From Human factors chapter 2 page 29:

Simulated Spaceflight Environments

Another strategy for gaining needed information is to simulate the conditions of interest and then see how people react to them. This could involve constructing a mock-up of a spaceship or habitat and then
having people do the kinds of things that we want them to do in space. As Harvey Wichman points out, a simulation is something like a play. The mock-up comprises the stage setting and props. The crew is the
cast, and their activities are prompted by some sort of script. This may result in a low-fidelity simulation, akin to when a cast first sits around in a circle and reads the script out loud. Or it may result in a high-
fidelity simulation, akin to a full dress rehearsal. (Editor: Both low and high fidelity sims have been our goal at Mars Society. We will increase fildelity each year and will be simulating much more closely a real Mars Habitat as time goes by.)

During the 1950S and 1960s, space-cabin simulation studies mimicked, in certain ways, environments and situations expected during the early space program. Astronauts have always trained in mock-ups, including shuttle simulators that react to the pilot's use of controls in the same way that airplane cockpit simulators respond realistically during make-believe airplane flights. There are several space station
mock-ups, and we can expect the best of these to be among the training grounds for tomorrow's astronauts. Simulations give us useful information, but as Jack Stuster points out, they also have severe limitations. Unless we are willing to invest phenomenal amounts of money and energy, we can simulate only key parts of a mission (such as launches and landings). A realistic simulation of a multiperson mission lasting weeks or months and involving all operational phases (launch, transit, surface exploration, and so
forth) would be enormously expensive. Another serious drawback is that in certain key areas simulators lack authenticity. We need to know how people behave in physically and emotionally challenging environ-
ments fraught with deprivation and danger, but ethics prevent us from putting people at high risk simply for the purpose of collecting behavioral data. (Editor: Some of the first shuttle simulations were crude plywood painted to resemble the interior and low resolution simulations on monitors where the windows into the cargo bay would be.)