Why not run an experiment for two months? During January, you agree to leave the thermostat alone. It gets set, then no one touches it. Then, during February it's a free-for-all. You can change the thermostat as much as you like. Anyone can. Then record the amounts at the end and see if there's a big difference.