Ewakaa's picture
Initial public release: TRACE v1.0.0
c427f62 verified
# Session Interpretation taxonomy — TRACE v1
# The controlled vocabulary the Task 2 generator draws from. Self-contained.
#
# Citations for underlying concepts: see taxonomy-v1.md.
task:
id: session_interpretation
name: Behavioral Session Interpretation
citation: >
Multi-session interpretation framework is TRACE's operationalization.
Underlying concepts grounded in CHH (2020) Ch. 6–7 (analyzing behavior change),
Iwata 1982/1994 (functional analysis), Hanley-Iwata-McCord 2003 (FBA review),
Bijou-Peterson-Ault 1968 (ABC recording), Michael 1993 (MOs).
# 12 session patterns (each drives trajectory + interpretation)
patterns:
- id: mastery_progression
name: Mastery progression
description: "Steady improvement toward mastery criteria."
trend: ascending
concern_level: none
default_escalation: 1
citation: "CHH Ch. 6."
- id: regression
name: Regression
description: "Decline in performance after previous mastery or near-mastery."
trend: descending
concern_level: high
default_escalation: 3
citation: "CHH Ch. 26 (maintenance)."
- id: plateau
name: Plateau
description: "Accuracy stable but below mastery criteria for extended period."
trend: flat
concern_level: moderate
default_escalation: 2
citation: "CHH Ch. 7."
- id: frustration_pattern
name: Frustration pattern
description: "Declining accuracy accompanied by behavioral indicators of frustration and escape-maintained responding."
trend: descending
concern_level: high
default_escalation: 3
citation: "Iwata 1994 (escape function); CHH Ch. 27."
- id: variable_performance
name: Variable performance
description: "Inconsistent accuracy across sessions with no clear trend."
trend: variable
concern_level: moderate
default_escalation: 2
citation: "CHH Ch. 5 (measurement reliability)."
- id: prompt_dependency
name: Prompt dependency
description: "High accuracy with prompts but failure at independent level."
trend: flat_high_prompted
concern_level: moderate
default_escalation: 2
citation: "Touchette & Howard 1984 JABA (time delay)."
- id: rapid_acquisition
name: Rapid acquisition
description: "Quick skill acquisition exceeding expected timeline."
trend: sharply_ascending
concern_level: none
default_escalation: 1
citation: "CHH Ch. 6."
- id: generalization_failure
name: Generalization failure
description: "Strong performance in training but failure to generalize to novel conditions."
trend: context_dependent
concern_level: moderate
default_escalation: 2
citation: "Stokes & Baer 1977 JABA; CHH Ch. 26."
- id: extinction_burst
name: Extinction burst
description: "Temporary increase in problem behavior during behavior reduction."
trend: temporary_spike
concern_level: expected
default_escalation: 2
citation: "CHH Ch. 24."
- id: skill_loss_after_break
name: Skill loss after break
description: "Performance decline following extended absence or break, with recovery."
trend: drop_then_recovery
concern_level: moderate
default_escalation: 2
citation: "CHH Ch. 26 (maintenance)."
- id: motivating_operation_shift
name: Motivating operation shift
description: "Responding drops when MO changes (e.g., satiation); recovers when MO restored."
trend: mo_dip
concern_level: moderate
default_escalation: 2
citation: "Michael 1993 JEAB; CHH Ch. 16."
- id: setting_event_trigger
name: Setting event trigger
description: "Accuracy or behavior changes correlated with an external setting event (illness, sleep, schedule change)."
trend: setting_event
concern_level: moderate
default_escalation: 2
citation: "Smith & Iwata 1997 JABA; Bijou & Baer 1961."
# Behavior functions
functions:
- id: escape
name: escape
reinforcer: "Termination or avoidance of an aversive stimulus (demand, task, interaction)."
typical_antecedents: [demand presentation, difficult task, non-preferred activity, transition to work]
typical_consequences: [task removed, demand withdrawn, redirected to preferred activity]
- id: attention
name: attention
reinforcer: "Attention from others (adult or peer)."
typical_antecedents: [adult attention diverted, peer interaction, low-attention period]
typical_consequences: [adult attends, reprimand delivered, peer interaction initiated]
- id: tangible
name: tangible
reinforcer: "Access to an item or activity."
typical_antecedents: [preferred item removed, restricted access, another peer has item]
typical_consequences: [item returned, access granted, activity restored]
- id: automatic
name: automatic
reinforcer: "Self-produced reinforcement (sensory, proprioceptive)."
typical_antecedents: [low-stimulation period, alone time, waiting]
typical_consequences: [no social mediation; behavior self-terminates or continues]
- id: unknown
name: unknown
reinforcer: "Insufficient data to determine function."
# Target behaviors (deceleration targets)
# For full operational definitions + citations see taxonomy-v1.md.
# Used when the generator samples 0–3 behaviors per log.
target_behaviors:
- id: tantrum
name: Tantrum
operational: "Co-occurring cluster of two or more of: crying/screaming, dropping to floor, kicking, hitting, throwing objects, lasting ≥ 3 s."
plausible_functions: [escape, attention, tangible]
typical_severity: moderate
- id: aggression
name: Aggression
operational: "Attempted or completed forceful contact directed toward another person (hit, kick, bite, scratch, pinch, throw-at)."
plausible_functions: [escape, tangible, attention]
typical_severity: high
- id: sib
name: Self-injurious behavior (SIB)
operational: "Any response that produces tissue damage or has potential to (head-hit, self-bite, face-slap, head-bang, skin-pick, self-pinch)."
plausible_functions: [automatic, attention, escape]
typical_severity: high
- id: elopement
name: Elopement
operational: "Full body crossing a designated boundary without adult approval."
plausible_functions: [escape, tangible, automatic]
typical_severity: high
- id: property_destruction
name: Property destruction
operational: "Hitting/kicking furniture or walls; throwing objects not meant to be thrown; tearing clothing, books, materials."
plausible_functions: [escape, attention, tangible]
typical_severity: moderate
- id: motor_stereotypy
name: Motor stereotypy
operational: "Repetitive non-functional motor movements (hand-flap, body-rock, finger-flick, spin)."
plausible_functions: [automatic]
typical_severity: low
- id: vocal_stereotypy
name: Vocal stereotypy
operational: "Non-contextual, non-communicative vocalizations outside appropriate conversational context."
plausible_functions: [automatic]
typical_severity: low
- id: non_compliance
name: Non-compliance
operational: "Failure to initiate a requested response within 5 s of an instructional prompt, OR active refusal."
plausible_functions: [escape]
typical_severity: low
- id: mouthing
name: Mouthing
operational: "Placing hand, fingers, or non-food object into the mouth outside scheduled meal/snack times."
plausible_functions: [automatic]
typical_severity: low
- id: pica
name: Pica
operational: "Placement of any inedible item past the plane of the lips, including ingestion."
plausible_functions: [automatic, attention]
typical_severity: high
- id: verbal_aggression
name: Verbal aggression
operational: "Yelling, cursing, name-calling, or verbal threats toward another person."
plausible_functions: [attention, escape]
typical_severity: moderate
- id: fecal_smearing
name: Fecal smearing (scatolia)
operational: "Reaching into diaper, pull-up, or pants to retrieve feces and/or spreading feces onto skin, clothing, walls, furniture, or other surfaces."
plausible_functions: [automatic, attention]
typical_severity: high
citation: "Piazza et al. 1996 JABA; Matson et al. 2008 (scatolia in ASD)."
- id: toileting_accident
name: Toileting accident (urine or bowel)
operational: "Session toileting data tracks both successful voids in the toilet (urine and BM) and accidents (voiding outside the toilet onto self, clothing, furniture, or floor). The deceleration target is accident frequency; in-toilet voids are recorded as context for toileting progress. Medical rule-out required before treating accidents behaviorally."
plausible_functions: [automatic, escape]
typical_severity: moderate
citation: "Azrin & Foxx 1971 JABA; Cicero & Pfadt 2002; LeBlanc et al. 2005."
# Measurement types used in session logs
measurement_types:
- id: accuracy_pct
short: "accuracy"
description: "Trial-based percentage correct (X/N)."
used_for: [skill acquisition]
- id: frequency
short: "freq"
description: "Raw count per session."
used_for: [discrete behaviors, target behaviors]
- id: rate
short: "rate"
description: "Count per minute or hour."
used_for: [target behaviors]
- id: duration
short: "duration"
description: "Total elapsed time behavior occurred."
used_for: [tantrum, on-task engagement]
- id: latency
short: "latency"
description: "Time from SD to response initiation."
used_for: [skill responsivity]
- id: partial_interval
short: "PIR"
description: "Proportion of intervals in which behavior occurred at any point."
used_for: [stereotypy, low-intensity behavior]
- id: momentary_time_sampling
short: "MTS"
description: "Behavior sampled at end of each interval."
used_for: [group observation]
# Escalation levels (safety-critical structured label)
escalation_levels:
- id: 1
label: "Continue monitoring"
description: "Current programming is effective; no changes."
- id: 2
label: "Adjust next session"
description: "Implement a specific programming change at the next session."
- id: 3
label: "Supervisor review"
description: "Elevate to supervising BCBA within 24–48 hours."
- id: 4
label: "Safety immediate"
description: "Stop program; contact supervisor / family; consider crisis-plan activation."
# Confidence levels (calibration signal)
confidence_levels:
- id: high
label: high
description: "Pattern and recommendation well-supported by log data."
- id: moderate
label: moderate
description: "Data supports the hypothesis but alternatives cannot be fully ruled out."
- id: low
label: low
description: "Insufficient data to rule in a specific pattern; recommend data collection before programming changes."
# Behavioral indicator phrases (pasted verbatim into session logs)
# Organized by cluster; Task 2 generator samples from here.
behavioral_indicators:
frustration:
- "Increased response latency (3–5× baseline)"
- "Pushing materials away from the work area"
- "Turning away from task or staff"
- "Vocal refusal ('no', 'I don't want to')"
- "Self-stimulatory behavior increase during demand periods"
- "Crying during demand presentation"
- "Elopement from the work area"
- "Attempts at property destruction (throwing materials, swiping items off the surface)"
- "Aggression attempt directed at staff (hit, kick, or scratch toward adult)"
- "Peer aggression (hitting or pushing peers during group activities)"
- "Self-injurious behavior onset during task difficulty (self-biting, hair-pulling, head-hitting)"
engagement:
- "Eye contact maintained throughout trials"
- "Quick response latency (within 2–3 seconds)"
- "Reaching for materials proactively"
- "Smiling during reinforcement delivery"
- "Requesting more trials verbally"
- "Appropriate sitting posture maintained"
disengagement:
- "Looking away during SD presentation"
- "Delayed responding (>5 seconds consistently)"
- "Stereotypic behavior during inter-trial intervals"
- "Leaving seat without permission"
- "Playing with non-task materials"
- "Echolalic responding (repeating SD rather than answering)"
# Teaching method metadata (for referencing methods in log programs)
teaching_methods:
- id: dtt
name: DTT
default_structure: "structured discrete trials, massed"
- id: net
name: NET
default_structure: "natural-environment embedded opportunities"
- id: task_analysis
name: Task Analysis
default_structure: "multi-step chaining, {chain_type}"
- id: fct
name: FCT
default_structure: "functional communication training, replacement response = {replacement}"
- id: prt
name: PRT
default_structure: "pivotal response training, child-choice, attempt reinforcement"
- id: bst
name: BST
default_structure: "staff-facing training program"