If you dig around you'll find out this study was lead by a PhD candidate, so it was most likely done to either generate supporting data for a thesis or fulfill some experimental research criteria. If it was for thesis work then many other data sets would be under consideration as well.Johnson McGee said:OT: This study has been done many times before in different ways, the defining characteristic of this one is a somewhat larger sample size which gives it a bit more reliability but it's still a weak conclusion. On the other hand, one could just look at census data and market reports for game purchases to show that there's no correlation between increased video game consumption and violence.
As for the weakness of the conclusion... enhh... it's not like they could come up with a stronger one with the data they collected. The study was never written up with the intention of publication and it seems that UQ only released the info they did to stir up the issue a bit and get their name in the papers. If the study was for a thesis or experimental research criteria the conclusions would only consider the data generated... because for a thesis the wider analysis and conclusions would be part of the thesis not the study and for experimental research course stuff, you're not encouraged to reach much beyond your generated data.