RBT Chaining and Shaping Questions

Shaping works only when reinforcement depends on the current approximation, so the RBT gives a bite only when Lily says the target sound, and passes the caregiver's concern about motivation along to the supervisor. Bites between trials are noncontingent reinforcement: Lily receives snack regardless of what she says, which reduces her motivation to vocalize and weakens the shaping contingency. A bite at the start of each trial is also delivered before any response, so it works like a bribe rather than reinforcement for the approximation. Reinforcing any sound, even one below the target, rewards earlier or unrelated vocalizations and stalls progress toward 'more.' Keeping delivery strictly contingent is what moves the behavior forward.

In a single-opportunity probe, assessment stops at the first error or hesitation beyond the set latency, and that step plus all remaining steps are scored incorrect. Because Ben did not start step 3 within 5 seconds, the RBT ends the probe there. Completing step 3 for Ben and continuing to score later steps describes the multiple-opportunity method, a different probe format in which each step is assessed. Waiting indefinitely ignores the latency criterion and would credit a step Ben cannot reliably start. Modeling step 3 and restarting the probe teaches during assessment, so the data would no longer show what Ben can do without help. Following the probe rules keeps baseline data accurate for the supervisor.

In shaping, the RBT reinforces any response that meets or exceeds the current approximation. 'Ba' meets the active step and 'ball' goes beyond it toward the terminal behavior, so both earn reinforcement, while 'b' falls below the step and is not reinforced. Reinforcing 'ba' only would put the full word on extinction, discouraging exactly the progress shaping is meant to produce. Reinforcing 'ball' only treats the final goal as the requirement, so Malik would miss reinforcement for the step he is currently learning. Reinforcing all three responses rewards 'b,' an earlier approximation, which can stall or reverse progress. The current step is a floor, not an exact match: at or above it earns reinforcement, below it does not.

Practicing the coat chain each time Hana actually needs her coat embeds instruction in a naturally occurring routine. The real situation supplies natural cues and a natural reinforcer, going outside, which promotes functional independence and generalization. Running ten back-to-back trials at the table removes the chain from its natural context, so Hana may perform at the table but not when she really needs her coat. Practicing only with a picture sequence teaches her to recite or point to the steps, not to perform them. Putting the coat on for her to save time removes every learning opportunity and builds dependence on adults. Natural routines give repeated, meaningful practice that matters in daily life.

A clicker that has been paired with edibles is a conditioned reinforcer, and it can be delivered the instant Theo meets the current approximation, marking exactly which shot earned reinforcement before the edible arrives. Immediate feedback is critical in shaping because small differences between approximations must be reinforced precisely. A larger edible at the end of a set is delayed and may reinforce whatever Theo happens to be doing at that moment instead of the target shot. Describing the shot after he finishes eating comes far too late to mark the approximation. An edible given before each shot is noncontingent, so it does not depend on meeting the step. The click bridges the gap between response and backup reinforcer.

In backward chaining, the RBT sets up the task so all earlier steps are already completed, and the learner performs only the final step, here placing the top slice of bread, then immediately contacts the finished sandwich and reinforcement. Setting out the materials so Zoe starts by opening the bread bag would begin instruction at the first step, which fits forward chaining or total task presentation. Prompting her through every step in order while fading help describes total task chaining, where all steps are taught on each trial. Teaching the first step and completing the rest for her is forward chaining, the reverse of the plan. Correct setup lets the final step end in the natural reinforcer.

Once a closer approximation occurs consistently and the advance criterion is met, the previous approximation is placed on extinction. The RBT withholds reinforcement for 'wa' and reinforces only 'wah-ter' or closer, so the new step becomes the standard. Reinforcing 'wa' because it used to be the target keeps the old approximation strong and can stall progress. Giving a smaller sip for 'wa' is still reinforcement; it continues to strengthen the earlier form rather than extinguishing it. Returning to reinforcing any vocal sound moves shaping back several steps without the supervisor's direction and would undo progress Kai has already shown. Reinforcing the new step while extinguishing the old one is what moves shaping forward.

A behavior chain begins with an initial discriminative stimulus, such as 'Time to wash hands,' which signals that reinforcement is available for starting the sequence and cues Eli to perform the first step. Without it, there is no clear signal for when the chain should begin. Physically guiding his hand right away delivers a prompt before Eli has had any cue or chance to respond, which skips the instruction and can build prompt dependence. Waiting silently leaves Eli without the signal that starts the chain, so a failure to begin would reflect missing instruction rather than a missing skill. Modeling the whole routine is a prompting strategy that may be used if planned, but it still follows the initial instruction.

Shaping reinforces successive approximations over many sessions, so it is too slow for a behavior that must be acquired immediately for health or safety reasons. Stopping at the curb near traffic is urgent; a learner cannot practice gradually closer approximations while at risk of injury, so the team relies on faster methods, such as prompting with close supervision. Speaking at a louder volume is a classic shaping target, because loudness can be reinforced in small, gradual increases. Sitting at the table for longer stretches fits shaping by duration. Saying 'cookie' more clearly is a typical example of shaping vocal approximations. Only the safety skill is a poor fit, because time matters more than gradual progress.

In a behavior chain, completing one step produces the stimulus that signals the next step, and that stimulus also functions as a conditioned reinforcer for the step just completed. So finishing each step reinforces the step before it, while the sticker at the end maintains the whole sequence. Saying each step reinforces the step after it reverses the relationship; a consequence can only strengthen behavior that came before it. Claiming that only the sticker reinforces the steps ignores the links that hold the chain together, and step order matters because each step sets up the next. The instruction at the start cues the first step, but it does not maintain the later steps on its own.