Shaping, Chaining, Instructional Formats, and Group Contingencies
Key Takeaways
- Shaping builds a NEW terminal response by differentially reinforcing successive approximations and withholding reinforcement for earlier forms as criteria advance.
- Chaining teaches an ordered sequence using a task analysis; forward starts at step 1, backward starts at the last step (contacting the terminal reinforcer first), total-task trains every step each session.
- Discrete-trial teaching arranges massed, adult-led trials (SD-response-consequence); naturalistic/incidental teaching captures motivation in natural routines and tends to generalize better.
- Group contingencies are independent (each person earns on own behavior), dependent (the group's access rides on one or a few members - 'hero' risk), or interdependent (the whole group must meet a criterion).
- First classify the behavioral unit - missing response (shaping), sequence (chaining), arranged trials (DTT), ongoing free-operant responding (naturalistic), or group-based access (group contingency).
Shaping: Building a New Response
Shaping is the differential reinforcement of successive approximations toward a terminal behavior that is not yet in the repertoire. The analyst reinforces responses that are progressively closer to the goal and stops reinforcing earlier, cruder forms once a closer approximation is established. Teaching a nonvocal learner to say "water" might reinforce a lip movement, then a "w" sound, then "wa," then "water."
Shaping requires a clearly defined terminal response, an identifiable starting response already occurring, reinforcers deliverable immediately, and data on progress. Criteria must advance gradually enough to keep responding strong but quickly enough to avoid cementing weak approximations.
Moving criteria too fast can put the new approximation on extinction (it never gets reinforced and responding breaks down); moving too slowly over-strengthens a crude form that then resists change. Shaping can target any measurable dimension — topography, frequency, latency, duration, or magnitude — not just the look of a response.
Shaping Is Not Prompting
This distinction is tested. Prompting adds an antecedent to occasion a response, then fades it. Shaping changes the response form itself through differential reinforcement of approximations — no added antecedent is required. If the scenario describes reinforcing closer-and-closer versions of a behavior, it is shaping; if it describes guiding/modeling and then removing that help, it is prompting/fading. The two are often combined in practice, but the exam wants the defining mechanism.
Chaining: Teaching Ordered Sequences
Chaining teaches a behavior chain — a sequence of responses where each step produces the SD for the next. You first write a task analysis (TA) that breaks the routine into teachable steps (e.g., handwashing, logging into a system, making toast). The chaining method depends on learner history, safety, where the natural reinforcer sits, and staff capacity.
| Method | Begins teaching with | Useful when |
|---|---|---|
| Forward chaining | Step 1, adding steps in order | Early steps cue later steps; learner tolerates delayed completion |
| Backward chaining | The LAST step (rest prompted) | Contacting the terminal reinforcer early is valuable/motivating |
| Total-task | Every step each session | Learner can attempt most steps with prompts; shorter chains |
In backward chaining, the trainer completes all steps except the last, the learner does the last step, and contacts the terminal reinforcer immediately — then the second-to-last step is added, and so on. This front-loads reinforcement, which can help unmotivated learners.
Instructional Formats: Discrete-Trial Versus Naturalistic
Discrete-trial teaching (DTT) arranges massed, adult-directed trials, each with a clear SD - response - consequence and a brief inter-trial interval. It is efficient for repeated practice of discriminations, verbal operants, and specific responses, and it produces clean data. Its weakness is that contrived trials can generalize poorly without explicit programming.
Naturalistic / incidental teaching (including milieu and natural environment teaching) captures the learner's current motivation (MO) in ongoing routines: the learner initiates, the teacher requires a target response to access the desired item, and the natural consequence reinforces. It tends to generalize and maintain better because teaching occurs under natural cues, but it yields fewer trials per minute and messier data.
- Free-operant arrangements allow many responses over time (fluency building, natural routines).
- Trial-based arrangements present discrete, contrived opportunities (DTT).
DTT and naturalistic teaching are not rivals so much as tools for different jobs: DTT excels at the rapid acquisition of discrete discriminations, while naturalistic teaching excels at generalization, maintenance, and spontaneous initiation. Many programs use DTT to establish a skill and then shift it into natural routines to promote durability. On the exam, choose by the stated goal — acquisition speed and clean data point to DTT; generalized, self-initiated use in everyday contexts points to naturalistic teaching.
Group Contingencies
A group contingency applies a single contingency to multiple people. There are three types, and the exam tests the distinction precisely:
| Type | Reinforcement depends on | Exam cue / risk |
|---|---|---|
| Independent | Each individual's own behavior against a shared criterion | Same rule for all; individual access; no peer dependency |
| Dependent | One or a few selected members' performance | 'Hero' procedure; peer pressure / scapegoating risk |
| Interdependent | The whole group meeting a criterion together | Builds cooperation; unfair if some lack the skill/supports |
In a dependent contingency, the group earns the reinforcer based on a target individual's performance — efficient but it can create coercive peer pressure or make one student a scapegoat. In an interdependent contingency (e.g., the Good Behavior Game), the whole group must meet the criterion. It promotes cooperation but is unfair if some members cannot yet perform without supports.
Exam Classification Routine
First, classify the behavioral unit, then map to the procedure:
- A missing response form that needs to be built from approximations → shaping.
- An ordered sequence of linked steps → chaining (pick forward/backward/total-task by motivation, safety, and skill level).
- Arranged, contrived opportunities with discrete SD-response-consequence → DTT.
- Repeated responding in natural routines capitalizing on motivation → naturalistic / free-operant.
- Reinforcement that rides on individual or group performance → group contingency (independent, dependent, or interdependent).
A clinician completes all steps of toothbrushing except the final step (rinsing), then has the learner perform rinsing and immediately delivers reinforcement. Over sessions, the clinician adds earlier steps one at a time from the end forward. This is:
A learner cannot yet say 'ball.' The therapist reinforces a 'buh' sound, then only 'ba,' then only 'bal,' then only 'ball,' withholding reinforcement for the earlier forms as each new approximation is mastered. This procedure is:
A teacher announces, 'If the entire class earns 90% on the spelling test, everyone gets 10 minutes of free time.' This is which type of group contingency?
Which instructional format would a behavior analyst typically expect to produce the BEST generalization and maintenance of a newly taught manding (requesting) repertoire?