Thursday, 13 December 2012

Violent Videogames: A lacklustre study

So, this week news broke that a researcher from Ohio State University named Dr. Brad Bushman has found a link to long term increases in aggression from playing violent videogames. A brief rundown of his findings can be found here.

The media reacted to this (at least in Britain) about as well as you'd imagine.

The Daily Mail: You scared yet?Source

Now, I'm not gonna jump up and down or shout loudly about how this college professor just doesn't understand Gaming. The fact is this doesn't do anything for the argument and even if he did have those views, recent events have shown them to be not entirely baseless.

No, instead I'm going to look at the study from what details have been released and try to show errors in his method. To preface the following I need to say that I am not a Psychologist, nor do I have a Doctorate like Dr. Bushman. I do have a Bachelors of Science and a understanding of scientific methodology though so I'm not completely shooting in the dark.

1)The link to Long-Term Aggression

A central claim of OSU's report is that this is the first study showing a increase in long-term aggression instead of a limited short-term boost. The problem is, the study only went on for three days. Compounding this, tests were conducted immediately after the games were played at the end of each day. Quoting the article:

"The results showed that, after each day, those who played the violent games had an increase in their hostile expectations"

 Putting aside that 'hostile expectations' is not the same as Aggression (one is an internalised view of the world, the other a very much external reaction to it) we're only really showing duplication of previous data that showed a long stint of playing these games increased short term aggression. Dr. Bushman himself says:
 “I would expect that the increase in aggression would accumulate for more than three days. It may eventually level off. However, there is no theoretical reason to think that aggression would decrease over time, as long as players are still playing the violent games"

Expect, may and theoretical are all words used when you cannot say something for certain. This is a common problem in every science because the margins of error in any experiment will always stop you being certain. However, here he has based a hypothesis (Aggression has a long-term buildup in violent game players) on those expectations while admitting that it's possible for aggression to level off over time. His final statement 'there is no theoretical reason...' is the most open for criticism. The existence of other studies showing only a short term rise in aggression is very much reason to expect a decrease over time.

"We would know more if we could test players for longer periods of time, but that isn’t practical or ethical"

Really? I think I know alot of Gamers who would be willing to go through a extended test pro-bono if it helped put this kind of silly nonsense to bed once and for all.

2) Aggressive Behaviour

As I mentioned above, the initial test only actually showed an increase in 'hostile expectations' through the test subjects filling in a ongoing story (this is the kind of silly stuff you have to do in the Social Sciences). The proof of actual aggression comes later on, as follows:

"Students in the study then participated in a competitive reaction time task, which is used to measure aggression. Each student was told that he or she would compete against an unseen opponent in a 25-trial computer game in which the object was to be the first to respond to a visual cue on the computer screen.

The loser of each trial would receive a blast of unpleasant noise through headphones, and the winner would decide how loud and long the blast would be. The noise blasts were a mixture of several sounds that most people find unpleasant (such as fingernails on a chalk board, dentist drills, and sirens). In actuality, there was no opponent and the participants were told they won about half the trials.)" (Subjects who played the violent games subjected losing opponents to longer and louder blasts then those who did not)

Wait, they won about half the trials? So in other words the other half of the time they were subjected to a blast of incredibly unpleasant noise?

Let me frame the problem with this as simply as possible. Did you ever play a game of slaps with a friend or sibling? Things start off fairly light until, a few goes in, something snaps and the slapping gets harder and harder. It's human nature to react when you are attacked either with flight or fight. Without some more information about the trends in both groups I find it hard to see how you could possibly eliminate one group simply having more people in the 'fight' category then another.

Certainly it seems dubious as a basis for establishing that these people are more aggressive then any other person getting a chalk board in the ear.

Buckle up Timmy, Here comes the pimp hand

3) The Test Group

"The study involved 70 French university students who were told they would be participating in a three-day study of the effects of brightness of video games on visual perception."

70 people. From this they are claiming to have evidence of a trend that is common throughout people who play games like Call of Duty. For the record, Call of Duty Black Ops 2 sold 7.4 million units in the US on launch day. Even better, Activision were disappointed in that number because it was down 14% from previous launches.

Most studies take a much larger population then this before they make such sweeping claims, and this should be doubly true of a subject like this which is bound to get snapped up by the media.

Apart from that we're talking about 70 university students. I understand the reasoning (16-25 being the largest age group of gamers) but lets face it, as this picture of my buddy Carey shows none of us are at our most emotionally stable during university.

Carey: After roughly *all* of the beers

I can't actually decide which is worse, that he asked to be included or that there are much, much worse photos out there.

I'll have to wait until the journal is published "at some future date" to get my hands on their actual data but for the moment I hope I got at least a few of you to agree there are some serious questions Dr. Bushman and colleagues need to answer.

For now let us say Skål! and drink together.

1 comment:

  1. I think the study's a bit more subtle than you're giving it credit for. If you look at the abstract from the Journal web page, you'll see that the crucial point was that the level of aggression displayed by violent video game players was higher on the third day than on the second, and higher on the second day than the first, thus demonstrating that having played violent video games on the first day was still affecting behavior on day 3. That's the long-term effect that the article demonstrated.

    The ethics point is a serious one: if it is genuinely increasing aggression, it's not ethical to do that to people for a longer period just to make a point. Obviously, just using volunteer gamers would not give you a random sample.

    This isn't the definitive study, and I don't think the author would claim that it is: but it does indicate that the effects of playing violent video games can extend over multiple days, which is a legitimate topic for further investigation.