
You see an article or statistic that confirms a belief you already have or directly opposes it. Do you require the same burden of proof in both instances? Can you justify why not?

Spoiler Example :

I occasionally share articles on Facebook that support my political views. Often I'll see something and immediately love it, but then when someone starts criticizing it intelligently I'll realize it was weak, poorly, written, and not very well constructed. I subconsciously overlooked these problems because I liked that it confirmed my views.

2) Observation
If you think something has changed, has it really changed or are you looking at it differently? Can the change be explained by a change in the conditions of observation.

Spoiler Example :

Most people believe the crime rate is increasing, it actually isn't, in fact it's at historic lows in much of the world. Certainly in the United States, but as media markets have expanded the number of crimes an average hears about has increased, thus leading a rational person to conclude that crime has increased.

3 Randomness
Can this thing that you think is meaningful be explained just as well by random coincidence? Human nature is to never believe in coincidence and always search for meaning. The truth is a lot of the things we consider meaningful are the result of random chance.


Anyway I've been trying to apply these 3 checks to things I say/think/do/read/write, with some success. Thought I'de share and hopefully hear tips on how to better recognize these things when we do them. It is after all very easy to pick apart others for these mistakes, while being completely blind to the fact you are making the exact same ones.

 

[SS-18 ICBM]

I think I've internalized #2 and #3 too much such that I'm afraid to make any statements without the qualifier "probably" or something similar.

 

[uppi]

1) Confirmation
You see an article or statistic that confirms a belief you already have or directly opposes it. Do you require the same burden of proof in both instances? Can you justify why not?

Extraordinary claims require extraordinary evidence. If I have an informed belief about a subject and there is a mountain of evidence supporting it, I am going to be much more critical of a single peace of evidence opposing it than I would be about a piece of evidence supporting it.

If there is contradicting evidence, there has to be a mistake somewhere. And there is a very high probability that the mistake is hiding in the single piece of evidence instead of the mountain of evidence being all wrong.

It might seem to be a folly at first glance, but it is necessary to make any progress. I would not be getting anywhere if I questioned the laws of physics every time a detector is broken. It is going to take me much more than that to propose that the laws of physics need to be amended.

3 Randomness
Can this thing that you think is meaningful be explained just as well by random coincidence? Human nature is to never believe in coincidence and always search for meaning. The truth is a lot of the things we consider meaningful are the result of random chance.

The problem is that this question can always be answered with yes. No matter damning the evidence is, there is always a non-zero chance of it being a random coincidence. The question is, where do you draw the line, where you believe it to be meaningful? That ties in with the first point: If you believe the conclusion, you are more likely to dismiss the probability of it being a coincidence.

 

[Borachio]

I think I've internalized #2 and #3 too much such that I'm afraid to make any statements without the qualifier "probably" or something similar.

Yes. This is possibly true for me, too.

Which can make convincing someone else of something that I'm almost certain is true almost impossible. I start having to include statements like "as true as I'm standing here". And then they start thinking I'm weird. (Which I am.)

 

[dusters]

I don't have anything to add to the topic, but i want to say that this is like a basic to critical thinking 101.

As a philosopher, one is trained to analyze, criticise and proof read every single statement, every single argument and check if the premises are awkward.

However, for a normal human being (who is not a robot who overanalyzes why the Sun seems to shine 5% brighter today compared to yesterday ), i expect that there is information which is nearly vital to be checked and then there is information which one can be put into "not sure, i can't be bothered to gather enough data to check if true/false" box.

 

[nc-1701]

Extraordinary claims require extraordinary evidence. If I have an informed belief about a subject and there is a mountain of evidence supporting it, I am going to be much more critical of a single peace of evidence opposing it than I would be about a piece of evidence supporting it.

If there is contradicting evidence, there has to be a mistake somewhere. And there is a very high probability that the mistake is hiding in the single piece of evidence instead of the mountain of evidence being all wrong.

It might seem to be a folly at first glance, but it is necessary to make any progress. I would not be getting anywhere if I questioned the laws of physics every time a detector is broken. It is going to take me much more than that to propose that the laws of physics need to be amended.

Let me give you a different example. Suppose you have a belief, suppose further that this belief is in fact accurate and true.
Now you come across an article making some claims which support your belief. You may quickly accept the article without proper scrutiny because it supports the belief, which you know to be true.
But in fact the article is poorly researched and it's conclusions are entirely false. Note this says nothing about our original belief, which is still true, but we would do well to still give the article proper scrutiny. Lest we expand our belief into a more expansive one that isn't true or begin supporting it and arguing for it using arguments and justifications which are not true.

The problem is that this question can always be answered with yes. No matter damning the evidence is, there is always a non-zero chance of it being a random coincidence. The question is, where do you draw the line, where you believe it to be meaningful? That ties in with the first point: If you believe the conclusion, you are more likely to dismiss the probability of it being a coincidence.

The last part is what's dangerous to clear thought. In general you have to be able to at least very crudely estimate how likely it was to have been random chance. In many cases we as humans dramatically underestimate this and assign meaning to things which are not meaningful.
But even improbable events have to happen, and it's best we keep that in mind. Getting struck by lightning is very rare, but lightning strikes many things every day.

 

[Borachio]

This isn't a bad book on the subject.

I suppose it's a light introduction to critical thinking generally.

 

[uppi]

Let me give you a different example. Suppose you have a belief, suppose further that this belief is in fact accurate and true.
Now you come across an article making some claims which support your belief. You may quickly accept the article without proper scrutiny because it supports the belief, which you know to be true.
But in fact the article is poorly researched and it's conclusions are entirely false. Note this says nothing about our original belief, which is still true, but we would do well to still give the article proper scrutiny. Lest we expand our belief into a more expansive one that isn't true or begin supporting it and arguing for it using arguments and justifications which are not true.

I agree, and I certainly do not want to give the impression that we should not give a claim its proper scrutiny. In fact I could readily cite a few article in poorly peer-reviewed journals, where I agree that what the authors claim is (probably) true, but their data is by no means sufficient to show that.

But I would say, that proper scrutiny depends on what I believe about the claim. If the result is what I expect, I do a few cross checks and then accept it. If I consider the result to be impossible, I really go out of my way to find the mistake and spend much more time on that, before I am convinced that the result is true (unfortunately, so far I have always found the mistake in such cases)

The last part is what's dangerous to clear thought. In general you have to be able to at least very crudely estimate how likely it was to have been random chance. In many cases we as humans dramatically underestimate this and assign meaning to things which are not meaningful.
But even improbable events have to happen, and it's best we keep that in mind. Getting struck by lightning is very rare, but lightning strikes many things every day.

Sure. Measuring random noise is a great case study how the human mind finds patterns, where there are none. I am just saying, that if you expect the effect, 95% confidence might be sufficient, but if you do not expect it, you might want 99.9%. And in fact, there are mathematical arguments why this is exactly what you should be doing.