Overview
The Three-Term Contingency (ABCs of Skill Teaching)
Antecedent (SD or instruction) → Behavior (client response) → Consequence (reinforcement or no reinforcement). Every teaching trial follows this structure. Also called the "three-term contingency" or "operant contingency."
Discriminative Stimulus (SD)
The specific antecedent stimulus that signals that reinforcement is available for a particular response. The RBT presents the SD clearly and consistently as written in the program. Example: holding up a picture card and saying "What is this?"
Reinforcement
A consequence that increases the future likelihood of a behavior. Positive reinforcement = adding something desired. Negative reinforcement = removing something aversive. RBTs use reinforcers identified by their BCBA through preference assessments.
RBT's Role in Skill Acquisition
RBTs implement skill programs exactly as designed by the BCBA. They deliver SDs clearly, prompt as specified, provide reinforcement per the schedule, collect trial-by-trial data, and report performance to their supervisor.
DTT vs NET Comparison
| Feature | Discrete Trial Training (DTT) | Natural Environment Teaching (NET) |
|---|---|---|
| Structure | Highly structured, therapist-led | Child-led, naturalistic |
| Setting | Tabletop, clinic, designated area | Play area, home, community |
| Motivation | Therapist-controlled reinforcers | Child's natural motivation (MO-based) |
| Massed trials | Yes — many trials in a row | No — embedded in natural activities |
| Generalization | Requires explicit programming | Occurs more naturally |
| SD delivery | Clear, consistent, controlled | Embedded in natural interactions |
| Best for | New skill acquisition, discrete skills | Generalization, communication, play |
DTT Components
Intertrial Interval (ITI)
Brief pause (1–5 seconds) between the end of one trial and the start of the next SD. Allows the client to reset and prevents response chaining across trials. The RBT pauses briefly, then presents the next SD.
Massed vs Distributed Trials
Massed = multiple trials of the same target in a row (faster acquisition). Distributed = trials interspersed across different targets in the session (better maintenance and generalization). The BCBA specifies which to use.
Correct Response Procedure
Client responds correctly → deliver reinforcement as specified (e.g., praise + token, preferred item for 5 sec) → record + → pause (ITI) → next trial.
Incorrect Response Procedure (Error Correction)
Client responds incorrectly or no response within time limit → deliver error correction procedure as specified by BCBA (e.g., 4-step: SD → prompt → response → reinforcement → transfer trial). Record −. Do NOT provide reinforcement for incorrect responses.
Natural Environment Teaching (NET)
Natural Environment Teaching (NET)
Teaching occurs within naturally occurring activities and interactions using items the client is already motivated to access (captures and contrives Motivating Operations). Skills are taught in context and are more likely to generalize.
Capturing vs Contriving MOs
Capturing = taking advantage of naturally occurring motivation (client reaches for a snack → SD "What do you want?"). Contriving = arranging the environment to create motivation (put desired toy out of reach).
Incidental Teaching
A form of NET where the teacher waits for the client to initiate (approach an item), then creates an opportunity to teach a target response before providing access to the item.
Prompting Hierarchy
| Prompt Level | Type | Description | Independence Level |
|---|---|---|---|
| Gestural (G) | Indirect | Point, gesture, nod toward correct answer | Moderate |
| Positional (P) | Indirect | Place correct item closer to client | Moderate |
| Visual (V) | Indirect | Written cue, picture, symbol | Moderate |
| Verbal (VB) | Direct | Partial or full verbal instruction | Low |
| Model (M) | Direct | Demonstrate the correct response | Low |
| Physical (PH) | Most intrusive | Hand-over-hand or partial physical guidance | Lowest |
Prompting Strategies & Fading
Most-to-Least (MTL) Prompting
Begin with the MOST intrusive prompt (often physical/full verbal), then systematically fade to less intrusive prompts as the client gains independence. Good for new skills — ensures success and minimizes errors. RBT starts with highest level, reduces as client demonstrates consistent correct responding.
Least-to-Most (LTM) Prompting
Begin with the LEAST intrusive prompt, add more support only if the client does not respond correctly within the time limit. Also called "prompt hierarchy." Allows client to attempt independently first. Good when client has some skill.
Prompt Delay (Time Delay)
Inserting a delay between the SD and the prompt. Constant time delay = same delay every trial (e.g., always 3 seconds). Progressive time delay = delay increases across trials. Encourages independent responding before prompt is delivered.
Graduated Guidance
A form of physical prompting where the therapist provides as much physical assistance as needed and immediately fades the level of pressure as the client begins to respond. "Shadow" the client's hands as they approach independence.
Prompt Fading Goal
The goal is always to fade ALL prompts so the client responds to only the natural SD. Prompt dependency occurs when clients only respond WITH a prompt — the RBT must implement fading as programmed by the BCBA.
Stimulus Control
When a specific stimulus reliably evokes a specific behavior. In DTT, the goal is for the SD alone to control the behavior (no prompts needed). Prompts are used to initially transfer stimulus control and are faded as control transfers to the natural SD.
Chaining Methods
| Method | Description | How RBT Implements | Best For |
|---|---|---|---|
| Forward Chaining | Teach first step first; prompt remaining steps | Reinforce completion of step 1, then 1+2, etc. | Skills where sequence matters and first steps easiest |
| Backward Chaining | Teach LAST step first; prompt preceding steps | Client completes final step independently, then works backward | Skills where final step provides natural reinforcement |
| Total Task Chaining | Client attempts ALL steps each trial; prompt as needed | Prompt any step where client needs help; fade prompts across trials | Clients with some prerequisite skills; faster acquisition |
Task Analysis
Breaking a complex behavior (e.g., hand washing) into small, sequential, measurable steps. Written by BCBA. RBT uses the task analysis to prompt and record + or − for each step each trial. Essential for chaining procedures.
Shaping
Shaping
Reinforcing successive approximations toward a target behavior. Start by reinforcing responses that resemble the target, gradually requiring closer approximations before delivering reinforcement. Example: teaching a child to say "ball" → reinforce "ba" → "bal" → "ball".
Differential Reinforcement in Shaping
Only reinforce the current approximation criterion — do not reinforce previous (already-mastered) approximations. Move the criterion forward as the client consistently meets the current level.
Reinforcement Schedules
| Schedule | Description | Pattern | Behavioral Effect |
|---|---|---|---|
| CRF (FR1) | Every response reinforced | Continuous | Rapid acquisition; fast extinction |
| FR (Fixed Ratio) | After every Nth response | Predictable | Post-reinforcement pause |
| VR (Variable Ratio) | After average of N responses | Unpredictable | Highest rate; most resistant to extinction |
| FI (Fixed Interval) | First response after fixed time | Scallop pattern (slow then burst) | Slowest before interval ends |
| VI (Variable Interval) | First response after average time | Steady moderate rate | Moderate, steady; resistant to extinction |
CRF vs Intermittent
CRF (continuous reinforcement) = reinforce EVERY correct response; used in initial skill acquisition. Intermittent = reinforce SOME responses; used for maintenance and to build persistence. RBT fades from CRF to intermittent as directed by BCBA.
VR: Most Resistant to Extinction
Variable ratio schedules produce the highest rates of behavior and are the most resistant to extinction because reinforcement is unpredictable (slot machine effect). Used to maintain behaviors that have been acquired.
Practice Quiz
Quiz Complete!
Memory Hooks
Flashcards (click to flip)
What are the three parts of a DTT trial?
SD (discriminative stimulus/instruction) → Response (client behavior) → Consequence (reinforcement or error correction). The three-term contingency: Antecedent → Behavior → Consequence.
Most-to-least vs least-to-most prompting — when to use each?
Most-to-Least: start with most intrusive prompt (physical/full verbal); fade to less intrusive. Use for NEW skills. Least-to-Most: start with no/minimal prompt; add support only if needed. Use when client has some skill.
Forward vs backward vs total task chaining
Forward: teach FIRST step first, prompt rest. Backward: teach LAST step first (natural reinforcer). Total task: client attempts ALL steps each trial, prompt as needed. BCBA chooses based on client and skill.
What is shaping?
Reinforcing successive approximations toward a target behavior. Start by reinforcing responses similar to the target, then gradually require closer approximations. Example: "ba" → "bal" → "ball".
CRF vs intermittent reinforcement — when is each used?
CRF (every response): used during INITIAL acquisition for fast learning. Intermittent (some responses): used for MAINTENANCE and building persistence. RBT fades from CRF to intermittent as directed by BCBA.
What is prompt dependency and how is it avoided?
Prompt dependency = client only responds when a prompt is present, not to the SD alone. Avoided by systematically fading prompts as planned by the BCBA so stimulus control transfers to the natural SD.
What is the intertrial interval (ITI)?
Brief pause (1–5 sec) between the end of one trial and the start of the next SD. Allows the client to reset. Prevents response chaining across trials. The RBT pauses, then delivers the next SD.
What is the difference between capturing and contriving motivating operations in NET?
Capturing: using naturally occurring motivation (client reaches for item → teaching opportunity). Contriving: engineer the environment to CREATE motivation (put toy out of reach). Both create teaching opportunities in natural contexts.
Study Advisor
Start with the three-term contingency (SD → Response → Consequence). Then learn DTT trial structure: SD → response → consequence → ITI. Know the difference between DTT (structured/tabletop) and NET (naturalistic/child-led).
Study prompting types (physical, model, verbal, gestural, positional, visual) and the two main hierarchies (most-to-least vs least-to-most). Understand prompt fading goal: transfer stimulus control to the SD alone.
Master chaining (forward/backward/total task), shaping (successive approximations), reinforcement schedules (CRF vs VR vs FR vs FI vs VI), and error correction procedures. Know when to use each and why.
High-yield: DTT = SD→Response→Consequence; MTL = most intrusive first (new skills); LTM = least intrusive first; Forward = first step first; Backward = last step first; VR = most resistant to extinction; Shaping = successive approximations.
DTT: SD→R→C; MTL=safety net first; LTM=try first; Forward chaining=step 1 first; Backward=last step first; Total task=all steps each trial; Shaping=closer approximations; VR=highest rate/most resistant; CRF=every response; Prompt fading=goal is no prompts.