Shaping

1. Shaping and Uses

Shaping is a procedure that allows the behavior technician to create new forms of behavior, rather than merely reinforcing existing behaviors. Shaping consists of systematically reinforcing successive approximations of a target response, while placing previous approximations on extinction.

Put simply, shaping is the use of reinforcement to make the form of a response gradually come closer and closer to the final targeted form.

Shaping involves reinforcing approximations of a response as it gradually becomes closer to the desired skill. Previous approximations are no longer reinforced as new approximations are gained.

Why is this? Because we know they can do better!

For instance, when teaching a patient with autism spectrum disorder (ASD) some of their first words, it is very unlikely that you will be able to prompt them to say the entire word correctly by modeling it. Instead, they will often be able to imitate some portion of the word.

IN CONTEXT

Here is an example of shaping.

When teaching a patient to ask for a cracker (assuming it’s one of their favorite foods), you might use shaping to gradually teach the patient to pronounce the entire word correctly, even though they can only say “Kuh” at first.

To do this, the behavior technician first ensures that the patient wants the cracker by doing a brief preference assessment and then holds up the cracker and says, “Cracker.” If the patient responds by saying “Kuh” or anything closer to the actual word “Cracker,” then the behavior technician immediately gives them a cracker.

On subsequent trials, the behavior technician would look for something closer to the actual word than “Kuh,” such as “Ahkuh.” When the patient says something like this, then the behavior technician immediately gives them a cracker and then requires this new response, or something closer, on subsequent trials in order to deliver the cracker.

This general process is continued until the patient can pronounce the entire word correctly.

Shaping has several benefits for us and for patients. It can do all of the following:

• Allows the patient to be immediately successful. Because of our acceptance of approximations of the behavior, the patient is not required to be 100% right on the first try.
• Allows the patient to access reinforcement. Because we deliver reinforcement for approximations; the patient does not have to wait to correctly perform the final behavior to access a reinforcer.
• Generally reduces patient frustration. By not requiring the final version of the target behavior in the beginning, we reduce the patient's frustration.
• Allows the patient to move at their own pace, although, it may be a lengthy process. The patient gradually gets closer to the target behavior at their own speed.
• Allows us to teach a skill that is otherwise difficult to prompt. Some skills can be difficult to prompt, especially if the target behavior does not currently happen and there is no way to prompt it.

EXAMPLE

If a patient is not saying the sound “b” in response to the S^D, “Say b,” and there is nothing we can do to prompt the patient to produce sound, we can use shaping. We cannot physically demonstrate to the patient how to produce the “b” sound as this is something that occurs vocally.

Note, a "non-example" would be if the patient does not clap his hands in response to the S^D “clap your hands.” This is a non-example because in this case, we can prompt the patient to clap their hands by demonstrating or by physically prompting the patient.

term to know

Shaping

Systematically reinforcing successive approximations of a target behavior, while placing previous approximations on extinction.

2. Shaping Guidelines and Examples

Your BCBA will let you know what procedures to use with your patient. If your BCBA has decided to use shaping to teach a new skill, these are the steps for conducting a shaping procedure.

step by step

You are working with a patient whose favorite item to play with is bubbles. However, currently they are not able to let people know when they want to play with them. You want to teach them to vocally request to play with bubbles so that they are able to ask several people (caregivers, siblings, etc.) for them whenever they want.

Target: “Bubbles.”

Step 1: Identify and reinforce the patient’s baseline: “Uh.”

Step 2: Select a new target approximation: “Buh.”

Step 3: Only reinforce the new approximation until consistent: “Buh” warrants a "Nice job, buddy!" while “Uh” gets the response, “Nice try, say ‘Buh.’”

Step 4: Select a new approximation and repeat the process: “Bub.”

Step 5: Continue to increase approximations until desired skill is achieved: “Bubbles.”

IN CONTEXT

Let's explore another example of shaping.

Victor can’t say the sound “b.” The behavior technicians have no way to prompt Victor to make him say this sound, so they need to shape the “b” sound.

Victor’s initial behavior (this is what the patient engages in prior to the intervention) is that when the behavior technician presents the S^D, “Say ‘b’,” Victor says “Ah” sometimes (sometimes he doesn’t respond).

Therefore, the target behavior (this is the goal for the intervention) in this scenario is that when the behavior technician presents the S^D, “Say ‘b’,” Victor says “b.”

Day 1: Victor sometimes says “Ah” in response to the S^D. The behavior technician reinforces this approximation. As the first step, we must reinforce the initial behavior (saying “Ah” sometimes), until Victor reliably says “Ah” when we present the S^D, “Say b.”

Day 3: Victor says “p” in response to the S^D. The behavior technician now reinforces this approximation and does not reinforce “Ah” as an approximation anymore. The behavior technician reinforces “p” because it is a closer approximation to “b” than “Ah” was. We no longer reinforce “Ah” because if we do, Victor will continue to say “Ah” and now we only want him to say “p.”

Day 7: Victor says “b” in response to the S^D. The behavior technician now reinforces the target behavior and does not reinforce “Ah” or “p” as approximations.

Now, Victor is starting to demonstrate the target behavior! This is our goal for Victor. As such, we do not reinforce “p” anymore because if we do, Victor will continue to say “p” in response to “Say ‘b’” and we only want to reinforce the best approximation, which is now “b.”

big idea

These are the overall guidelines for shaping behaviors:

1. Select the target behavior (e.g., “b”).
2. Reinforce the initial behavior until it is under S^D control (e.g., Victor says “Ah” consistently every time the behavior technician presents the S^D “Say b.”).
3. When the patient learns a better approximation (e.g., “p”), no longer reinforce the previous approximation.
4. Continue until the patient learns the target behavior.

think about it

Think of an example of a final behavior you might want to shape, the initial behavior, and some of the intermediate behaviors that you would reinforce along the way.