That could be possible designing a test for asking information about the scenario. You could ask the students a few questions about the configuration, objects, ... That would inform you if everybody perceived what you planned in your simulation.

Maurice, if I understand correctly. Are you proposing to recreate some element of the experience without creating all the other superfluous elements that would otherwise require you to build a fully immersive environment? I believe the answer is yes, you can target specific elements (cues) and you can fine tune which ones (multimodal/physical, psychological, functional) by characterizing them in terms of their contribution to the successful execution of the operational task. 

In terms of measurement, you can use subjective reactions but you could also use physiological sensors to determine if the experience is being produced. .

Interesting question.  I like Roberto's response as a start.  My interest in this is related to presence theory and the idea of 'perception of a real experience'.OR perception that a technology mediated experience is unmediated..i.e., that the user thinks they have (directly) experienced a 'real' thing...where 'real' could be defined a few different ways...or 'meaningful' could be substituted for 'real'.  (Sidebar: How 'real' does the experience have to be to be 'meaningful?)

Comparing how users talk about their experiences (report back, testimonials) with observations of users and again with physiological data could be really interesting.

I helped out on a project to test the acceptance of a serious-game in a secure forensic mental health setting. The conclusion was that a serious game is acceptable in the domain and potentially a very useful resource in assisting in the rehabilitation of service users. Team members are currently in the process of documenting this research.

What I think may be pertinent here is that a lot of work was done with the service-users to elicit environments, characters, storyline and gameplay of meaning to them. Great attention was paid to approaching a better ecological validity. Creative workshops with the service-users and service-providers were used to gather these requirements, and iteratively test and hone their implementation. Standard user-experience techniques were drawn on to facilitate and moderate this process. 

A final evaluation was held with service-users not involved in the creation of a test prototype game. The (group debrief, spoken) feedback was that the game reflected their experience, and that this significantly aided engagement, immersion and learning.

I will keep an eye on this question, I am about to do an MSc dissertation and thinking about ‘ecological validity’ as the subject.

I recently had the opportunity to try a high-stress decision making simulation that involved shoot-don't-shoot situations. I couldn't not believe how real it felt, not because of the physical fidelity of the simulation (which was ok) but because of the psychological fidelity. Other's had similar experiences which had lasting effects similar to what you would experience after encountering a high-stress situation. There was no scientific study being conducted but one of my colleagues brought and wore an EMG sensor were you could see the effects in real time. Very interesting... 

One of the aspects of a user perception is a workload, e.g. the workload in the full immersive virtual environment. As a measure for such an aspect I would recommend NASA TLX test which is a wide-used tool not only in the airspace industry. We used this approach in the comparative research of the virtual learning environment and it was effective.