Ready to Obey
The Dangerous Internet

Columbine and the Scientific Method

The Carriertom Into-wishen Research Institute is only about two inches from concluding that the scientific method is just a device for opinionated people to screen out evidence which points to conclusions they don’t want to hear about. For example:

Violent movies and television programs do not create violent viewers, says a University of Toronto professor who has just completed [Dec 2000] a comprehensive review of all of the research on the subject. "The scientific evidence simply does not show that watching violence either produces violence in people or desensitizes them to it."  [italics mine]

When junior Institute staff member Tom Fishandchips read those words, his heart sang. He ran down to the store and bought the Super Columbine Massacre video game. He’d long had his eye on it, but he was afraid to buy it lest people think he was violent. He loaded the game on his work computer and started spending all his off-time blowing sim-students to Kingdom Come, splattering blood and guts everywhere.

A blog reader might be repulsed at this point. Isn’t this post obscene, since real school shoot-em-ups are quickly becoming an American growth industry?

Get over it. There’s no proof, says the U of T report, citing scientific evidence. You got something against science?

Fishandchips’ continual playing soon got on the nerves of the other Institute members….guys like Sheepandgoats, Wheatandweeds, and Weedsandwheat. I mean, if you’re trying to write about God, it doesn’t help to hear assault weapons, sirens and SWAT team noise in the background. Driven to distraction, these eminent theologians also checked the research and came up with different studies, much to Fishandchips’ chagrin.

From the American Pediatric Association:

An APA spokesman testified before Congress in 2000: “Since the l950s, more than 3,500 research studies in the United States and around the world using many investigative methods have examined whether there is an association between exposure to media violence and subsequent violent behavior. All but 18 have shown a positive correlation between media exposure and violent behavior.” [italics mine]

The U of T study, upon which Fishandchips based his Columbine purchase, was one of the 18 clunkers! How could that be?

Note that the 3500 studies used many investigative methods. The U of T study used the scientific method. Could it be that the 17 other clunkers also used the scientific method? Could the scientific method be the least reliable when it comes to measuring people?

It‘s worth asking. If you practice the scientific method, you’re looking for cause and effect relationships. You think up experiments to test for such relationships. The experiments should be repeatable. Factors that would screw up the results, but are not what you are testing for, should be screened out.

These goals you can achieve in a laboratory. But it’s not so easy to do in real society. There are too many influences that go into making people what they are. Time, places, and circumstances might not be repeatable. And many relationships are like the chicken and the egg; they influence (and reinforce) each other. Fat chance you’re going to find out which came first, but unless you can do just that, the scientific method isn’t interested.

Sometimes with people, you must settle for correlations. Here are a few correlations we’re familiar with:

There is a correlation between condom non-use and sexually transmitted HIV.

There is a correlation between lead exposure and lower I.Q.

Between passive tobacco smoke and lung cancer.

Between calcium intake and bone mass.

Pediatricians accept all these relationships as fact, and practice preventative medicine based upon them. Yet, the correlation between exposure to media violence and aggressive behavior is stronger than any of the above relationships!

If we use the scientific method as a tool with which to investigate, fine. But God help us if we use the scientific method as the One True Tool….the only way in which we can know anything. In that case, there’s a lot of knowable things that we’ll never know.

Crestfallen, Fishandchips went back to the store to return Super Columbine Massacre. But he was too embarrassed to go in. He didn’t want them to think he was a violent person.


“Violent entertainment is aimed at children because it is profitable. Questions of right or wrong, beneficial or harmful, are not considered. The only question is ‘Will it sell?’”            Dr. David Walsh, author of Selling out America’s Children.


The LORD examines the righteous,
but the wicked and those who love violence his soul hates.
   Ps 11:5 NIV


To Sheepandgoats' surprise, the Super Columbine Massacre game is not a corporate creation for the purpose of raking in big bucks. It was created by an aspiring filmaker, who distributed it free and calls it an indictment of our times. "The game does not glorify school shootings," [it's creator] told The Washington Post. "If you make it far enough into the game, you see very graphic photos of Eric and Dylan lying dead. I can't think of a more effective way to confront their actions and the consequences those actions had."

A modern day Alfred Nobel - dynamite story, perhaps?

Defending Jehovah’s Witnesses with style from attacks... in Russia, with the book ‘I Don’t Know Why We Persecute Jehovah’s Witnesses—Searching for the Why’ (free).... and in the West, with the book, 'In the Last of the Last Days: Faith in the Age of Dysfunction'


Romulus Crowe

Hi Tom

You know, all those 3500 studies used scientific methods. There are many, many scientific methods ;)

You highlighted one of 18 studies that showed no correlation. (and Dec 2000 doesn't count as 'just completed' in science. That's history). Then you stated there were 3482 studies that said there was a correlation.

To a scientist, that says that yes, the correlation is true. More than 99% of studies support that conclusion. In any large-scale dataset you'd expect to get a few duff results. Better than 99% showing one result over the other is about as good as it gets.

I've noticed a great railing against science in many religions lately. I know there's this big fight between evolutionists and creationists going on, but in fact very few scientists study evolution at all. I don't.

I think evolution might be right, but I haven't studied the subject so I can't argue too well on the case. I'm not an evolutionist. I'm not even a biologist. Neither are the people running those studies.

Tarring all of science as 'wrong' because of one aspect that clashes with religion is no different than blaming every Christian for the Inquisition.

More than 99% of studies say there's a correltauion between violent media and violence in those who watch it. That's a scientific conclusion, based on scientific method.

We don't get everything wrong ;)


having read your post, i have a few issues with some of the things you said. one, i agree with romulus crowe, i have seen the religious trend to attack science at any provocation, and in most cases, i find it to be for non-sensical reasons. two, just because there is a corelation between to items does not mean that it is a cause and effect relationship, if that were true, then icecream would be illegal. as the sale of icecream rises, so to does the homicide rate. the answer is not that icecream causes murder, but that both sets of numbers rise when the temperature rises. when i observe the nature of our society, i dont see as much violence in it as would be expected if violence in media was a cause of real world violence. yes, just as i will admit that there are some people who would kill because of icecream, i would admit that there are those who would kill because of violent media, but i would also note that these are few and far between. what if an affinity for violence carried over into all aspects of an individuals life? then you would observe that they exhibited violent behavior, as well as consumed violent media. statistical correlations need to be taken with a grain of salt, and an awareness that there are numerous other factors that may need to be taken into account.


Hey Tom,

I caught this one browsing through the recent comments on your blog. I can't say that I'm not a bit jealous at how many hits you get. Maybe I should start writing more controversial bits, or matters dealing with entertainment and recent culture?

Anyway, I did my undergrad thesis on a topic very similar to this, and my mentor's area of study is media and aggression, so I have some familiarity with the literature. What you state is basically true - there is a fairly strong correlation between violence in media and aggressive behavior, but the story gets a bit more complicated. First off, almost every study looking into this studies short-term behavioral modifications. The studies that have examined long-term relationships are a bit of a crap-shoot. Some show positive correlations, some show negative correlations. It's also important to note that many of these are looking at "aggression" as opposed to strictly "violent behavior," which also goes to complicate matters a great deal.

In my personal opinion, and in the findings of my research, yes there is at least a short-term increase in aggressive behavior after exposure to violent media. There are a host of factors that contribute this correlation, among them being 1) identification with the violent character; 2) demonstration of expert knowledge/skills; 3) interactive rewards for aggressive behavior; and many others. To put it simply, reading a story about a villain who is absolutely repugnant and where the violent behavior is never actually described in detail is no where near as effective as playing a game where you can strongly identify with the character you are playing, where it gives detailed visualization regarding the execution of violent behavior, and where you get a reward, like points or achievement. The effect is also typically stronger in males, but there could be some selection bias involved with that.

So, are violent video games and television shows promoting violent behavior? It's possible. As I said, there is a short-term correlation for aggressive behavior, which may or may not transfer to violent behavior. Of course, the same could indeed be said about temperature, as Ricecake above notes - there is an incredibly strong correlation between environmental temperature and aggression. Above 90 degrees or so, and violent behavior spikes sharply, and thus we see yearly peaks in homicide rates during the warmest months. Temperature we can't do much about, but media perhaps we can.

Part of the problem, however, is effect size. To put it bluntly, literature, be it philosophical, theological, or political, has had a much larger effect on promoting violent behavior than any video game or television show we know of. Video games don't seem to have the same effect on inspiring entire countries to conduct war on other people as the great works of literature and religion have had in the past. Likewise, the demagogues of the news media, like our dear friends O'Reilly and Beck, have a much easier time promoting unrest than watching episodes of 24 back to back. There are likewise a host of factors as to why this is the case, but I won't dwell on that here.

So what should we do about violent media? We could ban it, and censor every book that has ever been in print (which would likely include most religious texts...there is an awful lot of blood and guts and calls to exterminate peoples in those books), and hope that we could create some new literature that would promote peaceful behavior. I think that approach would fail, and quickly. Part of the problem is that you're always going to have some people who are just crazy and lack the ability to feel human empathy. Part of the problem is the host of other factors that induce violent behavior. And part of the problem is that total ideological power has always turned violent eventually. There is something to be said for the release valve of letting people speak their minds occasionally.

On the other end, however, there are some positive things that we could do about video games and media. I saw a presentation last year at a conference I attended that discussed a finding using a game very similar to World of Warcraft (which, in my own research and in others have shown some positive effects for sociability and communication). Instead of rewarding only completing quests and killing creatures, the researchers added an avenue by which players could gain experience (reward) through helping other players or healing them. This ended up creating a strong negative correlation between game-play and aggressive behavior, even though players were still killing other creatures and completing quests. It's an interesting find, and if I hear anything about follow-ups to this study, I'll let you know. As it stands now, in my own opinion, I think that it would be a good idea to include elements like this into traditionally violent genres, but I don't think they should be banned outright. Even though there is a strong correlation in almost every study we've done between violent media and short-term increases in aggressive behavior, the larger question of long-term effects and translation to actual violent behavior is a lot thornier.


The comments to this entry are closed.