5.4 Shaping Successive Approximations

Building behavior by reinforcing closer forms

Shaping is a behavior acquisition procedure in which successive approximations to a target response are reinforced. It is especially useful when the final response is not occurring, occurs too weakly, occurs too briefly, or is too different from the goal to reinforce directly. Instead of waiting for the perfect response, the teaching plan identifies a current approximation that the client can do and reinforces responses that are closer to the target. Over time, earlier forms stop contacting reinforcement and the criterion moves toward the final skill.

The RBT must distinguish shaping from prompting. A prompt adds assistance to evoke a response. Shaping changes which naturally occurring or prompted response forms contact reinforcement. The two can occur together in some programs, but they are not the same. For example, a vocal imitation program may use a model prompt while shaping the sound from "buh" toward "ball." A motor program may prompt arm movement while shaping duration or range. The RBT follows the written plan for both the prompt and the reinforcement criterion.

Shaping requires clarity about the target behavior and the current step. If the final goal is requesting "water" using speech, a shaping sequence might begin with orienting toward the cup, then any vocalization while reaching, then a /w/ sound, then "wa," then "water." If the final goal is remaining seated for a group activity, shaping might begin with sitting for 10 seconds, then 20 seconds, then 45 seconds, then 2 minutes. If the final goal is writing a name, shaping might reinforce grasping the pencil, then tracing one line, then tracing letters with fewer prompts, then writing the first letter independently.

A frequent implementation error is reinforcing too broad a range of responses for too long. If the client has moved from reaching silently to vocalizing "o" for open, but the RBT keeps delivering the item for silent reaching, the older response may remain strong and the new response may not progress. Another error is moving the criterion too far ahead. If the current response is a quiet "ba" and the RBT withholds reinforcement until the client says "basketball" clearly, the client may stop responding or show frustration.

The supervisor sets the shaping steps; the RBT's accurate data helps the supervisor decide whether the step size is working.

Shaping also requires moment-to-moment judgment within written boundaries. The plan may define acceptable approximations, such as "reinforce any vocalization containing /m/ or /ma/ during snack mand opportunities." The RBT should not decide that a different sound is close enough unless the plan allows clinical judgment at that level and the RBT has been trained. When unsure, record exactly what was heard or observed. Objective notes such as "said 'mmm' while reaching toward crackers" are more useful than labels such as "good attempt."

Reinforcement must be immediate and contingent. If the client emits the current approximation and the RBT delays reinforcement while finishing a conversation, the relation between response and consequence weakens. If the RBT gives the reinforcer after non-criterion responses because the client is upset, the shaping path may become unclear. This does not mean ignoring dignity, distress, or assent. It means the RBT follows the written plan, uses approved supports, and contacts the supervisor when the procedure appears to increase distress or when the client is not contacting reinforcement.

Shaping can apply to many response dimensions. Topography is the form of the response, such as a clearer word or more accurate pencil stroke. Duration is how long the response lasts, such as standing in line. Force or magnitude may matter for responses like pressing a switch firmly enough. Latency may matter when teaching a quicker response after an instruction. The RBT should know which dimension is being shaped because the data must match the dimension. Recording only correct or incorrect may be too vague when the current goal is duration or vocal approximation.

The professional RBT stance is to keep the shaping path visible. Reinforce the current approximation, stop reinforcing earlier forms when the plan says to, avoid raising criteria independently, and document the exact response. When progress stalls, the RBT does not blame the client. The RBT brings objective data to the supervisor: what responses occurred, how often they met criterion, what prompts were used, what reinforcers were delivered, and what contextual variables may have affected responding.