President Obama is asking Congress to fund further research into the root causes of gun violence including -- as he famously singled out -- violent video games. But is more research actually necessary? And should the video game industry support it? We asked researcher Cheryl Olson, co-author of Grand Theft Childhood and an attendee at last week's meeting with Vice President Joe Biden, to share her thoughts.
This week, President Obama called for "research into the effects that violent video games have on young minds."
"We don't benefit from ignorance," he said in a nationally televised speech. "We don't benefit from not knowing the science of this epidemic of violence."
The Supreme Court settled this. Some researchers insist that there's evidence of a "causal link" between violent video games and aggression. So, what do we know, and where are we still in the dark? And why do those media violence researchers continue to debate and disagree?
To answer this, let me explain how I got entangled in this debate, how research works and what it shows, and what this means for game developers.
The research I've been involved in
Around 2002, a Congressman saw news about Grand Theft Auto 3, worried it was degrading our society and promoting youth violence, and ultimately made funds available to study this through the U.S. Office of Juvenile Justice and Delinquency Prevention. I wrote a proposal and won a $1.5 million grant to study the effects of violent video games on middle-school kids. At the time, all I knew about video games came from watching my son play Pokemon and StarCraft. (My expertise was in media influences on health; I'd produced videos on health and medicine before returning to grad school.)
This is typical. There's no central organization out there that decides what researchers ought to study. Somebody -- a government, foundation, corporation, or private individual -- makes funds available to study a topic, or a researcher has a study idea and goes out looking for funds.
Although science is supposed to be objective, often the funder or researcher has an agenda or point they hope to prove. Researchers get ahead in their careers by doing studies and getting the results published in academic journals. And researchers will plan studies differently depending on their background: for example, experimental psychologists do laboratory (actually office-based) studies, often on college students; the "research question" may test a theory about how people behave or think, with no immediate practical application.
As a public health researcher, I wanted to publish my findings, too. But I also wanted information that could help parents, doctors and teachers understand what's normal and what's unusual when it comes to violent video games and kids. This is known as "applied" research.
To start a video game research study like mine, you need to make a host of decisions, such as:
Who do I include in my study? I wanted groups of middle-schoolers, as they would represent the larger U.S. population of kids that age. It's easier to recruit college students for studies, but their brains and experiences are different from those of young teens, so I couldn't "generalize" my findings to 13-year-olds.
How do I measure their "exposure" to violent video games? I decided to ask teens to "list five games you've played a lot in the past six months," figure out their ratings, and use the M rating as a proxy for violent content. You can argue with my decisions; the best I can do is describe them clearly, so you know what my assumptions were.
Problems with my approach include: I'm ignoring the fact that T-rated games also have violent content. I'm assuming that games with violence against aliens or orcs will have the same effects as games with violence against humans. I'm not considering how "cartoony" or realistic the violence is. The list goes on. But I can't account for all of these factors in one study; it's too complicated.
"Exposure" would also include how much time they spend with violent video games: in a single play session, and across months and years. Can I expect a child to recall this accurately? Also, game technology is changing fast; how does that factor in? You can see how complicated this is getting!
How do I measure "aggression"? This gets to the heart of many disagreements about research findings. There is no widely agreed-upon definition of aggression. Do fleeting aggressive thoughts or feelings count? Can I generalize from some harmless "aggressive" behavior in a study, such as a brief blast of static-like noise, to assaulting someone in the real world?
Some researchers believe this to be true. I don't. I chose to measure common behavior problems, such as getting into at least one physical fight in the past year, or bullying other kids through words or deeds. I had to rely on what kids wrote on a survey, or "self reports" of their bad behavior. Their answers were kept private, but I had no way to double-check anything they shared. Again, this is typical.
To sum up, I found that the more M-rated games were on a child's list of five, the greater the odds that he (or she) had bullied and fought. Even though these are common middle-school behaviors, it seemed that parents ought to at least keep a closer eye on a child whose favorite games were all violent ones.
But then, we did more analyses, factoring in more of the stuff we asked about in the survey (with the help of Chris Ferguson at Texas A&M International University – I had run out of grant money). When we included questions measuring aggressive personality and stressful life events, the link (or "correlation") between playing M-rated games and bad behaviors went away.
So, does this mean that a child with aggressive personality is more drawn to violent games? That a child who is coping with a lot of stressful stuff will turn to violent games to blow off steam? Possibly. More research is needed before I can say that.
Note that because of the design of my study -- a survey at a single point in time -- I can't say whether exposure to video game violence "causes" anything. I can only calculate correlations: the odds of a relationship between two things -- like playing at least one M-rated game "a lot" and getting into a fight last year -- being greater than you'd expect by chance.
The Supreme Court said that none of the research submitted by the State of California "proved that violent video games cause minors to act aggressively... Nearly all of the research is based on correlation... and most of the studies suffer from significant, admitted flaws in methodology."
Now you have some idea of what that means. The "flaws" include generalizing from lab studies on a few college students (or surveys of a couple of hundred students in North Dakota) to all teens, from a paper-and-pencil measure of aggression to criminal assault. To make matters worse, there are myriad ways to subtly manipulate data, carelessly or on purpose, to get the results you want. For example, if your teacher volunteers your 7th grade class to fill out a survey about video game violence, and collects the survey papers herself, you kind of know what answers she's looking for.
If you'd like to know more about the messy business of video game research, click here and download the chapter at the bottom of the page.
What research means for game developers
I'd argue that there are moral and practical reasons for video game developers to support new research studies.
So far, there is no evidence that any violent video game caused or triggered any real-life murder. (There's also no evidence that training on a video game, by itself, can teach you to shoot a gun. Every school shooter studied by the FBI and Secret Service had practiced with real guns.) It's probably impossible to prove whether violent video games cause rare events such as murder, and especially mass murder. Given that playing violent video games is a statistically normal behavior for 13-year-old boys (and many girls), and that youth violence has been declining since the mid-1990s, it's hard to argue that the typical teen is harmed by them in any significant or lasting way.
However, these figures tell us nothing about atypical teens -- especially those who we know are at higher risk of violence because they see real violence in their homes or neighborhoods, and lack access to healthy outside activities and caring adults. Adding media violence to their already-violent worlds might harm these kids; or, they might use violent games as a safety valve to blow off stress and angry feelings. We owe it to them to find out. Once we do, we can work to maximize any benefits and minimize any harms.
If only out of self-interest, game developers should support research on the violent game use of juvenile offenders. If the principal fear is that violent video games promote or cue attacks against people or property, it makes sense to study the media use of juvenile offenders -- who have actually attacked people or property -- to see if their video game use differs from that of other teens. If their play patterns are the same as typical kids (or if they are exposed to less video game violence than the norm), that's powerful information for the game industry to have when the next lawsuit looms.
Joanne Savage, a crime expert at American University, wrote that "when I worked with young offenders, they spent much less time than other adolescents I have known watching television, going to movies or playing video games."
Supporting research for parent education would also be wise. At Vice President Biden's meeting last week with game industry representatives and researchers, he made two things clear. First, he'd seen no evidence linking violent video games and real-life violence. Second, many members of the public think there's a link, and game makers and sellers ignore this belief at their peril.
Don't assume that "Hey, the Supreme Court is on our side. We don't need to do anything. And we've already put out brochures." As if to highlight the ineffectiveness of industry outreach, the Vice President mentioned walking up and down the hallways of the Eisenhower Executive Office Building that morning, asking top officials and aides whether they'd heard of parental controls and game ratings. All professed ignorance. I found similar misconceptions and lack of awareness among my colleagues at Harvard. People were vaguely aware of years of headlines warning of video game dangers. Most never heard about the Supreme Court case.
The parents most concerned about video game violence are probably the ones least likely to be affected. Their children have involved parents who monitor and regulate their media use. But they are also the most vocal. And they deserve research-based information to relieve unfounded fears that hurt them and their children, and to guide sensible use of video games that will minimize any harms and maximize any benefits.
Like a good scientific study, a successful outreach campaign requires clear questions and upfront planning. Important aspects include:
What parents currently know and believe about video games. For example, what's in a typical M-rated game? What are ratings and content descriptors? How can I control my child's access, and how do I do that? What are potential benefits and potential risks of video games, including violent content? What do I consider to be violent or inappropriate content?
The experiences and inputs that gave rise to those beliefs. Had they seen video clips of violent game content, or read magazine articles? Did other parents share stories they'd heard about some awful game effect? Had they glimpsed, or actively watched, their children's play?
What parents currently do regarding video games. How do they monitor or limit their children's play? How do they choose what games to buy, rent or borrow?
People or organizations parents turn to or might trust to inform them about video games. This suggests who industry might seek out as partners or presenters (live or via video or Web).
Answering these questions might include everything from focus group discussions to following moms around GameStop or Wal-Mart and asking them to think out loud as they pick through racks of games.
The best message the video game industry can send may be: We want to give you options. We're parents, too. We want to give you more control over what your child sees and does with video games. This might include, for example, making reduced-violence versions of games sold in other countries available as options in U.S. games.
I'm not a developer, and I don't have a clear idea of the costs or permissions involved here. But I suspect this general approach is far cheaper and simpler than battling more efforts to slap labels on game boxes (members of Congress are working on these now) and other legal kerfuffles.
Cheryl K. Olson, Sc.D., was principal investigator for a two-year, $1.5 million Harvard research project to study the effects of video games on young teens, funded by the U.S. Office of Juvenile Justice and Delinquency Prevention. She coauthored a popular book based on that research, called Grand Theft Childhood: The Surprising Truth About Violent Video Games and What Parents Can Do. Her videos for parents on wise use of videogames can be found at: http://www.youtube.com/ckolsonscd