A08 Defintion Essay — Jon Gonzoph

[Might edit it a little before tomorrow, and it turned out to be exactly 1000 words, not counting Works Cited]

In most research situations, an ideal experiment is not needed. Normally, many studies on a similar topic will conclude similar results, and a researcher can use those results as a point of comparison to judge other studies by. If writing a paper with a thesis that agrees with the results, it is simple for a researcher to support their thesis by citing these studies. If the opposite, a researcher still has a concentrated group of conclusions to try to find a fault with, and finding studies that don’t come to these results is still a fairly simple matter.

Analyzing the effects of violence in video games does not follow the pattern shown above. It is a controversial subject, and for every study finding a link between violence in games and some negative effect is another claiming that no link exists. Even studies that agree on a general effect differ in concluding on the severity and length of this effect. In a case such as this, an idealized experiment is sorely needed, as it would allow a researcher a sort of template to compare any studies to, consequentially allowing them to separate the strong studies from the weaker ones. This idealized study cannot be found by comparing studies with similar results, for all that would lead to bias by the researcher; instead, only the methodology and the ways the studies interpret the results should be considered. Because the most debated effect of violence in video games is if it leads to increased aggression, I shall be contrasting several studies and trying to find which ones employ more stringent methods.

One vital aspect to consider when evaluating studies is how they measured increased aggression. For example, compare the study in Singapore by Sorick et. al. to the study on German adolescents performed by Moller and Krahe. In the first study, four different measurement scales and questionnaires are used; these scales have been in use for years and are thus believed to be fairly accurate. (Skoric). In contrast, not only did the second study only use two methods of measuring aggression, but one of those methods was taken from another of the author’s studies and had not been tested for reliability by an outside source. Both of these experiments used the Buss-Perry Aggression Questionnaire, but the first study used all 29 questions and the second study only uses 7 questions mixed with another 7 questions from a scale used for measuring a completely different type of aggression (Moller).

The conclusions to be drawn from this comparison are fairly evident. A study should use as many tools for measuring aggression as is reasonably possible, for that lessens the chance for anomalous results. Additionally, these tools must be rigorously tested before being used. If one were to use unproven tools, the data gathered is circumspect due to the possibility of warping the tools to allow the tester to manipulate the results to match with his or her expectations. This applies equally to both studies seeking to prove or disprove any negative effects in violent video games.

Another vital variable to consider when judging the worth of studies is the timeframe involved. It would be logical to assume that studies with a longer timeframe, such as 3 weeks in the case of the aforementioned Singapore study or an additional 30 month check as in the case of the German one, would produce more reliable conclusions. However, there is a widely held belief in the theoretical model known as the General Aggression model, which states that “media violence exposure not only leads to an immediate increase in aggression in a particular situation but also contributes to the development of an aggressive personality of the game player over time” (Moller). Following the GAM model would mean that even studies done in only a few hours would be relevant because any increased aggression found there would carry over into a long term effect. The problem is that GAM is not certain to be effective, with one study — “Internet Fantasy Violence: A Test Of Aggression In An Online Game.”  – finding results that directly contradict it. Over a month, the testers found that there was no difference in result between the group who played the same violent game for at least 56 hours and the control group that did not, whereas the GAM method would only be validated if there was a increase in aggression observed over this time (Williams).

Conclusions to this variable are more difficult to come to. Though the GAM has some studies that find it does predict behavior, the lack of consensus on this issue leads me to believe that the effects of short term studies should not be extrapolated over longer timeframes. Interpretations of results that subscribe to the GAM theory should also be analyzed closely, because it could skew the author’s views. However, it is not necessary to disregard those results entirely; just the specific sections that predict an increase in aggression over time without their own appropriate data to validate this conclusion. This conclusion only applies to the studies seeking to prove a link between video game violence and increased aggression.

In conclusion, though I have not been able to come to any sort of template indicating an ideal experiment for studying violence in video games, I am able to make some generalizations about which studies are superior. In regards to time, due to the uncertainty of the reliability of GAM, studies over longer periods of time will be preferred to those that only measure the effects of one gaming session. Studies that gather results using well-known and proven tools will be preferred over studies that use personally created or modified tool. In addition to this, studies that gather data based on multiple scales are preferable to those that only use one scale. Though other significant methodological differences exist, assembling a reference base of studies that meet these criteria is the first step on truly understanding the issue of violence in video games.


Works Cited:


1. Skoric, Marko M. et al. “Grand Theft Auto IV Comes To Singapore: Effects Of Repeated Exposure To Violent Video Games On Aggression.” Cyberpsychology, Behavior & Social Networking 14.10 (2011): 597-602. Academic Search Premier. Web. 7 Mar. 2012.


2. Möller, Ingrid, and Barbara Krahé. “Exposure To Violent Video Games And Aggression In German Adolescents: A Longitudinal Analysis.” Aggressive Behavior 35.1 (2009): 75-89. Academic Search Premier. Web. 7 Mar. 2012.


3. Williams, Dmitri, and Marko Skoric. “Internet Fantasy Violence: A Test Of Aggression In An Online Game.” Communication Monographs 72.2 (2005): 217-233. Academic Search Premier. Web. 7 Mar. 2012.


This entry was posted in x Definition Essay. Bookmark the permalink.

3 Responses to A08 Defintion Essay — Jon Gonzoph

  1. davidbdale says:

    I’m so glad you adopted this plan, Jon. I think it’s much more valuable than simply latching on to the studies that support one view or the other and ignoring the rest. I’ll be back if time permits with additional comments, but for now I just want to say I’m pleased with the scope and approach. (1000 words, you say!)

  2. I’m not exactly sure if this is still qualified as a definition essay, since the thesis mutated a bit from “this is what a perfect, ideal experiment is” to “Here are several factors that are important in a ideal study,” but I didn’t actually get near the perfect study itself.

  3. davidbdale says:

    Hey Jon. It’s still a definition essay.
    You can make it clearer that this is the case by saying somewhere in your long introduction that, although you haven’t found and can’t maybe even imagine the perfect study, you’ll analyze the flaws in several experiments to show how they deviate from what a good experiment would accomplish and hope, by what Michelangelo described as “carving away everything that doesn’t look like an elephant” to sculpt a perfect elephant: definition by negative attributes instead of positive attributes.

    The paragraph about Sorick and Moller is sort of helpful, but it gives us no indication how aggression is measured. Maybe you think it’s unimportant for us to know this and want us to concentrate instead on differences between the studies, but we feel left out. (If we feel left out too often or too long, we’ll stop reading.)

    *the data gathered is suspect?

    I don’t see why a longer timeframe is necessarily more likely to deliver reliable results, and I don’t think other readers will either. Maybe you will conclude that it’s easier to craft a perfect study for conclusions about short-term aggression effects and that longer studies are less reliable because of too many variables, but until you guide us to those conclusions, we don’t “logically” assume anything about the timeframe.

    No, of course short term predictors can’t be extrapolated to make long-term predictions, any more than we can judge how wired my kid will be tomorrow if I give him too much sugar today.

    The conclusions you come to are useful only if you’re looking for a particular type of result. You need to tell us earlier that the thesis you’re testing is that “prolonged violent gaming over several months produces more violent behavior even days after the most recent game-play” or whatever it is you’re looking to prove or disprove.

    I would think defining what we mean by aggression might be the first step, if that’s what we’re trying to measure. I’m still not sure whether the studies you read track hitting incidents, or attempts to intimidate other players, or playground hazing, or just a tendency to lash out verbally at the psychologists, Jon.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s