No self-respecting behavioral scientist would say that they have proven something conclusively with their research. It's just not how research works, and we don't start out with the assumptions that we can "prove" anything. The data support one hypothesis, but the way that this is stated is that the researchers have failed to reject the alternative hypothesis: video games produce [violent] behavior. The null hypothesis was rejected (video games have no effect on [violent] behavior.
In the article, the researchers say, "And the effects are that exposure to violent video games increases the likelihood of aggressive behavior in both short-term and long-term contexts. Such exposure also increases aggressive thinking and aggressive affect, and decreases prosocial behavior."
It's correlational data from a metanalytic study, but it is quite possible that the metanalytic study covers several true experiments. In any case, we should probably state that the evidence is mounting for causal link between violent video games and violent behavior.
I tried to use this language/RAD environment. Kept getting hit with show stopping bugs. There's a million of them. For starters, the plugin for web content apparently runs continuously as a process in the background. Don't know if they fixed that yet. Of if they will. Uninstallers and updaters didn't work (uninstaller corrupted my Windows registry for version 3.5). The memory management is awful, and leaks abound when using externals (for example, databases and browsers). The list goes on. And on. Right now this is a very pretty package, but it's not ready for the big leagues. The underlying code needs to be vetted. I'd recommend sticking with open source alternatives such as wxpython or Tcl/Tk if you want to avoid these issues. There is nothing more frustrating than having a show stopping bug and having to deal with a commercial vendor that won't give you the time day or treats you like a child. I think "opacity" is the word that describes them.
A big problem here is that there is a niche of programmers that wants/needs to create commercial applications, especially for entertainment and education needs, but the RAD tools have really gone to hell from the commercial vendors. I started out using Authorware in 1998, which was quite lovingly "discontinued," but was very stable. Then came Macromedia Director, which was god damned rock solid through MX 2004 (although essentially abandoned for Flash by Macromedia; they had like one engineer working on it at the time of the sale to Adobe). Adobe took Director and sent the code base to Bangalore, India, to a bunch of engineers who've made a complete mess out of things from what I can tell. I don't think its a coincidence that the CEO of Adobe is a graduate of the universities of India himself. Anyway, they're stuck at version 11.5, which is an unholy pile of junk (and it will cost you a cool grand, btw). Director, the program that built Macromedia, is essentially dead and along with it just about the only commercial programming alternative for the unwashed masses without computer science degrees.
The fact of the matter is that RunRev doesn't "dumb down" anything except the programmer. It's a complete struggle to use, but really only because it's not a stable programming environment. If it worked as advertised, we would be awash in RunRev apps. In my opinion, and it's just my opinion (as the owner of an Enterprise license from RunRev), Runtime Revolution is amateur night in the programming world. On the other hand, Tcl/Tk has a totally whacked syntax (upvar anyone?) but it is very fast and very very stable. Especially version 8.4 (you can get it from ActiveState for Windows). wxPython is probably the best choice, however, and you can create binary apps if commercial is your thing.
Is it just me, or is commercial development software in general becoming more buggy?