# DTT taxonomy — TRACE v1 # Self-contained: everything the DTT generator needs beyond shared primitives. # Citation: Lovaas 1987 JCCP; Smith 2001 Focus on Autism; CHH 2020 Chs. 17, 21. # # DTT fits array-based discrete-response skills. Only VB-MAPP domains where # the target is discrete selection / labeling / matching appear here. method: id: dtt name: Discrete Trial Training short: DTT citation: "Lovaas, O. I. (1987). JCCP, 55(1), 3–9. https://doi.org/10.1037/0022-006X.55.1.3; Smith, T. (2001). Focus on Autism, 16(2), 86–92." when_to_use: > Acquisition of discrete, decontextualized skills (receptive labels, tacts, matching, early imitation) — especially early in instruction when stimulus control must be established. # VB-MAPP domains appropriate for DTT (array-based discrete-response skills). # Other domains (Mand, Echoic, Motor Imitation, Intraverbal, Social, etc.) # belong to NET, PRT, or other method taxonomies. skill_domains: vbmapp: levels: - id: L1 name: Level 1 developmental_age: "0–18 months" - id: L2 name: Level 2 developmental_age: "18–30 months" - id: L3 name: Level 3 developmental_age: "30–48 months" domains: - id: tact name: Tact gloss: Labeling / naming L1: - tacts common objects (ball, cup, shoe) - tacts familiar people by name - tacts common actions (running, eating, sleeping) - tacts body parts - tacts common animals L2: - tacts colors of objects - tacts shapes (circle, square, triangle) - tacts adjectives (big/little, hot/cold) - tacts prepositions (in, on, under) - tacts emotions in self and others L3: - tacts community helpers and their roles - tacts categories (fruits, vehicles, clothing) - tacts features of objects (color, shape, function) - tacts past-tense events - tacts abstract concepts (same/different, first/last) - id: listener_responding name: Listener Responding gloss: Receptive language L1: - follows 1-step motor instructions (sit down, stand up) - selects correct item from an array of 2 - points to named body parts - points to named common objects - follows instructions involving objects (give me the cup) L2: - selects correct item from array of 4–6 - follows 2-step instructions - selects items by feature (find something red) - selects items by function (find something you eat with) - selects items by class (find an animal) L3: - selects items by multiple features simultaneously - follows conditional instructions (if X, then Y) - follows instructions involving temporal concepts (before, after) - id: vp_mts name: Visual Perceptual Skills & Matching-to-Sample gloss: VP-MTS L1: - matches identical objects - matches identical pictures - matches colors (identical chips) - matches shapes (identical blocks) - completes simple 3–4 piece puzzles L2: - matches non-identical items by category - sorts items into 2–3 categories - matches quantities (1–5) - matches upper-case letters - completes 8–12 piece puzzles L3: - matches associated items (sock-shoe, cup-plate) - sorts by multiple attributes simultaneously - matches upper to lower case letters - sequences 3–4 step picture sequences - reproduces block designs from model - id: lrffc name: LRFFC gloss: Listener Responding by Feature, Function, Class L1: - selects items by function (What do you drink from?) - selects items by feature (What is round?) - selects items by class (Find the animal) - selects by single function from array of 3 - selects by single visible feature from array of 3 L2: - selects items by function from array of 5–8 - selects items by feature from array of 5–8 - selects items by class from array of 5–8 - selects items by multiple features (round and red) - selects items by function when item not visible L3: - selects items given 2+ features/functions simultaneously - selects items by class with exclusion (animal but not a pet) - selects items by abstract features (something needed when cold) - selects items by comparison (which is heavier) - selects items by negative features (not round, not food) - id: reading name: Reading L1: - matches letters to identical letters - identifies own name in print - identifies 5–10 upper-case letters - matches words to identical words - tracks print left to right L2: - identifies all upper-case letters - identifies all lower-case letters - reads 10–20 sight words - sounds out CVC words (cat, dog, run) - reads simple 2–3 word phrases L3: - reads sentences with comprehension - reads short passages and answers questions - phonetic decoding for novel words - id: writing name: Writing L1: - traces horizontal and vertical lines - traces basic shapes (circle, cross) - copies basic shapes from model - traces letters of own name - writes own name from model L2: - writes own name independently - copies all upper-case letters from model - writes upper-case letters from dictation - copies simple words from model - writes numbers 1–10 - id: math name: Math L1: - rote counts to 10 - counts objects 1–5 with 1:1 correspondence - identifies numerals 1–5 - matches quantities to numerals 1–5 - identifies basic shapes in math context L2: - counts objects 1–20 with 1:1 correspondence - identifies numerals 1–20 - compares quantities (more/less/same) - solves single-digit addition with manipulatives - identifies coins by name L3: - adds single-digit numbers without manipulatives - subtracts single-digit numbers - identifies place value (ones, tens) - tells time to the hour and half-hour - solves simple word problems # Prompt hierarchies that can be used in DTT prompt_hierarchies: - id: most_to_least name: Most-to-Least (errorless) sequence: [full_physical, partial_physical, gestural, positional, independent] description: "Start with maximum assistance; fade systematically to independence." best_for: "Acquisition, early learners, safety-critical skills." - id: least_to_most name: Least-to-Most sequence: [independent, gestural, positional, partial_physical, full_physical] description: "Start with independent attempt; add assistance only as needed." best_for: "Learner with partial repertoire; promotes independence." - id: time_delay name: Time Delay (progressive) sequence: [0s_delay, 2s_delay, 4s_delay, 6s_delay, independent] description: "Increase latency between SD and prompt delivery across sessions." best_for: "Prompt-dependency prevention; ideal for fading." citation: "Touchette & Howard 1984 JABA." - id: graduated_guidance name: Graduated Guidance sequence: [hand_over_hand_full, hand_over_hand_light, shadow, independent] description: "Hand-over-hand guidance with fading physical pressure." best_for: "Motor skills, self-care chaining." - id: stimulus_fading name: Stimulus Fading sequence: [exaggerated_stimulus, moderate_fade, minimal_fade, natural_stimulus] description: "Modify the stimulus itself; fade cue salience over sessions." best_for: "Receptive discrimination, early reading." - id: stimulus_shaping name: Stimulus Shaping sequence: [modified_stimulus, intermediate_shape, target_stimulus] description: "Change the stimulus topography to approximate the target." best_for: "Complex visual discriminations." # Error correction procedures available for DTT # Citation: Cooper, Heron, Heward (2020), Ch. 21. error_corrections: - id: transfer_trial name: Transfer trial steps: > Re-present SD -> provide effective-level prompt -> reinforce prompted response -> distractor trial -> re-present SD independently. - id: four_step name: 4-step error correction steps: > Model correct response -> lead (do together) -> test (independent) -> distractor -> retest. - id: backstep name: Backstep steps: > Return to previous prompt level that produced success -> successful response -> re-attempt at target prompt level. - id: simple_correction name: Simple correction steps: > "No, watch me" -> model -> re-present SD -> differential reinforcement for correct response. - id: errorless name: Errorless (preventive) steps: > Prevent errors by using high prompts from the start; fade gradually. Paired with most-to-least prompt hierarchy. # Reinforcement schedules relevant for DTT (acquisition and thinning) # Citation: CHH (2020) Ch. 13, 22. # Note: DRO/DRA/DRI are behavior-reduction schedules handled in FCT taxonomy. reinforcement_schedules: - id: crf name: CRF (continuous reinforcement) description: "Every correct response reinforced." typical_phase: "Acquisition." - id: fr2 name: FR-2 (fixed ratio 2) description: "Every 2nd correct response reinforced." typical_phase: "Thinning as accuracy stabilizes." - id: vr3 name: VR-3 (variable ratio 3) description: "Average of every 3rd correct response reinforced." typical_phase: "Maintenance, resistance to extinction." - id: token_economy name: Token economy description: "Tokens delivered per correct response; exchanged at session end for a back-up reinforcer appropriate to the learner (for children: candy, cookies, brief leisure-time access, preferred activity, screen time, or a favorite sensory item)." typical_phase: "Group / classroom settings, mastered / generalization phase." # Mastery criteria common for DTT programs # Citation: CHH (2020) Ch. 28. mastery_criteria: - id: mc_80_2sessions text: "80% accuracy across 2 consecutive sessions with at least 10 trials each." - id: mc_90_3sessions text: "90% accuracy across 3 consecutive sessions." - id: mc_80_generalization text: "80% accuracy with 2 different therapists and 2 different settings." - id: mc_80_independent text: "80% accuracy at independent prompt level across 3 sessions." - id: mc_90_first_trial text: "90% first-trial-correct across 5 consecutive sessions." - id: mc_fluency text: "Fluency criterion: correct responses within 3-second latency at 90% accuracy." - id: mc_natural_env text: "Demonstrated use in natural environment across 3 independent instances."