A rat is put into a box with two compartments (A and B) separated by a low barrier, or hurdle, Compartment A is painted white and has a floor made of metal rods through which mild electric shocks can be delivered to the animal’s feet. Compartment B has a plain wooden floor and is painted in black.
Suppose at the beginning of the experiment, the animal is placed in compartment A and the shock is turned on. In response to a shock of moderate intensity, the rat will run and move about in the shock compartment; in course of its more or less random movement, it eventually gets over the hurdle into compartment B, where there is no shock.
The rat is then removed from the “sale” side of the apparatus-compartment B, and after a time, placed back in the shock compartment, compartment A. Again, when the shock comes on, the rat will move around and eventually find its way into the “safe” compartment. The experiment continues, in this way, with the rat being placed back in A after running to B.
The first few times the shock is given on the first few trials, in other words, the rat is slow to make the appropriate response of jumping the hurdle into the non-shock, or “safe” compartment. But as more and more trials are given, the animal learns to leap over the hurdle very soon after the shock comes on. In other words, it learns to make the response that terminates the noxious shock stimulus. This is escape learning, and it is based on negative reinforcement.
We might have done a similar experiment with a rat in an operant chamber. In this case, a mild shock would be applied to the rat’s feet and a lever press would shut off the shock, thus, allowing the animal to escape from it. Or we might have used dogs or monkeys instead of rats: but remember, whatever the species or apparatus, escape learning is based on negative reinforcement.