What Is Shaping?
Shaping is the procedure of differentially reinforcing successive approximations of a target (terminal) behavior. Each approximation is a behavior that more closely resembles the final desired response. Behaviors that meet the current criterion are reinforced, while behaviors that fall below the criterion are placed on extinction.
B.F. Skinner originally demonstrated shaping by reinforcing pigeons for successive approximations toward complex responses such as turning in circles and playing ping-pong. In applied settings, shaping is used whenever the terminal behavior does not yet exist in the individual's repertoire, meaning it cannot be reinforced directly because it never occurs.
When to Use Shaping
Shaping is appropriate when:
- The target behavior is not currently in the individual's repertoire.
- The behavior cannot be effectively prompted or modeled.
- The behavior involves a continuous dimension (duration, intensity, latency, accuracy) that can be gradually modified.
- Other procedures such as chaining or prompt fading are not suitable for the specific behavioral goal.
Common applications include increasing verbal output from non-vocal individuals, building social interaction skills, increasing exercise duration, improving handwriting legibility, and developing functional communication responses.
Step-by-Step Shaping Procedure
Step 1: Define the Terminal Behavior
Clearly operationally define the final target behavior. Specify what the behavior looks like, how long it should last, how intense it should be, or whatever dimension is relevant. For example, if the terminal behavior is "the client will request preferred items using a full sentence," specify the exact response form: "I want [item], please."
Step 2: Identify the Starting Behavior
Determine the closest existing behavior in the individual's current repertoire that resembles the terminal behavior. This is your starting point. For the sentence request example, the starting behavior might be the client reaching toward items or making any vocalization in the presence of a preferred item.
Step 3: Plan the Successive Approximations
Outline the steps between the starting behavior and the terminal behavior. Each step should be a small, achievable increment. For example:
- Any vocalization in the presence of a preferred item
- A vocalization that approximates a word (e.g., "wah" for "want")
- A recognizable single word (e.g., "want" or the item name)
- A two-word combination ("want cookie")
- A three-word phrase ("I want cookie")
- The full sentence ("I want cookie, please")
Step 4: Reinforce the First Approximation
Begin by reinforcing every occurrence of the starting behavior. Use a strong reinforcer delivered immediately following the response. Continue reinforcing at this level until the behavior is occurring at a stable, elevated rate.
Step 5: Raise the Criterion
Once the current approximation is established at a stable rate, shift the reinforcement criterion to the next approximation. Responses that met the previous criterion are now placed on extinction, and only responses that meet the new, higher criterion are reinforced. This is the differential reinforcement component of shaping.
Step 6: Continue Through Successive Steps
Repeat the process of reinforcing the current approximation until it stabilizes, then shifting the criterion upward. Continue until the individual is performing the terminal behavior.
Critical Considerations
Step Size
If steps are too large, the individual may not make the transition and behavior may decrease. If steps are too small, the process becomes unnecessarily prolonged and the individual may become "stuck" at intermediate levels. The practitioner must balance efficiency with achievability, adjusting step sizes based on ongoing data.
Timing of Criterion Shifts
Shift the criterion only when the current approximation is occurring reliably. Premature shifts lead to failed transitions and potential frustration. A common guideline is to wait until the approximation occurs correctly on approximately 80% of opportunities before moving to the next step.
Avoiding Ratio Strain
When raising the criterion, the reinforcement rate temporarily drops because the individual must now exhibit a more demanding response to earn reinforcement. If the jump is too large, the drop in reinforcement can produce ratio strain, leading to decreased responding, emotional behavior, or behavioral breakdown.
Clinical Example: Increasing Exercise Duration
A client's treatment goal includes walking on a treadmill for 20 minutes. Currently, the client walks for approximately 2 minutes before stopping. A shaping plan might include: reinforce 2 minutes of walking until stable, then shift to 4 minutes, then 7 minutes, then 10 minutes, then 14 minutes, then 18 minutes, then 20 minutes. At each step, reinforcement is delivered for meeting the current duration criterion, and the client is not required to walk longer until that step is established.
Common Mistakes
- Reinforcing too many different response forms simultaneously: Reinforcement should be specific to the current approximation. Reinforcing broadly makes it unclear which dimension is being shaped.
- Moving backward without recognizing regression: If the individual's behavior drops below the current criterion, the practitioner may need to temporarily return to a previous step rather than continuing to withhold reinforcement.
- Not using data to guide criterion shifts: Shaping decisions should be based on objective data, not clinical impression alone. Graph the behavior and use visual analysis to determine when to advance.
BCBA Exam Relevance
Shaping is a foundational concept on the BCBA exam. You should be prepared to define shaping, distinguish it from other teaching procedures like chaining and fading, identify appropriate applications, design successive approximation sequences from clinical vignettes, and troubleshoot common shaping problems such as ratio strain and premature criterion shifts.