Howard has had a longtime interest in psychology and human behavior and enjoys digging into the reasons why we do the things we do.
Shaping is a technique for behavior modification that uses the principles of operant conditioning. It was researched by B. F. Skinner with his work on pigeons and other animals.
Rather than reinforcing a single desired response, as in straight operant conditioning, shaping reinforces a series of responses that are progressively closer to the end goal.
This process is also referred to as behavior shaping or the method of successive approximations.
Examples of Shaping in Animals
Let's take someone who wants to train a bird to fly in a clockwise circle. Waiting for the bird to do this spontaneously would take a long time. On top of that, the bird would have to do it enough times to learn it gets rewarded for flying like that.
This behavior could be shaped instead. At first, the bird would get some food for flying to the right. Once this behavior is established, the bird stops getting rewarded simply for going right. Now, it must fly to the right and also curve equivalent to a quarter circle. Next, it only gets the reward for making a half circle. This continues to three quarters of a circle, a full circle, and then multiple circles.
Eventually, the bird is only rewarded for flying the target number of circles in a row.
What if the goal is to train a dog to ring a bell? Just like in the example above, it might never do this on its own.
The first step in shaping this behavior could be to reward the dog with a treat if it faced the bell. Next, the dog is rewarded if it moves closer to the bell. After that, the reward is only given if the dog moves right in front of the bell. The next step is to withhold the reward until the dog touches the bell. Finally, the dog is only rewarded for applying enough force to ring the bell.
Shaping could also be used to train a dog to roll over. The dog would be progressively rewarded for sitting, lying down or crouching, lying on its side, turning to its back, and finally rolling right over.
When you see an animal do something elaborate, it has almost certainly been trained using shaping techniques.
The value of shaping in modifying animal behavior is clear, but is it useful for humans?
Examples of Shaping in Humans
The value of shaping is also seen in modifying human behavior.
When a young child is learning to form letters and numbers, they don't do it perfectly. At first, the teacher acknowledges roughly formed characters. After a while, the teacher only praises smoother looking characters. Eventually, the child consistently forms proper letters and numbers.
Shaping is also at work when we practice something alone. If you're learning to play a song on the piano, you don't get a good feeling if the sequence of notes doesn't sound like the song at all. Trying again, a few notes in a row are played correctly and at the right speed. You reward yourself with an acknowledgement that you're getting closer. This process continues as longer sections of the song are reproduced properly and rewarded with a sense of accomplishment. Eventually, you're only satisfied by playing the entire song correctly.
Imagine a young boy who continues to crawl around instead of walking. He doesn't respond to encouragement or commands to walk.
First, the parent has to observe the boy's behavior without interference for a while. This allows them to identify the times when the boy's behavior is closest to the target behavior of walking all the time. They see that he stands to reach things he wants, to wash his hands, or while playing a particular game.
Now, the parent rewards the boy with praise every time he stands. If he stands to take a toy off a shelf, the parent might say, "You can reach that all by yourself. Good job!" When he goes back to crawling, nothing is said about it.
When the boy stands to use the sink, the parent acknowledges the accomplishment.
Every time the boy stands up, the parent praises and engages him. When he returns to the floor the attention stops.
Gradually, the boy will stand and walk for longer periods of time, and the reward is given every time. Eventually, the boy won't feel anything is being gained by crawling around.
As you've probably noticed, shaping requires an accurate identification of the smaller or simpler steps that lead to the main goal. Failing to recognize them properly will result in missed opportunities to reinforce the behavior, or in reinforcing the wrong behavior.
One name for this planning stage is task analysis, which was developed by R. B. Miller. It starts with defining the target behavior or performance. Each step along the way is then identified, broken into sub-steps or sub-skills.
Be aware that the first sub-step could be far from the end goal. For complex tasks, or tasks that are simply outside someone's current capabilities, this is the only way to get there. Find the behavior someone performs that is the closest to the desired behavior, regardless of how small a resemblance it bears to the final goal.
With the first and subsequent sub-steps identified, we know what needs to be rewarded and what needs to be ignored. This process of rewarding desirable behaviors and ignoring undesirable ones is called differential reinforcement.
Reinforcement of each sub-step should continue until it occurs reliably. Jumping to the next sub-step too quickly would require too great a change in behavior for the subject.
When is Shaping Effective?
Shaping is a sound method for attaining skills that:
- are new,
- are complex,
- take persistence and practice, and
- require improved accuracy and speed.
When is Shaping a Poor Strategy?
Shaping is time-consuming and requires special attention. It should only be used when simpler methods are ineffective.
If someone responds to straight instruction and reminders, or can model a behavior demonstrated to them, shaping isn't necessary.
- Skinner, B. F. (1953). Science and human behavior. New York: Macmillan.
- Miller, R. B. (1962). Analysis and specification of behavior for training. In R. Glaser (Ed), Training research and education: Science edition. New York: Wiley.
- Krumboltz, J. D., & Krumboltz, H. B. (1972). Changing children's behavior. Englewood Cliffs, NJ: Prentice-Hall.
This content is accurate and true to the best of the author’s knowledge and is not meant to substitute for formal and individualized advice from a qualified professional.