Warning: Arm Flapping Leads to Heart Attacks

by Maxpower & Beauty

Nothing is more irritating than people who confuse correlation with causation.

It happens more than you may think, everyday people are drawing erroneous conclusions from their observations. This does not only infect unscientific areas such as pseudoscience but has also been highlighted in some respected peer reviewed scientific journals.

For those of you who may have slept through your high school science courses;

Correlation measures the patterns of responses between two variables, either positive or negative. This means that there is a relationship between the items in question. For example watching violence on TV is related to an increased preponderance to violent action in children. Correlation is easy to see on an everyday basis, you don’t need to have a study or a model set up, anyone can see the relationship.

Causation happens when one action leads directly to another, for example when given a general anesthesia you will pass out within seconds. It is the drug and nothing else that is causing you to pass out. However, proving that one thing caused another is difficult. You need to show that nothing else could have influenced the result in any way. Going back to the TV example, it has been shown that children who watch violent TV programs are more likely to engage in violent behavior. They are positively correlated– more violent show watching, more violent actions. This does not mean that watching violent TV shows causes a person to be violent. It could be that bratty kids that hit their siblings prefer to watch violent TV shows. Perhaps the kids that watch the violent TV shows have parents that are violent. The only way to conclusively prove that watching violent TV programs causes violence in children would be to conduct an experiment to eliminate all of the other possible causes.

One of the movements in which I most strongly associate the confusion of correlation and causation is the environmental protesters. In the case of global warming through the increased release of so-called ‘greenhouse gases’ the environmental organizations have unwittingly or purposefully mixed the definition of correlation and causation.

Environmental organizations will be the first ones to point out that there is a link between the amount of greenhouse gases released into the atmosphere and the slight overall warming trend we experienced in the last half of the Twentieth Century. On the face of it, and this is how the groups want you to take it, there seems to be a strong association. Hey look, there are more cars out on the roads now than in any point in history, and those environmental groups are saying there is more greenhouse gas emissions now than at any point in time. Well that must mean that we are causing the increase of temperature, I think this is a classic case of human’s overestimating out impact on the universe. Just because we feel in control of our surroundings that doesn’t mean we know how to impact the weather patterns of the planet short of triggering a nuclear winter.

While there may be a correlation between the amount of greenhouse gases released into the atmosphere and the amount of cars on the road, there is absolutely no established causation. How can anyone be certain that when I take my car for a spin I am contributing to the end of the world? (beyond the inevitable entropy which will occur when all energy is burned up in the universe) You can’t. The model that you would need to set up and then conduct regressions on would have an infinite amount of different factors which could impact. What about the variation in the sun’s magnitude which occurs in cycles? Interestingly the amount of the energy arriving all over Earth in one second is equivalent to the amount of energy humans uses in a year. Or what about natural variations in climate which go back into prehistory (the ice age being the last example) and are even documented as happening on human record? The warming of the earth in the medieval period is well documented historically.

The latest example of this fear-mongering in the media is that giant iceberg which just broke off from Antarctic. Did anyone listen to the media describe it? ‘A giant chunk of Antarctica broke off yesterday in a dramatic realization of the problems associated with global warming.’ They all ignored the scientists comments that while this part of the ice sheet was getting thinner, many other places were at record ice depths. Maybe the increasing amounts of greenhouse gases have actually increased the thickness of some parts of Antarctica. I could prove that with correlation.

So even while pseudoscience environmentalists can not prove a causation between greenhouse gases and the warming of the earth, many take it one step further with the help of the media. This is what really frosts my globe, every time there is some sort of unusual weather activity, whether it be a record high temperature set on one day, or a freak snow storm in July, its always linked to global warming. Oh sure, so now based on a tenuous link between some gases and the warming of the planet we link the unusually warm (or wet, or cold, or dry) weather at any one point in the world at any one point in time with the systematic warming of the atmosphere by humans? This breaks down to the idea that prior to humans there was no freak weather at all, because we cause it all don’t ya know.

Even scientists, people you think should know better, can confuse correlation and causation. A paper was published in the respected British Medical Journal (1997, Vol 314 for those with too much time on their hands) that concluded that people who move their arms around a lot are prone to developing coronary heart disease. Huh? So if I wave my arms around I am risking a heart attack? Excuse me if I don’t believe you. You may be able to convince me that the two are related, even if only coincidentally. You cannot convince me that the former causes the latter. The author does at one point state that other factors may also impact the development of coronary heart disease. Yeah thanks, it also says he did not receiving any funding for this project and it shows.

If you are not into reading academic journals, you can find confused people mixing up correlation and causation in the popular media. Flipping through a newspaper you could find headlines like: people who travel a lot get more cancer; taking the Pill leads to cervical cancer; eliminating X food from your diet will make you lose weight. I think you are getting the idea. I’m sure that you could think of many other things that could explain these relationships besides the ones that are mentioned.

You are probably thinking, “So what’s your point? Who cares if people mix up correlation and causation? You are boring me.” Well, it is actually quite important. Many individual people, pressure groups and policy makers make statements about an observable relationship, claiming it to be causal. This has serious implications for policy, laws, punishments, and so on. These people are trying to convince you that they are right, hoping that you won’t be able to prove otherwise (or hoping that you couldn’t be bothered). You need to be skeptical about everything that you read or hear, especially if someone claims that their product, idea or action will make something happen. Unless you can prove so with an experiment or complicated linear regression model, it is likely untrue. Something very trivial could account for or influence that outcome. Question everything. Use your brain. Flap your arms as much as you want.

Maxpower & Beauty