Introduction to Shaping
Shaping is a paradigm of experiments used by B.F. Skinner to test operant conditioning and behaviorist theory. The process of shaping involves reinforcement of successive approximations of a target behavior. The method requires that the experimental participant perform actions that are at first rewarded, then gradually changed to encourage the presentation of a specific, pre-selected action. Thus, shaping is a very useful tool for training animals, such as dogs, to perform a difficult task .
The method of shaping involves a calculated reinforcement of behavior by the trainer. Initially, rewards are given for even crude approximations of a desired action. Then, the trainer rewards a behavior that is one step closer, or one successive approximation nearer, to the target behavior. As the participant moves through each behavior trial, rewards for old behaviors are discontinued in order to encourage progress toward the desired behavior. In this way, shaping uses operant conditioning principles to train a subject by rewarding proper behavior, and discouraging improper behavior.
Shaping relies heavily on the reinforcement or punishment of successive approximations of certain behaviors. Reinforcement always strengthens the behavior that came before it, and can be either positive or negative. Positive reinforcement involves the subject attaining access to something desired (a reward), while negative reinforcement involves removing something negative, or aversive, from the situation. Both types of reinforcement encourage a behavior, and thus differ from punishment which discourages behaviors. Punishment can also be positive or negative. Positive punishment inflicts an undesirable consequence on the participant, whereas negative punishment removes access to the desired reward. Each type of reinforcement or punishment is meant to train the subject towards a particular target behavior.
Example of Shaping—Training a Rat to Press a Lever
B.F. Skinner performed shaping experiments on many animal test subjects, including rats. He would place the rats in an apparatus that monitored behaviors, called a Skinner box. The target behavior for the rat was to press a lever that would release food to the rat. When the rat initially entered the box, it had no pre-disposition to press the lever. Skinner would reward the rat for successive approximations toward pressing the lever such as taking a step in the right direction, standing on its hind legs, or touching the lever. Each reward led to an increase in approximations of the desired behavior, culminating in the rat pressing the lever and receiving food. This process has been replicated with other animals, even humans, such that it is now common practice in many training and teaching methods.