Clicker training is essentially a form of “operant conditioning,” a learning method developed and studied by Harvard psychologist, B. F. Skinner. Around 1936, Skinner created the “Skinner Box,” a tool that demonstrates the operant conditioning theory. Here’s how it works: A rat is placed into a nearly soundproof box. For a few days, a mechanism automatically drops pellets into a tray, and the rat moves to the tray to get them. Eventually, just the sound of the mechanism starting to drop the pellets motivates the rat to go to the tray. Next, a lever is raised near the food tray. As the rat moves around, “operating” in its environment, it eventually touches the lever and presses it downward, releasing a pellet from the food supply. The rat now knows that when it presses the lever, food will appear in the tray. The rat will press its little heart out to get as many pellets as it can. Then, the experimenter will change something. For example, the lever will be adjusted so that the rat has to push down very firmly to get the pellet. The animal soon learns this new element to the game, and adjusts its behavior accordingly. The idea is simple: Behavior that is reinforced using reward (positive reinforcement) has an increased probability of occurring in the future.
But can it work on organisms more intellectual than rats? Just go to Las Vegas or Atlantic City and observe any bank of slot machines, and then answer that question. Sometimes the machines pay and sometimes they don’t, but the small rewards the machine offers are enough to keep the player shooting coins down the slot and pulling the lever. When there’s a payout, a noise occurs, often a siren or a bell and then the exciting ching-ching of coins fall. The player is positively reinforced with noise and then coins, generally at pretty regular intervals, and so the game continues. Eventually, in most cases, the house skims a percentage of coins off of the top of nearly every play, leaving the player with nothing in the end. It’s less the idea of the big payout that keeps the player hooked as it is the brain’s natural conditioned response to winning sometimes. And so it is with rats – and dogs.
When the rat pressing the lever only gets a pellet every ten presses, or say, every few minutes, it will speed up the rate in which it presses the lever when it recognizes that there’s a patterned schedule. But, if a pellet no longer arrives when the lever is pushed, the rat will press the lever anxiously, trying everything it can to get the pellet. This is called the “extinction burst.” When no pellet arrives, the rat will stop pressing – the behavior is extinguished, at least for the moment. It’s the same with the slots; when one machine stops paying out after a few pulls, the player will rapidly make a few plays, then get frustrated and move on to the next machine.