This is interesting. What sort of psychological research are you thinking could only be conducted with old games, as opposed to newer games, which one can purchase without need to compile and build?
The release of the source code is actually more of a symbolic signal that the competitive value of the source code has reached a certain level. And that is within n orders of magnitude of the level where there is enough energy and money to pay lawyers to pursue, say, trademark violation or copyright infringement against anyone conducting such research.
So to be perfectly safe, one would be ultra-paranoid and stick to games released in the 20th century and then frame it as a kind of historical landmark in the video game market: what do video games in the 20th century say about what we knew about operant conditioning up to that point?
Example: suppose a behaviorist foundation and ask questions about really minute details related to visual and motor processing. Another: observe the expected and actual effects on behavior.
Would you conduct a Stanford Prison Experiment style of experiment if it were done with video games? That is, the players are associated with guards and prisoners, and they play out similar roles that the original experiment outlined, but in a virtual world?
The prison experiment example already has a volunteer playerbase, it's called "Jailbreak mod" and versions exist for various iterations of Counter-strike and probably other FPS games, and almost certainly Roblox. It is role-playing, and attracts certain kinds of personalities. I have played it.
You may be able to do some of it with mods - but you want to compare how people react with different settings, sprites, sounds, or even gameplay rules.
