Discrete Trial Teaching (DTT) — Comprehensive Notes
Discrete Trial Teaching (DTT) — Comprehensive Notes
Overview
- DTT is a highly structured teaching method used to teach specific discrete responses usingABA principles.
- Typically conducted at a table with one instructor and one student; eye contact, exchange of materials, and clear interaction are emphasized.
- Distractions are minimized (e.g., pens stored in a box) to create a controlled learning environment.
- The program is defined by the supervisor with clearly specified stimuli and expected responses.
- DTT is contrasted with natural environment teaching (NET) and pivotal response teaching (PRT), which are more child-led.
- DTT aims to build a foundation of a skill in a structured setting and then transfer it to daily life.
- DTT is used when a client struggles to learn in natural environments; it is not intended to be used lightheartedly but as a concrete tool for skill acquisition.
- The three-term contingency (antecedent, behavior, consequence) is foundational to DTT and operant conditioning.
The Three-Term Contingency and Discriminative Stimulus (SD)
- Three-term contingency basics:
- Antecedent (A) → Behavior (B) → Consequence (C)
- The antecedent makes the behavior more or less likely; the consequence reinforces or punishes the behavior over time.
- Discriminative Stimulus (SD): a stimulus that comes to control a specific response because reinforcement occurs in its presence and not in its absence.
- SD is part of the learning history and signals reinforcement availability for a particular response.
- Examples to illustrate SD in daily life: firetruck card -> red response (color), doorbell ringing -> answering door -> package (reinforcement).
- In DTT, the SD is the cue that tells the learner which response will be reinforced.
- Stimulus control and discrimination are central: we want responses to be triggered by the SD and not by irrelevant cues.
DTT Formula: SD, Response, and Reinforcement/Feedback
- DTT uses a simple sequence:
- Discriminative Stimulus (SD) → Response (B) → Reinforcement or Informational Feedback (C)
- Consequences depend on the appropriateness of the response.
- Responses must be observable, measurable, and clearly defined in the program (no ambiguous states).
- The difference between a behavior and a response: a behavior is any observable action; a response is the specific, expected action when the SD is presented.
- The order and clarity of SD, response, and consequence help establish stimulus control and consistent teaching.
Responses in DTT: Types, Characteristics, and Criteria
- Correct responses: meet the program’s defined criteria for the given SD.
- Incorrect responses: do not meet the defined criteria.
- Self-correction (correction): the learner initially emits an incorrect response, then corrects themselves to the correct response.
- Important: mark self-corrections clearly because prompt cues or nonverbal indicators could influence performance.
- Nonresponse (NR): the learner provides no response within a specified time window (commonly 3–5 seconds; supervisor may adjust).
- Responses must be defined in objective, measurable terms and recorded consistently by the team.
- The “stuffed animal test” is used as a heuristic to distinguish observable responses from internal states (like emotions) that are not considered responses in DTT.
- Approximate responses may be accepted in some programs (e.g., pronouncing a target similarly but not exactly), but team consensus and supervisor approval are required.
- Multiple responses may be accepted for some targets to support generalization; all acceptable responses must be listed in the program.
- Examples:
- SD: show a red card, ask for color; expected response: “red.”
- SD: “cat says”; acceptable response: “meow meow” or “meow” in some programs.
- SD: “give me the big block” with a big vs. small block; expected: learner hands the big block.
Consequences: Reinforcement and Informational Feedback
- Consequences are either reinforcement for a correct response or informational feedback for incorrect/self-correct/nonresponse.
- Reinforcement must be contingent on the correct response and consistently delivered to build stimulus control.
- Reinforcement examples: social praise, tokens, tangible items; the type and magnitude should be matched to effort and the learner’s preferences.
- Informational feedback is neutral and non-aversive; it should clearly indicate the need to try again without demeaning the learner.
- Contingency and consistency are critical: reinforcement should be delivered immediately after a correct response and withheld after incorrect responses.
- Pairing praise with other reinforcers is common to ensure learning translates to natural environments.
Prompting: Least-to-Most (LTM) and Transfer of Stimulus Control
- Prompting should be least intrusive and gradually faded to transfer control to the SD.
- Prompt timing: prompts are given immediately after the SD and before the response, to facilitate transfer of control from prompt to SD.
- Types of prompts:
- Verbal prompts (partial verbal prompts): e.g., saying part of the correct answer.
- Gesture prompts: pointing or tapping a location.
- Physical prompts: full, partial, or light touch to guide the learner.
- Most-to-least prompting is used rarely with a new skill (errorless learning) but can be used at the start if necessary.
- Fade strategy: move from full prompt → 3/4 prompt → 1/2 prompt → 1/4 prompt → no prompt, while sometimes alternating left/right positions to prevent location-based learning.
- Important concept: transfer of stimulus control means the learner should respond correctly to the SD without prompts over time.
- A no-prompt trial should follow a prompted trial to assess independent responding.
- When prompt is a part of the response criteria, the prompted response can be considered correct as part of the acquisition phase.
- If prompting becomes ineffective or leads to prompt dependence, adjustments to prompting strategies are made by the supervisor.
- “Reset the field” option may be used to reduce aversive experiences or when prompts hinder progress (used under supervision).
Phases of DTT (Mass Trial, Expanded Trial, Discrimination Training, Random Rotation)
- Mastery criteria (typical):
- The client must score 80–100% correct across 2–3 different individuals in 2–3 days at each phase before moving to the next phase.
- Mastery in random rotation specifically is required before moving on; supervisors may adjust criteria per client.
- Mass Trial (initial phase):
- Repeated presentation of the same target; typical plan is 10 trials per target, but may be fewer if the task is more effortful.
- Field size: typically 2–4 items; beware if only two items are present, guessing may produce ~50% accuracy.
- Notation on data sheets: m/k/d stands for Mastered Known Distractors (MKD) when a target is mastered across random rotation.
- Example: target = circle; spacing demonstrates “give me circle” responses; once mastery is achieved by two or more assessors over 2–3 days, move to the next target.
- Expanded Trial: (retention across distractors)
- Tests whether the client can retain the target response across distractors beyond the original target.
- Pattern example: Target → Known Distractor → Target → Distractor1 → Distractor2 → Target (and so on), increasing the number of distractors.
- Known distractors are distractors that the learner has previously mastered; if using multiple programs, distractors may come from other programs as needed.
- The pattern is continued until mastery across Distractors 1, 2, 3 is demonstrated; then return to the target.
- If errors occur after no-prompt, continue the pattern; if errors occur on the target after a no-no prompt, consider returning to mass trial.
- Discrimination Training: (two targets with built-in prompts)
- Purpose: improve discrimination between two targets that may look or sound similar.
- Built-in prompting: start with target 1 at full positional prompt, then fade to 3/4, 1/2, 1/4, and finally no prompt; then switch to target 2 with the same prompting sequence.
- The prompt is considered part of the response criteria and is counted as correct when used (e.g., circle with full prompt, then circle with 3/4 prompt, etc.).
- If errors occur, adjust prompts (e.g., move closer to the learner for the incorrect target using more intrusive prompts).
- Randomization: the order of targets and distractors can be varied; no fixed pattern should be relied upon.
- Random Rotation: (no pattern; higher-level discrimination)
- Tests discrimination across multiple targets and distractors in a random order.
- If a learner makes an error, a no-no prompt is used after the first incorrect response to guide a retry, followed by prompting only after a second consecutive error.
- After prompting, the learner should perform an independent trial to ensure transfer of learning.
- The order of targets and distractors is randomized to prevent anticipation and promote generalization.
- The learner must demonstrate mastery (often 80% or higher) without prompts across multiple trials before graduation to the next set of targets.
Mastery Data and Notation
- Mastery across phases is tracked with data sheets; notations include MKD (Mastered Known Distractors) and m/k/d on the day sheet.
- Data-based decisions are made by supervisors; criteria can be adjusted based on client needs and progress.
Error Correction and Phase-Specific Strategies
- Mass Trial
- Errorless learning is common for new targets; start with full prompt and fade to independent responding.
- After an incorrect response, prompt immediately and then provide an opportunity to respond independently (prompt → independent trial).
- End the session on a correct response when possible to maintain motivation and a positive learning end.
- Expanded Trial
- Correct errors after target with a no-no prompt if the error occurs after a target response.
- If errors occur after distractors, continue the pattern without changing the overall sequence; only the target-specific errors trigger prompting.
- Discrimination Training
- After an error, increase the prompt back toward a full prompt for the item that was incorrect, while keeping the other item in a lower prompt level.
- If the learner struggles with the second target, revert to a more prompt-supported position and re-test.
- Random Rotation
- Use a no-no prompt after the first incorrect response to a target or distractor, then provide an immediate prompting trial if needed.
- After prompting, ensure an independent trial is provided to confirm retention.
- Maintain a no-pattern, randomized order to support true discrimination, not memorization of order.
Practical Considerations for Running DTT
- Pacing: keep the time between trials short to maintain engagement and reduce drift (e.g., avoid long breaks between trials).
- Motivating Operations: conduct frequent preference assessments; establish motivating deals or tokens to increase readiness and motivation.
- Readiness and Transition: reinforce learning readiness skills like transitioning to the table, attending to materials, and listening to instructions.
- Reinforcement delivery: deliver reinforcers immediately after a correct response; pair social reinforcement with other preferred reinforcers for broader learning relevance.
- Consistency: reinforce the same correct response consistently across all staff; inconsistent reinforcement can confuse the learner and impede stimulus control.
- Appropriateness of reinforcers: match the reinforcer to the difficulty and effort required; avoid overly long or aversive reinforcers that disrupt the flow of learning.
- Avoid prompt dependence: fade prompts as quickly as possible to transfer control to the SD; begin with least intrusive prompting and adjust as needed.
- Ethical and practical concerns: DTT should not be the only approach; integrate with naturalistic teaching to support generalization and functional use in real life.
Examples and Illustrations (Key Scenarios)
- SD: a red color card is shown and the learner is asked to name the color; correct response is “red” (orange would be incorrect unless the program allows approximations).
- SD: a cat is shown and the learner is asked to say the cat’s sound; correct response could be “meow meow” (or shorter forms if allowed by the program).
- SD: “give me the big block” with big and small blocks; correct response is the learner hands the big block.
- Expanded/Discrimination scenarios demonstrate how to escalate or adjust prompts as distractors are added or altered.
Ethical and Real-World Relevance
- DTT provides a strong foundation of discrete skills before transitioning to real-life, functional use.
- The disciplined structure is a means to an end, not an end in itself; the goal is generalization and independence across settings (home, school, community).
- Team alignment and clear, agreed-upon response definitions prevent confusion and promote consistent reinforcement across different staff members.
Quick Recap of Core Concepts
- DTT is a structured, table-based method using the three-term contingency with clearly defined SDs, responses, and consequences.
- The SD controls behavior; reinforcement increases the likelihood of the target response in the presence of the SD.
- Responses must be observable and fully defined; various forms of responses (correct, incorrect, self-correction, nonresponse) are managed with phase-appropriate error correction.
- Phases (Mass Trial, Expanded Trial, Discrimination Training, Random Rotation) provide a progressive framework to build skill and generalize learning.
- Mastery criteria and data collection guide progression; MKD notation tracks durability of learning across distractors.
- Prompting, fading, pacing, motivating operations, and reinforcement strategies are all critical for effective DTT implementation.
Practice Prompts and Transfer of Control (Summary Tips)
- Prompt after SD, not before or after the response; the goal is to transfer control to the SD.
- Start with least intrusive prompts; fade quickly to avoid prompt dependence.
- Consider a reset of the field if prompts are aversive or if a learner is overwhelmed.
- Use pacing, motivating operations, and smooth transitions to maintain engagement and maximize learning opportunities.
Key Formulas and Notation (LaTeX)
- Three-term contingency: ext{Antecedent}
ightarrow ext{Behavior}
ightarrow ext{Consequence} - Discriminative Stimulus (SD) and response relationship: SD
ightarrow ext{Response (B)}
ightarrow ext{Reinforcement/Feedback (C)} - Contingency effectiveness (example):
- Mastery criteria (typical):
- Field size in Mass Trial:
- Prompt levels: full, 3/4, 1/2, 1/4, none (as part of the fading sequence)
- Three-term contingency: ext{Antecedent}
References to Practice and Assessment
- After covering these notes, you’ll engage in hands-on practice with actual DTT sessions under supervision.
- You may encounter variations in mastery criteria based on client needs and supervisor decisions.
Final Takeaway
- DTT is a powerful, structured approach to teaching discrete skills; when implemented with fidelity, it builds a solid skill foundation and supports later generalization to daily life. Proper pacing, prompting, reinforcement, and data-based progress checks are essential for success.