Another Study Attacks Violent Video Games, Claims To Be "Conclusive"

Killer Orca is one of many to tell us about a new study on the effects of violent video games on kids. The latest meta-study that analyzed research from 130 different reports claims to have "conclusively proven" that violent video games make more aggressive, less caring kids. "The team used meta-analytic procedures — the statistical methods used to analyze and combine results from previous, related literature -- to test the effects of violent video game play on the behaviors, thoughts, and feelings of the individuals, ranging from elementary school-aged children to college undergraduates. [...] Anderson says the new study may be his last meta-analysis on violent video games because of its definitive findings."

Comments

Uh... no.

[dgatwood]

Nothing other than a double-blind study with random selection of test subjects can truly be considered "conclusive", IMHO. All studies that I've seen thus far are hopelessly thwarted by selection bias.

Re:Funny

[vell0cet]

Actually... there were MORE violent criminals before video games were invented.

Youth crime and violence have been steadily decreasing since the introduction of the playstation in 1995. And apparently, they haven't been this low since the sixties.

"As violent videogames have become more popular in the United States and elsewhere, violent crime rates among youths and adults in the United States, Canada, United Kingdom, Japan, and most other industrialized nations have plummeted to lows not seen since the 1960s." - Texas A&M International University researchers Christopher Ferguson and John Kilburn

There are some graphs out there from the US Department of Justice that show exactly this trend.

A Meta-Analysis of bad studies creates a bad Meta

[vtechpilot]

My wife's PhD thesis was a Meta-Analysis, and I helped her create some new tools for doing the math behind the analysis so I feel like I have a pretty good grasp on the topic. The process (greatly simplified) is this. Dig through hundreds of articles published in peer reviewed journals on the topic you are examining, and find as many as you can that test the specific theory you are studying. All the articles included in the meta-analysis must test the same theory. Next you need to reverse engineer the numbers reported in the article. This can be a bit tricky since each article may have reported their result using different statistical tests. Occasionally some articles don't have all the relevant numbers and you have to contact the author. Once you have all that data together the math is relatively straight forward.

Presuming that all the other articles that you feed into the process are based on high quality research, then a Meta-Analysis can give you an insight to the overall strength of the results of the theory being tested. As you might imagine this process can easily be a Garbage In Garbage Out sort of situation. The researcher performing the meta-analysis must have the ability to identify bad studies that overlooked key moderating variables, or were simply done poorly and remove these bad studies from their analysis. If you want to attack this meta-analysis, attack the articles it was based off of. A meta-analysis by itself is not 'conclusive' just because of the method it represents. The analysis itself must be performed on many many well done studies in order to have any credibility of its own.