Getting Desensetized

There's a new study forthcoming in the Journal of Experimental Social Psychology, which supposedly finds that people who play violent video games experience an increase in aggressive behavior and a decreased reaction to violent imagery.

"It's already well known that playing violent video games increases aggressive behavior and decreases helping behavior," said Brad Bushman, a researcher at the University of Michigan. "But this study is the first to link exposure to violent video games with a diminished reaction to violent images."

The research was conducted by Bushman, a U-M professor of psychology and communications studies and a faculty associate at the U-M Institute for Social Research (ISR), along with fellow researchers Bruce Bartholow, an assistant professor of psychology at the University of Missouri-Columbia, and Marc Sestir at the University of North Carolina at Chapel Hill.

"Most of us naturally have a strong aversion to the sight of blood and gore," Bartholow said. "Surgeons and soldiers may need to overcome these reactions in order to perform their duties. But for most people, a diminished reaction to the effects of violence is not adaptive. It can reduce inhibitions against aggressive behavior and increase the possibility of inflicting violence on others."

On the surface, this study looks potentially damning for the video game industry. However, after a closer look at the details of the study, one can't help but question the results. Take a look at some of the research techinques used to come up with these findings:

For the study, the researchers asked 39 male undergraduates how often they played their five favorite video games, and how violent the games were. The researchers also assessed participants' irritability and aggressiveness, asking them how much they identified with statements like the following: "I easily fly off the handle with those who don't listen or understand" and "If somebody hits me, I hit back."

Next, the researchers outfitted participants with electrode caps to obtain EEG data, including the average amplitude of a particular type of brainwave, known as P300, which is believed to be an indicator of how people evaluate a stimulus, such as a photograph. After doing so, the researchers showed participants a series of images. The content of the images was emotionally neutral (a mushroom, a man riding a bicycle), violent (a man holding a gun to another man's head) or negative but nonviolent (a dead dog). While participants viewed the images, their brain waves were recorded.

After viewing the images, participants were told they were competing in a reaction time task with another person to see who could press a button faster following a tone. The slower person would supposedly receive a blast of noise through a pair of headphones, with the intensity and duration of the blast determined before each round by the previous round's winner. Actually there was no partner, but participants' tendency to administer long, loud blasts of noise is a widely used, reliable measure of aggression.

What the researchers found was that participants who routinely played violent video games responded less to violent images, as measured by a diminished amplitude of their P300 brainwaves. But this was not true of their response to other, equally negative, nonviolent images. The researchers also found that the smaller the brainwave reaction to violent images, the more likely participants were to behave aggressively in the reaction time task by blasting their "partner" with loud, unpleasant noise.

Read at face value, the research seems a bit flawed to me. The noise test is a key example. Basically, each of the 39 male participants were told to hit a button as fast as possible. The loser would get a blast of loud noise ... with the volume and length of the noise determined by the winner of the previous round. If the blast is detemined over a period of sequential rounds, then each "winner" knows that he's subject to losing the round and getting hit with the consequence he set. On the other hand, if the participant thinks that he's setting the level for the next set of competitors, he's still likely to set the bar at the same level he would expect to be given to himself. Truthfully? That's not violence ... that's competition at best, and pissing contest between guys at worst.

As far as the image test goes, these participants were college students ... not teenagers. Before I graduated high school, I was already reading pulp fiction stories by the likes of Raymond Chandler (a great author) and Dashiel Hammett. I also had seen much worse than a picture of a man with a gun to his head in my history classes. Would a picture like that get much of a stimulus reaction from me? Probably not ... though not because of anything like video games. And to be perfectly honest, I've been around guns before and even had them pointed at me. And yeah, I'm sure THAT caused a peak in my P300 brainwaves at the time. By their own admission, there was no change in how participants were affected by the "negative but nonviolent" imagery.

I'm more than a little curious to take a look at the full study once it's published. Unfortunately, with all of these studies trying to prove one thing or another, I've become more than a little desensetized ... not to violence, but to these damned studies. Here's the frightening thing about studies, with a little time, money, and ingenuity, and with the right test group, I could put together a study showing that cute and fuzzy bunnies can drive people to acts of road rage.

Too often, researchers go into a study looking for a particular answer. Many times that expectation of a predetermined outcome causes those responsible to skew things in their favor, intentionally or not. The scientific method is supposed to begin with the question, not the hypothesis:

• Define the question
  
• Gather information and resources
  
• Form hypothesis
  
• Plan experiment
  
• Do experiment and collect data
  
• Analyze data
  
• Interpret data and draw conclusions that serve as a starting point for new hypotheses
  
• Communicate results

This method is generally meant to be cyclic in nature. Once the data is interpreted, new hypothesis can be formed an tested as well. I've seen far too many cases though in which researchers spend too much time trying to prove their hypothesis, instead of letting the raw data lead them to new ideas and interpretations.

Anyway, that's enough of a soapbox moment for the moment. What are your thoughts?