This question came to me when I witnessed a collegue's behaviour on an application freezing due to a large operation. And it's something I've certainly done myself.
An app freezes for more than a couple seconds and we start clicking at it. Repeatedly. Often in no particularly meaningful place, as if that is going to somehow "wake it up." Even more bizarrely, users who have learnt that all those clicks will queue up and fire off when the program responds again will specifically choose to click in an area that has no known interactive ability (like an empty part of the chrome).
So users are clicking when they know it will have no effect and in areas that they know has no function, in hope that it will bring their program back to life. We all do it. But why?
Also, it seems the longer a program stops responding, the more random actions we start to throw at it. After a few seconds, it's clicking. Soon after, its keyboard tapping (the spacebar seems a favourite; but conversely people seem reluctant to try the Enter/Return key). Then the Escape key in heavy rapid fire (power users may break out the Ctrl-Alt-Del or other combos at this point).
Until finally, we reach the guaranteed fix-all of unresponsive programs : kicking the tower.
Why do we do this?
For the rest of the story: http://ux.stackexchange.com/questions/39413/why-do-users-click-randomly-and-rapidly-when-an-application-hangs