5.3 Prompting Hierarchies, Fading, and Transfer of Stimulus Control
Key Takeaways
- Prompts are antecedent stimuli added to an SD to occasion a correct target behavior; they must be systematically faded to establish independent stimulus control.
- Most-to-Least (MTL) prompting begins with the most intrusive prompt to ensure accuracy during skill acquisition, whereas Least-to-Most (LTM) prompting fosters independence during maintenance or fluency.
- Prompt fading methods include time delay (constant and progressive), prompt physical fading, and stimulus fading (modifying physical features of the SD).
- Transfer of stimulus control occurs when the learner's response shifts entirely from relying on an artificial prompt to responding independently to the naturally occurring SD.
- Prompt dependency is prevented through vigilant data collection, systematic fading schedules, and withholding high-magnitude reinforcers for prompted responses.
Prompting Hierarchies, Fading, and Transfer of Stimulus Control
In Applied Behavior Analysis, a prompt is defined as a supplementary antecedent stimulus introduced to occasion a correct response in the presence of a discriminative stimulus ($S^D$) that does not yet reliably evoke that behavior independently. While prompts are essential tools for accelerating learning, an un-faded prompt acts as an artificial crutch. The ultimate objective of all behavioral instruction is transfer of stimulus control—shifting the learner's independent response from dependence on the ABAT's artificial prompt to the naturally occurring $S^D$.
1. Classification of Response and Stimulus Prompts
Prompts are broadly categorized into two main structural divisions: Response Prompts (which act directly on the learner's behavior) and Stimulus Prompts (which act directly on the instructional stimuli).
PROMPT CATEGORIES
│
┌────────────────────────────────┴────────────────────────────────┐
▼ ▼
RESPONSE PROMPTS STIMULUS PROMPTS
(Operate on the learner's output) (Operate on antecedent stimuli)
├── Full / Partial Physical ├── Positional Prompts
├── Full / Partial Vocal (Verbal) ├── Redundancy / Feature Prompts
├── Modeling Prompts └── Movement / Highlight Prompts
└── Gestural Prompts
Response Prompt Types (Ranked Most to Least Intrusive)
- Full Physical Prompt: Hand-over-hand physical guidance to complete the entire motor movement (e.g., guiding a child's hand to trace the letter 'A').
- Partial Physical Prompt: Light physical guidance at the wrist, elbow, or shoulder to initiate or guide movement without controlling the full action.
- Full Vocal (Verbal) Prompt: Echoic model providing the exact vocal response required (e.g., ABAT says "Say ball" -> learner says "Ball").
- Partial Vocal (Verbal) Prompt: Providing the initial phoneme or syllable of the target vocal response (e.g., ABAT says "Baa..." for target "Ball").
- Modeling Prompt: Demonstrating the exact visual or motor behavior required for the learner to imitate (e.g., ABAT claps hands to prompt learner to clap).
- Gestural Prompt: Pointing, glancing, tapping, or waving toward the correct stimulus to signal the target response.
Stimulus Prompt Types
- Positional Prompt: Placing the target stimulus closer to the learner than distractor stimuli within a visual array.
- Redundancy Prompt: Exaggerating a physical dimension of the target stimulus (e.g., making the correct flashcard larger, brighter, or a distinct color relative to distractors).
2. Instructional Prompt Hierarchies
Choosing the correct prompt hierarchy depends on clinical assessment data, learner learning history, and the specific phase of skill acquisition.
| Hierarchy Strategy | Operational Definition | Clinical Application | Advantages / Disadvantages |
|---|---|---|---|
| Most-to-Least (MTL) | Instruction begins at the highest level of prompt intrusion (e.g., Full Physical) and systematically step-fades to less intrusive prompts (Partial Physical -> Gestural -> Independent) across trials or sessions as criterion is met. | Skill Acquisition phase; Errorless Learning protocols; learners prone to frustration. | Advantage: Minimizes errors, builds high success rates. <br>Disadvantage: Risk of prompt dependency if fading criteria are not strictly followed. |
| Least-to-Most (LTM) | Instruction begins by offering an opportunity for an independent response. If an error or non-response occurs, the ABAT systematically increases prompt intrusion (e.g., Independent -> Gestural -> Model -> Physical) within the trial. | Skill Maintenance phase; Fluency training; learners with strong prerequisite skills. | Advantage: Always affords independence opportunities. <br>Disadvantage: May allow frequent errors during early acquisition. |
3. Systematic Prompt Fading Procedures
To achieve transfer of stimulus control, prompts must be systematically faded according to operationalized protocols defined in the Behavior Support Plan.
Time Delay Fading Procedures
Time delay procedures introduce a controlled temporal interval between the presentation of the $S^D$ and the delivery of the prompt:
- Constant Time Delay (CTD): A fixed delay interval (e.g., 3 seconds) is established after an initial block of 0-second delay trials. The $S^D$ is delivered, the ABAT waits exactly 3 seconds, and if no independent response occurs, the controlling prompt is delivered.
- Progressive Time Delay (PTD): The time delay interval gradually increases systematically across trials or sessions (e.g., 0s -> 1s -> 2s -> 3s -> 5s) to allow the learner growing windows of opportunity to respond independently.
Stimulus Fading
Stimulus fading involves systematically altering the physical characteristics of a stimulus prompt until only the natural $S^D$ remains. For example, when teaching word reading, a picture of an apple is embedded within the letters of the word "APPLE." Across successive teaching sets, the picture's opacity is reduced (100% -> 75% -> 50% -> 25% -> 0%) until the learner responds purely to the printed text.
4. Transfer of Stimulus Control
Transfer of stimulus control is demonstrated when the prompt is completely eliminated and the learner's response occurs reliably under the exclusive control of the naturally occurring $S^D$.
Phase 1 (Prompt Dependent): [SD: 'What is this?'] + [Prompt: 'Say Dog'] ──► Learner: 'Dog'
│
▼ (Fading Schedule)
Phase 2 (Partial Transfer): [SD: 'What is this?'] + [Prompt: 'D...'] ──► Learner: 'Dog'
│
▼ (Complete Transfer)
Phase 3 (Independent): [SD: 'What is this?'] ──► Learner: 'Dog'
Preventing and Mitigating Prompt Dependency
Prompt dependency occurs when a learner waits for a prompt before responding, despite possessing the capability to respond independently. To prevent prompt dependency, ABATs must:
- Enforce Strict Differential Reinforcement: Deliver high-magnitude reinforcers (e.g., primary reinforcer + enthusiastic praise + tokens) for unprompted independent responses. Deliver low-magnitude reinforcers (e.g., neutral praise only) for prompted responses.
- Avoid Shadow Prompting: Unintentional non-verbal cues—such as hovering hands near the target card, leaning toward the correct answer, or eye-gazing at the stimulus—must be strictly monitored and eliminated.
- Adhere strictly to supervisory fading criteria: Never maintain a prompt level beyond the mastery criteria established by the QBA/QASP-S/BCBA supervisor.
Prompt Dependence and Demand Fading
Prompt dependence occurs when a learner responds correctly only when prompts are present and fails to respond to the natural $S^D$ alone. It is a primary risk of intrusive prompt hierarchies that are not faded on schedule.
ABAT actions to prevent prompt dependence:
- Follow written fading criteria exactly (time delay increases, prompt-level reductions).
- Differentially reinforce independent responses more richly than prompted responses.
- Avoid inadvertent shadow prompts (hovering, eye gaze, body orientation toward the correct stimulus).
- Report stalled independence data promptly so the supervisor can revise the prompt plan.
Demand fading is a related antecedent procedure used when escape-maintained behavior is high. The technician begins with a reduced response requirement (fewer trials, easier steps, shorter duration) and gradually increases demands to the terminal criterion specified by the supervisor. Demand fading is not the ABAT independently "going easy" on a learner—it is a programmed plan component. Combining demand fading with high-probability request sequences and differential reinforcement often improves compliance while preserving instructional integrity.
An ABAT is teaching a client to point to their shoes. The ABAT taps the shoe flashcard with their index finger to prompt the client. Which prompt category does this represent?
Which prompting hierarchy begins by providing an immediate unprompted test opportunity, and systematically increases prompt intrusion only after an error occurs within the trial?
An ABAT presents the instruction 'Say Apple' while holding up an apple flashcard. Across 5 sessions, the ABAT shifts from saying 'Apple' (full vocal) to saying 'App...' (partial vocal) to saying 'A...' (phoneme) to saying nothing. What clinical process is being demonstrated?