Make it a double blind test! Even the experiment's facilitator won't know which one is which. The facilitator should have a colleague configure the 4 Tavors and label the outside of each with a randomly generated number from 1-4, then give them all back to the facilitator in a big container. The facilitator runs the test for each volunteer subject. After the test, the subject ranks the triggers on a questionnaire and puts it in an envelope before handing back to the facilitator. Votes are counted after it is all over, then translated back to the real triggers by the colleague. This way, the facilitator can run a safe set of experiments and yet be clueless to the id of the triggers.
Edit: The implementation of that concept is notional, since the differences between single and two stage triggers might be obvious to the keen observer, and because test subjects might think aloud. But you get the idea