"Violent Pac-Man" researcher responds

Yesterday we had a little fun with a piece written by About.com's Aaron Stanton, in which he critiqued a study done by Dr. Kimberly Thompson, an Associate Professor of Risk Analysis and Decision Science at Harvard. Thompson, creator of the Kid Risk Project, testified before the U.S. House of Representatives June 14 to suggest changes to the ESRB rating system. Dr. Thompson took time to respond to Stanton's piece and succumb to a Q&A.

Her biggest issue came with the ESRB not playing games before issuing a rating, citing Elder Scrolls IV: Oblivion as an example of a title found be more violent than initially rated. Thompson also regales stories of dealing with the ESRB and their accusations that her studies are flawed. You can read her testimony here (PDF file).

Of her video game studies at the Kid Risk Project, Thompson has found suggestive themes (violence, sex, profanity) in games of all ratings. While you may find her definition of violence to be very broad compared to conventional wisdom, Thompson is an advocate of self-regulation. In regards to the "ultra-violent Pac-Man" study, Thompson writes,

"I think that it is important to keep in mind is that games rated E are played by children as young as 2 and 3 years old, and the developmental psychology literature indicates that young children do not have the developmental capacity to distinguish reality from fantasy until approximately age 6 or 7 (of course this varies)."

If you notice any hint of ill temper, please understand that she feels her work and research has been misconstrued. We appreciate that Dr. Thompson took time out of her schedule to respond as cordial as possible.

From Dr. Thompson,

"I will be happy to answer your questions, but I would like to ask that you please take the time to actually read the studies that we have performed and my actual House testimony (link) or note that gamepolitics.com has an MP3 of my testimony and Pat Vance's testimony (link). I ask this because your questions reflect a focus on just one of our studies and mis-attribute my testimony to July 2006 Senate hearings, when in fact I testified in June 2006 testimony in hearings held by the House.

"I believe that you may have been misled by the article by Aaron Stanton that leads off with this incorrect assertion of me testifying in July in the Senate and continues with inaccuracies from there. Please note that my research team has conducted numerous studies, and that you can read about them all from Kid Risk Homepage (go to the hot topic of video games) ... I will only emphasize here that as we noted in our study, "Finally, for historical comparison, we assessed the content of 8 classic arcade games that have been rereleased as E-rated compilations or paired with E-rated remakes of the original games." We did not include the arcade games that you're asking about in our statistical analyses of E-rated games and we are well aware that these games were created well before the industry established the ESRB and during an era with vastly different visual technology that makes comparison with modern games problematic..

"In short, we are always surprised when gamers focus on vintage arcade games when discussing our research, since these games were not part of the sample of 55 E-rated games that we examined and hence were not part of the statistical results. We only played and coded some older arcade games because we were interested in examining the progression from abstract and repetitive portrayals of violence in early video games to more realistic portrayals in modern video games. However, quite contrary to what some people appear to allege about our research, we never have and never would hold up these games as evidence against the accuracy of the ESRB rating system.

"With respect to all of our studies, I will also emphasize that we performed separate studies of different categories of game ratings (E, T, and M), because they are played by young people of very different developmental levels and comparisons between them would be inappropriate. We have never and would never use the percentage of violent game play to make a ridiculous claim that a game like The Legend of Zelda is more "violent" than a game like Grand Theft Auto: Vice City, for example, although critics of our work like to throw out such statistics and attribute such claims to us. In contrast to what you may believe, we are aware of the developmental differences in children of different ages, and we have been very careful to consider this in our research.

"I think that it is important to keep in mind is that games rated E are played by children as young as 2 and 3 years old, and the developmental psychology literature indicates that young children do not have the developmental capacity to distinguish reality from fantasy until approximately age 6 or 7 (of course this varies). I'm sure that as a young child you probably were not frightened of ghosts trying to kill you, but the concept is one that does frighten many young children. In the context of coding the arcade games, we faced the challenge of applying our definition of violence (which we clearly and transparently include in our papers or on the Kids Risk web site).

"I assure you that coding the arcade games was a challenge, because we had to decide whether the ghost chasing Pac Man (or Ms. Pac Man) had the intent to harm or kill, which we decided is the case because in fact you do die if caught, lives are limited, and hence the need for more quarters (or starting over if you're playing on a console or computer).... So, even if you're good at Pac Man, if the ghost is chasing you with the intent to kill you (and in those rare instances where it kills you) or when you kill the ghosts, we count that (i.e., chasing a character with the intent to kill is violent, eating yummy screen-clearing dots is harmless). As we have noted in our papers, people can reasonably disagree with us, but we did not believe that it was consistent to not count this as violence even though it is quite abstract. We are also surprised when critics of our research omit important information about our studies in which we have addressed this issue more. For example, we specifically examine not just the percentage of violent game play, but also other crucial factors like the severity of the portrayal of injuries and suffering, the numbers of human and nonhuman deaths, the types of weapons used, and the reward system.

"In particular, the severity of the portrayal of injuries and suffering is often what people think of when saying that one game is more "violent" than another (which is more relevant from a developmental psychology perspective for games rated T and M). For example, does the game contain minor auditory or visual representations of injury and pain that primarily serve to notify the player that a character has been injured (e.g., characters like Mario grunt or flash red when injured), or does the game contain graphic representations of injury and pain that serve to exaggerate or focus attention on suffering (e.g., characters screaming in agony or bleeding excessively when injured or when otherwise physically tortured)?"


And now for the interview:

Do you feel that the violence portrayed in games like Pac-Man and Mario Bros. is harmful to minors? In what way does it affect their growth to warrant a rating exceeding that of the pre-designated E?

I have not studied whether or not violence or any other content in video games or other media is specifically harmful to young people. Our research focuses on helping people understand the actual content in video games and making sure that everyone understands their responsibilities for protecting children from any media that may be harmful in the context of our self-regulatory environment. You can check out the new guide that I recently posted on Media and Kids that offers 10 essentials (concepts and actions) parents need to make smarter media choices for and with the kids in their lives. If you are interested in the body of literature that indicates harmful effects, then you should contact Dr. Joanne Cantor, Dr. Craig Anderson, or Dr. Vic Strasbuger. I believe that parents need to pay attention to their kids and to what their media experiences because all media are educational, whether intended or not. I will also note for you that the ESRB has assigned content descriptors for violence to games in the Pac Man series, which you can see for yourself by searching www.esrb.org. I am a huge advocate for self-regulation and for better parenting (I believe self-regulation means responsibility is required by all).

Does the lack of scientific evidence supporting claims that children are influenced by video games affect your study or opinion of it?

The absence of evidence is not the same as evidence of absence. There are plenty of studies that suggest that exposure to violent media increases aggressive behavior (again, talk to Dr. Anderson about this). No one has ever ever done a randomized clinical trial to assess the impact of any media product. At this point, the evidence of harms from video games includes a number of observations, and concern among pediatricians and adolescent medicine physicians who deal with patients who have experienced media-related health issues (ranging from addition to desensitization and aggressive behavior, etc.) The important thing to realize about self-regulation is that we are depending on parents to make good media choices with and for their kids to avoid bad outcomes.

How do you feel the ESRB should be overhauled?

This is in my actual House testimony so please read that. The top of my list is that the ESRB should actually play video games before it assigns ratings to games. Note that we have never said that the ESRB needs to play every part of each game (again read my testimony), but we think that the ESRB should actually experience a sample of the actual game play in addition to receiving information from the publishers about the most extreme content.

To what extent are video games used as a scapegoat for politicians and activists?

This is not a question that I have researched so I'm not sure how to answer it. My impression is that every industry thinks that it is the scapegoat for politicians and activists. This is America.

Is any form of violence acceptable in universal games?

Violence is part of life. I am comfortable deciding what is appropriate for me and my family, but I would not determine acceptability for anyone else. Our research seeks to help make parents aware of the violence and other content that may be of concern to them in video games and to make sure that they actually pay attention to their kids and their kids' experiences with games.

How did you come up with your method of calculating gaming violence? How do you respond with detractors who say that your range of violence (e.g. Pac-Man "chomping" a ghost) is too broad?

We developed and consistently applied definitions for violence and other content. I wrote a lot more about this particular point in my message above, so please refer back to that. Please note that in reading all of our studies, you should get a good sense of all of our research.