Friday, January 24, 2014

How the Rooster is Responsible for the Sunrise

I have some very bad news for you. Your way of forming beliefs is very flawed. And so is mine. It's a condition of being human called "cognitive bias". If you ever feel like crying one day, or just rocking for hours in a fetal position, Google "cognitive bias" and spend some time sorting through at how terrible humans are at processing and interpreting information.

There are hundreds of cognitive biases, some more destructive than others, but one of the more relevant ones is called false causation. Also known as "false cause", "correlation not causation" or for any logic nerds, "post hoc ergo propter hoc" (after that, therefore because of that).

Let's say you begin working at a new job. Your new employer brews coffee each and every morning for the employees. The coffee is some off brand, it's not terribly good, but it's free. You quickly discover each time you drink the coffee, your mouth swells up, you start sniffling, and you just feel poorly in general. You drink a different brand of coffee at home, name-brand, and this never happens. So you quickly conclude that this off-brand coffee is causing this reaction.

There is an actual correlation. Drink off-brand coffee at work, immediately have a bad reaction. If you avoid drinking the coffee at work, then you never have a bad reaction. This is a correlation. Our cognitive bias immediately leads us to conclude that the correlation is the cause.  Unbeknownst to us, it's actually the cheap, off-brand dish soap that is used to clean the coffee pot that is causing the allergic reaction, not the coffee.

This is a harmless example. However, this bias manifests much more strongly in our cognition in other, more relevant areas of our lives. Not the least of which is superstition.  Maybe if we all grow our beards, we will win the world series! You can bet your bottom dollar the entire Red Sox baseball team will be sporting beards on opening day in 2014.

This bias is why people of antiquity believed things like sacrificing children could control the weather, or that slaughtering livestock could ward off sickness.  It's also how politicians can easily fool people.  Stop me if you have heard this one before, "unemployment has gone way down since I have been in office" or "during the Congressman's term, violent crime has double in his state."

It's one of the oldest tricks in the book. Politicians take credit for the good, and distance themselves from the bad. Their opponents hold them responsible for the bad, and disconnect them from the good.  I titled the blog thusly because I remember a presidential debate when I was a kid. I can't remember who it was exactly or which party said it but, the phrase uttered was "my opponent taking credit for the improvement in our economy is like the rooster taking credit for the sunrise."  That statement always resonated with me, and it's a prime example of correlation and causation.

"There are three kinds of lies: lies, damned lies, and statistics." 

This phrase was made famous by Mark Twain. Twain speaks on how persuasive and misleading statistics are, mainly because of our cognitive biases. Unfortunately, this leads people to distrust any statistics that disagrees with their point-of-view because "statistics cannot be trusted."

Ironically, statistics are generally very accurate when conducted by a reputable organization. The reason that statistics are misleading is not because we cannot trust them to be accurate, it's that we cannot trust ourselves to make sense of the numbers because of the false causation bias.

Here is a good example, 99.8% of inmates in American prisons believe in God. That means only 0.2% of American prisoners are atheists, even though atheists account for about 10% of the American population. I have presented this information before to believers who cry, "bullcrap!" or "nonsense!" They absolutely refuse to even accept the correlation exists. That is because they believe that correlation equals causation (religion causes bad behavior, atheism causes moral behavior), the causation cannot possibly be true, therefore the correlation is not true.  This is a flawed way of thinking.  The correlation exists, but the causation probably does not.

Again this gets back to the idea that statistics are not unreliable in and of themselves, our ability to interpret and apply meaning to statistics is what is flawed and should be approached with apprehension and skepticism.  

Regarding the prison population in America, prisons are required to observe religious practices of inmates. Claiming a religion gives special rights and privileges to inmates that the non-religious do not receive. That is also why there is a high number of Muslim prisoners in America. Islam does not cause people to end up in American jails. When an inmate gets to jail, they often convert to Islam, because Muslims are permitted out of their cells eight times per day for prayer in many institutions. That is far more likely an explanation for the over-representation of theists in prisons and the under-representation of atheists.

We've Kicked God out of Schools, therefore School Shootings

I once saw a televangelist say that prayer was removed from public schools in 1962 by the Supreme Court. Then he began spouting off all of these statistics about how much worse things are in schools now. School shootings, drugs, etc. and claimed that we kicked God out of schools and therefore this is the results we get. This same mentality comes up after every school shooting in America. "We don't allow God in schools, so he can't save those children."

This is nonsense in the highest degree.

First of all, the fact that said televangelist is currently serving a 10-year sentence in Federal prison for 58 counts of tax evasion and fraud notwithstanding, his entire premise is wrong. Prayer was not removed from American schools in 1962. The Supreme Court disallowed faculty-led prayer or religious programs in publicly-funded schools. Since 1962, the Supreme Court has repeatedly and consistently upheld student's right to pray, form religious clubs, even use the school property for student-led religious events and meetings.  Public schools are required to allow students time for prayer.  The Supreme Court has been very clear on that.

Secondly, there were 24 school shootings in America in the 30 years prior to 1962. There were 25 school shootings in the 30 years after 1962. 

Thirdly, even if the fraudulent sleezebag's claims were accurate (and they certainly were not), then he has simply demonstrated a correlation and nothing more.  Determining causation is often really, really difficult.  In fact, when statistical studies are published the finders of said studies are required to offer numerous explanations for causation and then apply confidence levels to those explanations. The people who understand statistics and correlation the most, are the ones most apprehensive to state causation the most.

It should come as no surprise, then, the people who understand correlation and causation the least, are the most likely the ones to make a false connection between the two, and then vote a politician into office based on a flawed method of thinking.

No comments:

Post a Comment