I have always been a staunch supporter of differentiating between these two concepts, correlation and causality. But recently I have come to the conclusion that confusing the two is at the root of most of the evil in this world. Evil, for all intents and purposes, is defined here as ‘ignorance and the actions inspired by it’.

All too often, people see a correlation between two things and assume a causal relationship exists when there isn’t one, like Sex Ed and having sex, for example. Many people see the correlation between Sex Education and having sex (mostly because they both have the same word) and assume causality when there is none. Becoming more educated about sex doesn’t cause sex, and having sex certainly doesn’t make you more educated about it. There is no causal relationship between the two, only a correlation.

Those individuals who say that teaching abstinence is the best sex education constantly confuse correlation and causality. Basically, Sex Education does have a causal relationship with safe sex. More specifically, sex education causes a higher percentage of the sex being had to be safe sex. But it does not cause more sex to be had, unless you count the sex being had when the people having safe sex outlive the diseased uneducated people having unsafe sex. But the abstinence preaching people refuse to make this distinction and, instead, willfully or unwittingly confuse correlation with causality.

But that’s just meant to be an example. I’ll go further into the sex education issue in another post.

Does X Cause Y, or Vice Versa?
Sex education is not meant to be the focus of this article. Correlation and causality are the subjects at hand. The sex education example is meant to illustrate one of the basic concepts in correlation. When we look at the relationship between two things, it is rather easy to see that there is some kind of relationship (correlation) between them. But it can be rather difficult to discern the nature of that relationship, like when you try to determine if it is a causal relationship or not.

Basically, you have two entities, X and Y. You can see that there is a relationship between X and Y because everywhere that X is, so is Y. But without seeking a deeper understanding, you cannot tell if X causes Y, or if Y causes X, or if both are caused by Z, or if both are caused by completely separate but equally related events.

Another example… X is pressing the power button on your TV remote, Y is turning the TV on. Here, there is a clear causal relationship. X causes Y. Y does not cause X. And both are not caused by some other external factor.

But, what about when you press another button on that remote. Let’s say you press the ‘Channel Up’ button and the TV turns off. Is there a causal relationship? Usually the answer would be ‘No’. Channel Up doesn’t turn the TV off. We know this because the instruction manual says that there is no causal relationship. The Channel Up button advances the channel up by one channel.

However, there is a correlation, a relationship between the events, they happened at roughly the same time. In this case, we can be pretty certain that the TV turning off (Y) did not cause you to hit the Channel Up button (X). But if the remote is malfunctioning then it is possible that hitting the Channel Up button (X) caused the TV to turn off (Y). But it is also possible that the power went out, or a fuse was tripped, in which case an external factor (Z) caused the TV to turn off (Y).

So you see, assuming that correlation automatically means that there is a causal relationship can lead you down some very flawed paths, especially if you don’t bother to seek a deeper understanding. In this case, you might want to investigate whether your other electronics still have power or not. You might go to the fuse box and reset the fuse. You might check to see if the TV is still plugged in. You might hit the Channel Up button again to see if it turns the TV on, or you might turn the TV on and hit the Channel up button again to see if it turns the TV off. The point is, you see the correlation, and you investigate and seek a deeper understanding in order to determine the nature of the relationship. A correlation is present, but is there causality?

Flawed Paths
Correlation does not equal causality. It’s an easy trap to fall into. One that many people routinely fall into when reading news articles that talk about studies and results and data. In fact, I presented data in an article (click here) not too long ago, that easily could have been skewed by assuming causality from the simple correlation, which is why I made it a point not to do that.

But here’s the problem… Let’s say that you saw the data from my IQ vs. Religiosity article in a newspaper or on CNN’s website. What do you suppose the headline would have been? “IQ vs. Religiosity” doesn’t really inspire the masses to read it, unless they are particularly interested in the subject matter (what a concept!). But that’s not what News Organizations are interested in. They need to grab your attention, because they want your business (i.e. money).

So what would their headline be? I don’t hold corporate media in a very high regard so I would guess that they would completely bastardize the whole point of the data. Here’s a few examples I came up with:

Study: Religion leads to lower IQ

or

Religious Not As Smart As Others

or

USA: Successful Combination of Religion and High IQ

Can you recognize the flawed causal relationships put forth in those headlines? It’s not a stretch to imagine these headlines appearing on your favorite cable news outlet. But there is problem, those headlines are misleading. The confusion between correlation and causality completely falsifies the data, and makes it out to say something that it doesn’t.

This confusion is directly responsible for other things, like why you thought eggs were bad for so long, and all of the sudden they’re good for you again. For the longest time, it was reported that eggs caused higher cholesterol, but that conclusion was drawn from a simple correlation between eggs and cholesterol without further investigation into the causal relationship. Turns out that the correlation between eggs and cholesterol isn’t what made them “unhealthy”. Only when you added saturated fat to the mix, which caused the cholesterol to stick to the walls of your arteries, did eggs and their cholesterol become a worrisome product.

The Roots of Evil
What about more pressing issues, like politics, or war. Given the state of affairs today, one example comes to mind rather quickly: Terrorism and Islamic faith. The correlation between Islam and terrorist acts is undeniable, but all too many people are assuming a causal relationship without gaining a deeper understanding. Many think that the Muslim religion itself is to blame for terrorism. They scoff at those who say that Islam is a religion of peace, not unlike (I imagine) the Pagans must have scoffed at the Christians when they held themselves up to be a peaceful, loving folk during their quest to take over the world. They scoff because they see a very intense correlation, from which they assume causality. And from this, ignorance is born, because they think they’ve discovered the root of the problem - The Muslim religion, in this case. And since they discovered that, they stop searching and focus efforts on correcting (what they think is) the problem.

But what are the consequences if their assumption is wrong? It basically creates a problem for the express purpose of solving it, where there was none before. We create and define an axis of evil, which rivals the Saturday morning cartoon dichotomy between He-Man and Skeletor and thus further reduces the complicated issues to an oversimplified, child-like battle between The Good and The Evil. In this case, The Good are the ones who raped, pillaged, and burned a long time ago, and The Evil are the annoying copycats that decided to do it centuries later. But I digress…

The assumption of causality, in this case, has led to efforts that focus on the wrong problem, leaving the actual problem of the causes of terrorism to go unmitigated. Does Islamic faith (X) cause terrorism (Y)? Are both X and Y caused by some external factor? Or are both caused by separate but equally correlated factors, like geographical location (for X) and poverty/despair/education (for Y)? Using violence to fight violence, claiming hordes of lives that we define as collateral damage in our fight (not unlike terrorists in that respect) makes us look like hypocrites (evil) to most of the world. It is the ignorance and the actions inspired by it that are responsible for creating that perception.

So, what happens if we flip the script and look at the terrorist’s reasoning when it comes to correlation and causality. One of the central reasons for many terrorist attacks is American (or Western) occupation of foreign lands. And as a result their targets are often citizens or servicemen of these countries. This is born out of a correlation between the citizenship of their targets (X) and the occupation by the associated country (Y), a correlation where they see a causal relationship and find fault with their target to justify their attack. But is there a causal relationship between the civilians and their country’s occupation? I suppose it’s indirectly possible, but you can’t really know without being aware of their voting record, or their feelings/thoughts on the situation. Even the soldiers are not the cause, for they are only following orders. Again, a deeper understanding is required to know the nature of the correlation (relationship). And out of the ignorance that results from not seeking that understanding rises the true evil that murders innocent civilians.

A Brief History
How can we examine a correlation/causality conundrum involving a misconception about religion without addressing the false causalities within religion itself? This differs from other examples a bit, because religion itself is generally thought of as part of a search for deeper understanding. In other words, it is not that a deeper understanding is not a goal, it is that a true deeper understanding was, at that time in our history, nearly impossible to achieve. Therefore, the confusion between correlation and causality is still a function of ignorance.

Let’s use the Bible since that is what people in my country are most familiar with. The correlation exists between what the characters were saying at the time (X) and the events that were the subject of their “prophecies” (Y). In this case, there is not an immediate confusion between correlation and causality. That is to say that X was not thought to cause Y, or vice versa. Instead, in a quest for deeper understanding, both events were attributed to an outside factor, the influence of god. There are endless examples of events attributed to god that were beyond human understanding at that time, like the huge floods, the plagues, the pregnant virgins walking around, fire raining from the sky, etc.

I don’t want to go into the more rational explanations for the events attributed to god many many years ago. I just want to show how a lack of understanding contributes to applying a causal relationship even though the data does not back up the assertion. The point is, the less you know (read: the more ignorant you are) about how things work, the more likely you are to make the mistake of assuming causality based on nothing more than correlation.

The Gist
I could go into numerous other examples of false causality, like sexuality and rape, or immorality and homosexuality, or bad driving and utter stupidity, or your tardiness and the presence of red lights/police cars, or the length of your thesis and your computer’s probability of freezing. But I don’t think that is necessary to help you understand this concept, its significance, or its rippling effects any more than you already do, or any more than you can on your own.

My main goal is to illustrate the pervasive and everyday presence of this confusion between correlation and causality and the need for a deeper understanding to be sought by all.

Share This