Thursday, August 28, 2014

Video Game Science - Correlation vs. Causation

Hello again, Parents and Gamers!

As I'm sure we all saw it coming, little to nothing has happened in the video game industry over the course of this Sunday. However, rather than going our regular route of simply leaving you with a game from Kongregate, we decided instead to do a little write-up. One of the most common questions that we receive whenever we cover scientific research about video games is concerning correlation and causation. More specifically, we are often asked why we implicitly specify if results are "correlative". Allow us to explain.

Fun with Graphs! Thanks to

When referencing scientific research, whether it be about video games, movies, music, medicine, psychology, or any other field, the two of the most common words you will hear when results are reported are "correlation" and "causation"; but what do they mean?

Sadly, a fair number of people take these words as meaning the exact same thing. They do not. In fact, assuming that correlation and causation are the same can lead to such incredible scientific error that mixing these two up in, say, medicinal research, could have dire results. Allow me to explain.

Correlation is best defined as a "mutual relationship between two or more things". In other words, if we are talking about scientific research, there will be an obvious connection between two actions or reactions. To give you an interesting example, consider this fictional piece of research: over the past 30 years, the length of the average lady's skirt has gotten shorter and shorter. Coincidentally, the average number of home-runs hit by the New York Yankees has progressively gotten higher and higher. Now just hold onto that one for a second.

Now, let's consider Causation. The simplest way to describe causation is "the relationship of cause and effect". Can't get much more simple than that. In scientific research, that would mean that one thing that is happening is clearly leading to another thing happening; an action leading to a reaction. To provide the same example, allow us to consider the New York Yankees again: over the last 30 years, the team has progressively adopted more home-run hitters into the ranks as well as a number of coaches that have really made a focus on working on the team's batting average, in turn the Yankees have progressively gotten more and more home-runs.

Both of these examples are technically relationships, therefore they are correlations, however only one is a cause and effect relationship.

This now brings us full circle. In this fictional example, we have seen two linked trends. On the one hand, we have correlation stating the the length of women's skirts and the number of home-runs hit by the New York Yankees over the last 30 years; on the other hand, we also have a correlation between the number of home-run hitters and coaches that have been hired on followed by the number of home runs hit. Common sense tells us that the increase in overall talent has led to more home-runs, nonetheless the data is still there suggesting the short-skirt relationship.

Now you ask, what does this have to do with video games?

You may be interested to know that, despite the multitude of research and reports that are published or talked about in the media, science still has not determined what effect video games may have on players. There have been dozens of studies that have been posted on both sides "showing" that video games are both bad and good for you. How can this be?

Again, this is a matter of correlation versus causation. For all of the claims that researchers have made about video games, very few have ever been able to prove causation. This means that the video-game research has found trends and relationships between certain factors, let's say violence in video games and tendency towards violence in real life, however they have been unable to prove that one thing is causing the other. Is it the violence in the game that is causing it, or is it possibly a violent family history? Is the individual turning to the violent game BECAUSE of a tendency towards violence already? Despite all the speculation, these factors have yet to prove cause and effect.

Before you ask: No, I'm not trying to make a statement concerning violent video games and violent tendencies; I just want to use an example that most people will be able relate to concerning video game science.

So, from now on, whenever you hear reports of scientific research "proving" something, whether it be video games, psychology, whatever: look into it! The research can most certainly show a link between the factors considered, i.e. a correlation, but does it prove causation? Most journalistic reports, whether on the TV or Internet, will always reference where the research is being published, thereby allowing you to find out for yourself. If you're interested in the research, take a quick look and you can figure out whether the researchers proved cause-and-effect or if there is simply a link!

Referring Articles:


Hopefully this has been helpful for anyone interested in research. We would like to thank you for joining us again and encourage you to stop by tomorrow for a new article of The News. As always, if you have any questions, simply shoot it to us in an e-mail at the address provided or just leave it in the comment section below. We will get back to you as quick as we can. Thanks again and we hope and have a wonderful day.

No comments:

Post a Comment