One thing we checked early was the response of the sensor to simple harmonic motion. We hung a 30cm aluminum bar from a loosely coiled spring, and attached the sensor to the bar with velcro. When disturbed from equilibrium, the bar (and sensor) underwent simple harmonic motion.

We could measure the period of this motion with a stopwatch, and compare it with the period observed in the wireless data. This is the experiment that revealed that the internal clock on the XBee was not to be trusted! After we moved time measurement function to the computer program we got good correspondence between the two period measurements.