This sounds like an interesting topic. I do not know the research, so do not know if there are scenes typically used, but I doubt there are any standard stimuli for such experiments.
It would be easier to conduct experiments using 3D still photography than 3D video, but you could construct scenes with foreground objects in focus and background objects in varying degrees of out-of-focus with video or with still photography. Obviously, you could do it using various different camera lenses, but you could also use a green screen (or blue screen, or black screen) method on your foreground objects, and then put those images in front of a background that you blur to varying degrees in software. I think using green screen methods with 3D would be pretty neat.
My personal experience is that I get less eye fatigue with large DOF 3D - everything in focus. The blurred objects with short DOF strain me some when I try to get them into focus but can't. I wonder how you will measure eye fatigue. Will you ask the viewers to rate it, or will you have a performance measure such as accuracy in counting objects in the video over time, or what?