• Question: I didn't understand the joke you told (about causality), can you explain it?

    Asked by to Ian on 19 Jun 2014. This question was also asked by .
    • Photo: Ian Stephenson

      Ian Stephenson answered on 19 Jun 2014:

      Good question…

      Correlation is when you measure two things and the APPEAR to be linked. For example playing video games, and getting arrested for violent crimes. If I took 1000 people you could ask them if they played FPS video games, and if they had ever been arrested. You might find that people who play FPS games are more likely to have been arrested.

      Newpapers LOVE this stuff. Someone does a study like this and find a correlation between games and violence. However the obvious reaction is “Wow – games cause people to be violent in real live, lets ban games”. But THAT IS INCORRECT.

      There’s an easy to make mistake on seeing that FPS players are more likely to get arrested than non-player, so say “hey – that’s evidence that video games make you violent”. An even worse mistake is to suggest that banning games would reduce violence, yet we see this all the time.

      The mistake is that correlation (people who play FPS get arrested more), is to assume that means games CAUSE you to be arrested more. It’s much more likely (but less good for reporters and people who don’t like video games) that people who are violent, like playing violent games, so being violent causes people to play FPS. In which its entirly possible that games REDUCE real violence, because at least some people who would go out and actually be violent decide to stay in and shoot there mates in a video game! Or maybe just more boys play video games, and more boys get arrested.

      All the study tells is is there is CORRELATION which is not the same as CAUSALITY. Scientific studies are careful not to mix them up, but journalists are less careful. Try reading some newspaper reports on science and you’ll find lots of jumping to conclusions.

      So the joke is:
      a:I used to think correlation meant causality
      b:really?
      a: but them I took a statistics course and now I don’t
      b: so the statistics class really helped?
      a:well… maybe…

      There’s a correlation between taking the class, and understanding the difference. However that DOESN’T necessarily mean it was the class which CAUSED person a to change his mind….

Comments