The shaping procedure is a useful strategy to build new skills in learners. The definition of shaping is the differential reinforcement of a series of successive approximations to a target behavior. It’s okay if that sounds like a jumbled word salad - let’s break it down.
Reinforcement is the process of making a behavior more likely to occur in the future.
Differential reinforcement is using different amounts of reinforcement for certain behaviors depending on how similar they are to the target behavior (the closer the behavior is to the target behavior, the larger the quantity of reinforcement).
A successive approximation is any behavior that is similar to the target behavior.
The ultimate goal for shaping is the target behavior you set out for the learner.
When using shaping, you will select a behavior that is somewhat similar to the target behavior. You will start reinforcing this behavior so it becomes strong in the learner’s repertoire. Then, you will select a topography of behavior that is closer to what the target behavior should look like. You would put the old topography of the behavior on extinction, and then only provide reinforcement for instances of the new behavior.
One of the key things when shaping behaviors with our learners is to differentially reinforce behaviors when they are closer to the target behavior. This means when a successive approximation is closer to the target behavior in a trial, we give more of that reinforcer. For example, if the learner is on a token economy schedule of reinforcement, instead of giving 1 token you can give them 2 tokens for responses that are closer to the target behavior. And we do not give a reinforcer for approximations that are less similar. By differentially reinforcing successive approximations (aka similar behaviors) that are closer to the target behavior in a shaping procedure the probability of our learner reaching the target behavior will increase.