Applied behaviour analysis

1. Introduction

Version followed: contemporary assent-based, trauma-informed ABA as set out in Hanley's Practical Functional Assessment and Skill-Based Treatment (PFA/SBT; Practical Functional Assessment and Treatment Notebook, rev. March 2018) and Hanley (2020) "happy, relaxed, engaged" (HRE), with the BACB Ethics Code for Behavior Analysts (2020, in force 2022), Taylor, LeBlanc & Nosik (2019) on compassionate care, and Breaux & Smith (assent). Lovaas-era early intensive behavioural intervention (EIBI) and discrete trial training are described historically; contemporary discrete-trial teaching is given only in its assent-based form. Aversive procedures are described historically only.

Applied behaviour analysis is not a single therapy but an applied science and service framework derived from Skinner's radical behaviourism. Baer, Wolf & Risley (1968) defined it by seven dimensions: applied (socially important behaviour), behavioural (observable and measurable), analytic (a functional relation between procedure and change must be shown), technological (procedures described so others can replicate them), conceptually systematic, effective (change large enough to matter in the person's life), and general (change carries over to new people, places and materials). [source: Wikipedia ABA summarising Baer, Wolf & Risley 1968]

Core idea of change: behaviour, including behaviour others find challenging, is maintained by its relation to the environment (what evokes it and what follows it). Change comes from (a) identifying the conditions and reinforcers that maintain a behaviour, (b) arranging environments so that the person's needs are met safely, and (c) teaching skills that reach the same outcomes more efficiently. The contemporary form adds that teaching should happen only when the learner is safe, relaxed and willingly engaged, and that the learner's assent and withdrawal of assent shape the programme. [source: Hanley 2018 notebook; UNL ASD Network HRE handout citing Hanley 2020; Beadle & Tirella 2023 citing Breaux & Smith]

Scope note for psychotherapists [synthesis]: in many jurisdictions designing and supervising ABA programmes is a regulated activity (e.g. BCBA certification in the US; front-line staff work as registered behaviour technicians under supervision). BACB Code 2.01 limits practice to services within one's education and training. Use this guide to understand, collaborate with, or (if credentialed or supervised) deliver the procedures; seek supervision before running a functional analysis.

2. Framework

ABA services are usually described as either comprehensive (many skill domains, high weekly hours, historically EIBI) or focused (one or a few target behaviours, e.g. severe problem behaviour). PFA/SBT is a focused, problem-behaviour pathway that can sit inside a comprehensive programme. [source: comprehensive vs focused distinction from BACB/CASP funder guidelines; uncertain]

Phase 0. Intake, consent, assent and medical screen

Goals: obtain and document informed consent from the legal decision-maker; plan how assent will be read for this learner; involve the learner and family in goal selection; rule out or refer for medical contributors (pain, sleep, constipation, seizures, medication effects). [source: BACB Code 2.09, 2.11, 2.12]
Format: one or more meetings with caregivers and, as far as possible, the learner.
Move on when: consent documented, assent indicators and withdrawal indicators operationally defined, medical referrals made where indicated. [source: Beadle & Tirella 2023; BACB 2.11, 2.12]

Phase 1. Building HRE and rapport

Goal: establish a context in which the learner is simultaneously happy, relaxed and engaged; learn preferences and communication style. [source: UNL HRE handout citing Hanley 2020; Gover et al. 2022]
Format: free-access, no-demand sessions with preferred items freely available.
Move on when: the learner reliably shows HRE with the clinician; for learners with a history of severe problem behaviour the handout recommends at least 5 consecutive minutes of HRE before any adult-led time. [source: UNL HRE handout]

Phase 2. Practical functional assessment

2a. Open-ended interview with the people who know the learner best: define the problem behaviour and its precursors, the evoking situations (establishing operations) and the suspected reinforcers. [source: Hanley 2018 notebook]
2b. Interview-informed synthesized contingency analysis (IISCA): brief (about 5-minute) alternating control and test sessions, typically ordered control, test, control, test, test. [source: Hanley 2018 notebook]
Format: the interview, observation and IISCA can be completed in about 30-minute sessions or less. [source: Commonwealth Autism best-practice blog]
Move on when: problem behaviour (or its precursors) is near zero in control and in reinforcement intervals and reliably occurs in test, the learner responds quickly to the reinforcers, nothing escalates, and no emergency procedures or persistent distress occurred. [source: Hanley 2018 notebook]

Phase 3. Skill-based treatment (SBT) Shaped in sequence within the same contingency the IISCA identified:

Functional communication response (FCR): simple, then intermediate, then complex.
Delay/denial tolerance response.
Contextually appropriate behaviours (CABs) during delays, increasing in number and difficulty.
Reinforcement thinning toward the natural density of the setting.

Format: short sessions in an "open-door" model without physical management. [source: Hanley 2018 notebook; ACT Community Hanley workshop page]
Move on (step to step) when: the current response is fluent and problem behaviour stays low; move to the next reinforced delay only after the expected CABs occur. [source: Hanley 2018 notebook]
Session count: not specified in the sources. [uncertain]

Phase 4. Generalization and maintenance

Extend to relevant people (teachers, other caregivers), then to relevant contexts (school, home), then over longer periods; experienced implementers first, less experienced caregivers after mastery, with treatment-integrity checks. [source: Hanley 2018 notebook]
Move on (to fade-out or monitoring) when caregivers implement with integrity, effects hold across people and settings, and goals selected by learner and family are met. [synthesis]

Phase across all: skill acquisition (communication, daily living, play, social, academic) using task analysis, chaining, prompting, shaping, naturalistic teaching and assent-based discrete-trial teaching (section 4). Continual evaluation of graphed data drives every change. [source: BACB 2.18; Wikipedia ABA technique summaries]

Dose. Lovaas-model EIBI ran up to about 40 hours a week of one-to-one teaching for two or more years [historical; source: Lovaas 1987 as summarised by UO Autism History Project and Wikipedia]. Contemporary focused PFA/SBT dosage is set by the problem and the data rather than a fixed hour count [uncertain: no figure in the sources]. Hours should be justified by the learner's goals and assent, not by a standard package [synthesis].

Historical framework (Lovaas/UCLA Young Autism Project, from 1970). Children aged under about four received intensive home-based one-to-one discrete-trial teaching delivered by trained student therapists and parents, progressing through curricula from attending and imitation to language and school readiness, then into mainstream preschool. Lovaas (1987) reported that 9 of 19 children in the intensive group performed in the normal range in first grade; critics (e.g. Schopler) questioned sample assignment and the outcome measures. Early UCLA work used contingent aversives (slaps, and electric shock in some studies) as a stated last resort for self-injury; Lovaas later said learning often did not generalise or last. Described here historically; no aversive procedure is described operationally in this guide. [source: UO Autism History Project; Wikipedia ABA]

3. Guiding principles

Social significance first. Choose targets that matter in the person's life, chosen with the person and family. Practical meaning: ask "who benefits and how?" for every goal. [source: Baer et al. 1968 via Wikipedia; BACB 2.09]
Measure observable behaviour. Define targets so a third party could count them. Practical meaning: write each target as what you would see or hear. [source: Baer et al. 1968 via Wikipedia]
Demonstrate function before treating. Practical meaning: no reduction plan before a functional assessment. [source: Baer et al. 1968; Hanley 2018; BACB 2.13, 2.14]
Technological description. Practical meaning: every procedure is written down well enough that a substitute could run it. [source: Baer et al. 1968; BACB 2.14]
Generality is designed, not hoped for. Practical meaning: plan for new people, places and materials from the start. [source: Baer et al. 1968; Hanley 2018]
Effectiveness is judged by the data and by the person's life. Practical meaning: graph, review often, and change what is not working. [source: BACB 2.18]
Happy, relaxed, engaged comes first. Teaching is placed inside HRE time, not the other way round. Practical meaning: if HRE is lost, return to it before teaching anything. [source: UNL HRE handout citing Hanley 2020 and Evans 2023]
Assent is ongoing and can be withdrawn. Assent withdrawal is information that something in the environment is aversive. Practical meaning: pause, change something, and honour "no" however it is communicated. [source: Beadle & Tirella 2023 citing Breaux & Smith; BACB 2.11]
Trauma-assumed / trauma-informed. Assume the learner may have experienced trauma, including from past services. Practical meaning: prioritise safety, choice and control; avoid procedures that recreate coercion. [source: Rajaraman et al. 2022 abstract; ACT Community Hanley workshop page]
Positive reinforcement as the default. Practical meaning: build up what the person can do; do not plan around suppressing behaviour. [source: BACB 2.14]
Least intrusive effective procedure; restrictive or punishment-based methods only as a last resort after less intrusive methods have failed or when the risk of harm from the behaviour outweighs the risk of the procedure, with required reviews (e.g. human rights committee). Practical meaning: in assent-based practice, build programmes that do not need them at all. [source: BACB 2.15; synthesis for the second sentence]
Safety, no physical management. Practical meaning: assess around precursors rather than severe behaviour; design an open-door context. [source: Hanley 2018; ACT Community Hanley page]
Respect self-determination and do not target harmless autistic traits. Practical meaning: do not set goals to eliminate stimming, enforce eye contact or produce masking unless the behaviour causes harm and the person agrees. [synthesis, prompted by autistic critiques summarised in Wikipedia ABA; BACB core principles on dignity and self-determination]
Compassionate, collaborative relationships with caregivers. Practical meaning: listen actively, show empathy, explain rationale without jargon, share decisions, apologise for mistakes. [source: Taylor, LeBlanc & Nosik 2019]
Medical needs considered. Practical meaning: refer for pain, sleep or other medical contributors before or alongside behavioural treatment. [source: BACB 2.12]
Accountability through supervision and integrity checks. Practical meaning: measure whether implementers run procedures as written. [source: Hanley 2018 notebook]

4. Intervention inventory

Grouped as A. relationship, consent and assessment; B. skill-based treatment of problem behaviour (PFA/SBT); C. skill-acquisition procedures; D. caregiver and system procedures; E. historical procedures (described, not taught). All example wording is illustrative unless quoted and cited.

A1. Informed consent, assent planning and goal co-selection

Purpose / when: before any assessment or intervention, before major changes, and whenever circumstances change. [source: BACB Code 2.09, 2.11]
Steps:
1. Explain in plain language what will be assessed, what procedures may be used, risks, alternatives and the right to stop; document consent from the legal decision-maker and renew it when the plan changes.
2. With caregivers (and the learner where possible), list how this learner shows willingness (approaches, smiles, reaches for materials, cooperates readily) and how they show withdrawal (precursor behaviour, slowed or prompted responding, turning away, protest, saying "no", stiff posture, covering face, leaving). Write both lists as operational definitions. [source: Beadle & Tirella 2023]
3. Ask the learner directly what they want help with, using their communication system; ask caregivers what they love about the learner and what would make daily life better. [source: Hanley 2018 notebook for caregiver question; synthesis for the learner interview format]
4. Agree goals; drop or reshape goals that target harmless traits rather than wellbeing or safety. [synthesis]
Example language (illustrative): "You can tell us to stop at any time, and if [name] shows us they don't want to continue, we'll stop and change what we're doing."
Watch for: consent obtained only for a package of hours; goals written about compliance rather than skills the person values. Working when the learner and family can each name why each goal matters to them.

A2. Medical and physiological screen

Purpose / when: at intake and whenever problem behaviour starts or changes abruptly. [source: BACB 2.12]
Steps: ask about pain, ear or dental problems, gastrointestinal symptoms, sleep, seizures, menstrual cycle, medication changes; refer to the relevant physician; document referral and follow up. [source: BACB 2.12 for referral and documentation; synthesis for the checklist]
Example language (illustrative): "Before we look at what's happening around the behaviour, I want to make sure nothing is hurting."
Watch for: new self-injury aimed at one body area (e.g. head-hitting near an ear) suggests a medical cause. [synthesis]

A3. Establishing happy, relaxed and engaged (HRE) / pairing

Purpose / when: first sessions with every learner, and again whenever HRE is lost. Teaching opportunities are placed within HRE time. [source: UNL HRE handout citing Hanley 2020; Gover et al. 2022]
Steps: [source: UNL HRE handout]
1. Gather preferences and communication style from the learner, people who know them and preference assessments (A4).
2. Make the space safe and comfortable with preferred items freely available.
3. Remove all demands; signal that the learner decides what happens next.
4. Get to the learner's level, respond to their bids, comment in ways that do not upset them.
5. Do not touch their items unless invited; when invited, join as a shared experience rather than supervising.
6. Rotate a variety of preferred items and activities.
7. For learners with a history of severe problem behaviour, allow at least 5 consecutive minutes of HRE before adult-led time; between teaching opportunities, return to learner-led HRE for at least 30 consecutive seconds.
Example language (illustrative): "You pick. I'll hang out here." / "Wow, that tower is tall."
Watch for: HRE requires all three (happy, relaxed, engaged) at once. A learner who is engaged but tense, or quiet but avoiding you, is not in HRE. Working when the learner approaches you, invites you into play and stays calm when you arrive.

A4. Preference assessment

Purpose / when: identifying items, activities and social interactions to make freely available and to use as reinforcers; repeat when motivation drops or assent is withdrawn. [source: Beadle & Tirella 2023 (repeat assessment after withdrawal); Wikipedia ABA]
Steps: formats commonly used in the field include asking the learner and caregivers, free-operant observation (watch what the learner chooses with everything available), paired-stimulus (offer two items at a time and record choices), and multiple-stimulus without replacement (array of items, chosen item removed, repeat). Rank items by selection. [uncertain: formats from the standard behaviour-analytic literature, e.g. Fisher et al. 1992; DeLeon & Iwata 1996; primary papers not drawn on]
Example language (illustrative): "Which one?" (holding two items at equal distance).
Watch for: preferences shift within a day; free-operant observation avoids removing items, which suits learners for whom item removal is distressing. [synthesis]

A5. Open-ended interview (PFA step 1)

Purpose / when: first step of practical functional assessment for problem behaviour. [source: Hanley 2018 notebook]
Steps: [source: Hanley 2018 notebook]
1. Interview together the people who spend most time with the learner; aim for consensus.
2. Ask them to recount two recent serious episodes in detail, then ask about triggers, early warning signs (precursors) and how people usually respond.
3. If information is still missing, ask hypothetical questions: what are the warning signs, when would staff need backup, what reliably stops it, what reliably sets it off.
4. Ask what they love about the learner and what the learner enjoys.
5. Explain next steps and address concerns.
6. Output: a definition of the problem behaviour class and its precursors, the evoking situation(s) and the suspected synthesized reinforcers (often several together, e.g. escape plus access to a tablet plus adult attention delivered the learner's way).
Example language (illustrative): "Talk me through the last time it got really hard. What was happening right before?" / "What does it look like a minute before it escalates?"
Watch for: caregivers' guilt or blame; use the compassionate-care repertoire (D2). Working when you can predict a situation that will reliably evoke the precursors and what will reliably end them.

A6. Interview-informed synthesized contingency analysis (IISCA) (PFA step 2)

Purpose / when: confirming the hypothesised contingency experimentally and safely before treatment. Requires supervision by someone trained in functional analysis. [source: Hanley 2018 notebook; BACB 2.01, 2.13]
Safety and consent steps: identify the range of intensities and injury risk beforehand; target precursors or early warning behaviours rather than severe behaviour; plan to deliver the reinforcers noncontingently during any crisis; keep non-treating caregivers on their usual procedures. [source: Hanley 2018 notebook]
Steps: [source: Hanley 2018 notebook]
1. Build the test condition first from the most challenging, easily replicated situation in the interview.
2. Build the control condition by removing only the synthesized reinforcement contingency: in control, the suspected reinforcers are available freely.
3. Use consistent cues to signal each condition (e.g. your body position).
4. Run sessions of about 5 minutes, typically control, test, control, test, test.
5. In test, when a precursor or problem behaviour occurs, deliver the synthesized reinforcement for about 30 seconds, then re-present the evoking situation.
6. Adjust: if control shows problem behaviour, revise the control; if test shows little or delayed problem behaviour, run more tests or revise the test.
7. For highly reactive learners, lengthen and vary reinforcement intervals, or run the analysis in a natural setting with a familiar person.
Example language (illustrative): test: "Time to put the tablet away and come do some work." Reinforcement interval: "Okay, you can have it, no work."
Watch for / success criteria: near-zero problem behaviour in control and reinforcement intervals; quick responding to reinforcers; no escalation; no persistent distress; no emergency procedures. Stop and revise if any of these fail. [source: Hanley 2018 notebook]

A7. Measurement, single-case evaluation and assent data

Purpose / when: throughout; every decision rests on graphed data. [source: BACB 2.18; Baer et al. 1968 via Wikipedia]
Steps:
1. Define each target and record its frequency, duration or percentage of opportunities.
2. Record assent-withdrawal indicators alongside skill data, graph and analyse them together. [source: Beadle & Tirella 2023]
3. Graph sessions; review at fixed intervals; where feasible show control by a design such as reversal or multiple baseline across settings or behaviours. [source: Wikipedia ABA]
4. Collect social-validity information (do the learner and family like the goals, procedures and results) [synthesis; the BACB Code addresses this as stakeholder preference, 2.14].
Watch for: improving skill data with rising withdrawal data means the programme is coercive; change it. [synthesis based on Beadle & Tirella]

B1. Functional communication response (FCR) training

Purpose / when: first SBT skill; gives the learner an efficient, socially acceptable way to obtain the same synthesized reinforcers that problem behaviour produced in the IISCA. [source: Hanley 2018 notebook; FCT lineage from Carr & Durand 1985]
Steps: [source: Hanley 2018 notebook unless marked]
1. Start from HRE and the control condition; present a mild version of the evoking situation.
2. Teach a simple request first (the notebook's example is "My way"), using modelling, role play, feedback and prompts faded as quickly as the learner allows (C3).
3. Reinforce each request immediately with the full synthesized reinforcer at first.
4. Shape toward an intermediate, then a complex request that includes a polite opener and a request phrase suited to the learner's communication system (speech, signs, device, picture card).
5. Problem behaviour no longer produces the reinforcers in the standard format; see B5 for partial-extinction and full-assent options.
Example language (illustrative): modelling: "When you want to keep playing, you can say 'my way'. Let's try it." Learner says it; adult: "Sure, your way!" and returns the item.
Watch for: the request must work immediately and reliably at first or problem behaviour will be more efficient. Working when the learner requests independently in the evoking situation and precursors fall.

B2. Delay and denial tolerance response

Purpose / when: once the FCR is fluent; teaches the learner to cope when the request cannot be granted right away or at all. [source: Hanley 2018 notebook]
Steps: [source: Hanley 2018 notebook]
1. After a request, give a clear, consistent signal that it is delayed or denied.
2. Teach a tolerance response (notebook example: "Okay, no problem" or a non-vocal equivalent).
3. Reinforce the tolerance response intermittently, sometimes by granting the original request.
4. Keep delay lengths unpredictable; never state the exact requirement.
Example language (illustrative): "Thanks for asking nicely. Not right now." Learner shows tolerance response; adult: "Thanks for being cool with that! Here you go."
Watch for: precursors after the denial signal mean the step was too big; shorten the delay. Working when the learner accepts "not now" calmly in varied contexts.

B3. Contextually appropriate behaviours (CABs) during delays

Purpose / when: after tolerance; the learner does what the situation calls for during the delay (e.g. completes a task, cleans up, transitions) before the reinforcer returns. [source: Hanley 2018 notebook]
Steps:
1. Choose final CAB targets with the learner's family and teachers. [source: Hanley 2018 notebook]
2. Start with easy responses, then harder ones; increase the number required gradually (the notebook gives a progression from 1 response up to around 20). [source: Hanley 2018 notebook]
3. Deliver the reinforced delay only after the expected CABs occur.
4. Build CAB chains across the classes Hanley names: relinquishing, transitioning, starting, varying, enduring and persisting. [source: ACT Community Hanley workshop page; details of each, uncertain]
5. Offer choice models within the chain (e.g. which task, in which order, where) to reduce escalation. [source: ACT Community page names choice models; specific formats, uncertain]
Example language (illustrative): "Not yet. Let's do two of these first, then your way." After two: "Done! Your way."
Watch for: rising assent-withdrawal indicators as requirements grow; drop back a step. Working when the learner completes the expected chain and problem behaviour stays low.

B4. Reinforcement thinning and chained schedules

Purpose / when: after CABs; to bring reinforcement toward the natural density of the setting. [source: Hanley 2018 notebook]
Steps: [source: Hanley 2018 notebook]
1. Start rich and frequent; thin gradually (the notebook mentions working toward around 1 reinforcer per 10 responses).
2. Make reinforcement durations variable and unpredictable.
3. Once chains are established, delays can be embedded so the schedule feels continuous to the learner.
4. Structure the day into learner-led, adult-led and transition periods; reinforce skills intermittently (roughly 1 in 5 occasions).
Example language (illustrative): "Good job sticking with that. Free time!"
Watch for: resurgence of problem behaviour when thinning moves too fast; return to the last stable step. [source: resurgence named on ACT Community Hanley page; response is synthesis]

B5. Responding to problem behaviour during SBT: partial extinction or full assent format

Purpose / when: deciding what happens when problem behaviour occurs in treatment. [source: Hanley 2018 notebook]
Steps:
1. Choose, with caregivers and supervisor, between partial extinction (problem behaviour produces less of the reinforcer than the skill does, avoiding the escalation full extinction can cause) and the "full assent format", which keeps reinforcement available without extinction. [source: Hanley 2018 notebook]
2. In either format, if severe behaviour occurs during teaching: an empathetic statement, then return to learner-led HRE. If low-level behaviour occurs: an empathetic statement and encouragement to keep going. If severe behaviour occurs during HRE: an empathetic statement, give what the learner likely needs without requiring a request, and adjust the environment (dim lights, calming music, space). [source: UNL HRE handout]
Example language (illustrative): "That's really hard. You're done for now, here's your tablet."
Watch for: extinction bursts (a temporary rise) and the risk that inconsistent extinction makes behaviour more persistent. [source: Wikipedia ABA] Working when skills reliably outcompete problem behaviour without escalation. [synthesis]

B6. Teaching functional refusal and self-advocacy

Purpose / when: for every learner, early; makes assent withdrawal communicable and safe. [source: Beadle & Tirella 2023]
Steps:
1. Teach a refusal response the learner can produce easily: a picture icon, sign, or words such as "stop", "not now", "no".
2. Honour it every time in the early stages; remove the instruction, re-establish motivation, and try again later or with a changed instruction.
3. Do not treat refusal as non-compliance to be overridden. [source: Beadle & Tirella 2023 citing Bannerman et al. 1990]
Example language (illustrative): "You pressed 'not now'. Okay, we'll stop."
Watch for: fear among caregivers that honouring refusal teaches avoidance; pair with B2 and B3 so the learner also learns to tolerate and engage. [synthesis]

B7. Generalization and maintenance programming

Purpose / when: after SBT is stable with the primary implementer. [source: Hanley 2018 notebook]
Steps: [source: Hanley 2018 notebook]
1. Add relevant people one at a time (other caregivers, teachers).
2. Add relevant contexts (home, school, community).
3. Extend over longer time periods.
4. Check treatment integrity; experienced implementers first, less experienced caregivers only after mastery.
Example language (illustrative, to caregiver): "You'll watch me run it, then you'll run it with me here, then I'll step back."
Watch for: problem behaviour reappearing with a new person usually signals integrity drift (e.g. inconsistent signals or delays). [synthesis]

C1. Task analysis

Purpose / when: any multi-step skill (handwashing, dressing, using a device). [source: Wikipedia ABA]
Steps: perform or watch the skill done competently; list each observable step in order (e.g. handwashing: turn on tap, wet hands, apply soap, scrub, rinse, turn off tap); test the list by having someone follow it literally; record which steps the learner does independently at baseline. [source: Wikipedia ABA; testing step is synthesis]
Example language (illustrative): "First water on."
Watch for: steps too large for this learner; split them further. [source: Beadle & Tirella 2023 recommend breaking steps down further when assent is withdrawn]

C2. Chaining (forward, backward, total task)

Purpose / when: teaching a task-analysed sequence. [source: Wikipedia ABA]
Steps:
- Forward: learner does the first step independently, is helped through the rest; when step 1 is mastered, require steps 1 and 2, and so on.
- Backward: learner is helped through all steps except the last, which they complete independently; then the last two, and so on, so the learner always finishes the task and contacts its natural outcome.
- Total task: learner attempts every step every time, with help only where needed. [source: Wikipedia ABA]
Example language (illustrative): backward chain, jacket: help with arms and zip start, then "You pull it up!"
Watch for: pick backward chaining when finishing is itself rewarding; total task when the learner already has many steps. [synthesis]

C3. Prompting and prompt fading

Purpose / when: helping a correct response happen so it can be reinforced, then removing the help. [source: Wikipedia ABA; ASAT DTI page]
Steps: [source: Wikipedia ABA]
1. Choose prompt types: stimulus prompts (e.g. position the correct item closer, gesture) or response prompts (verbal, model, physical guidance).
2. Choose a strategy: most-to-least (start with the most supportive and fade), least-to-most (start with the least and add help only as needed), errorless (prompt immediately so errors are unlikely).
3. Fade systematically, e.g. physical guidance from hand-over-hand to light touch at the wrist to no contact.
4. In assent-based practice, ask before physical prompting and stop if the learner pulls away. [synthesis, consistent with Beadle & Tirella 2023 and Hanley's no-physical-management model]
Example language (illustrative): "Can I help your hand?" / gesture toward the cup.
Watch for: prompt dependence (waiting for help); needing more than two prompts on a familiar instruction is an assent-withdrawal indicator, not just a skill deficit. [source: Beadle & Tirella 2023]

C4. Shaping

Purpose / when: the target response does not yet occur at all, e.g. first spoken words. [source: Wikipedia ABA]
Steps: reinforce the closest current approximation (e.g. any vocal sound while reaching for bubbles); once it is reliable, reinforce only closer approximations ("b", "bu", "bubbles"); move criterion in small steps; drop back if responding stops. [source: Wikipedia ABA; dropping back is synthesis]
Example language (illustrative): learner says "buh" reaching for bubbles: "Bubbles!" (blow bubbles).
Watch for: raising the criterion too fast produces frustration and withdrawal.

C5. Discrete-trial teaching (assent-based contemporary form)

Purpose / when: structured teaching of discriminations, imitation, receptive language, academics, especially for learners who need many clear practice opportunities; combine with naturalistic teaching for generalization. [source: ASAT DTI page citing Smith 2001]
Steps: [source: ASAT DTI page citing Smith 2001; assent elements are synthesis drawing on UNL HRE handout and Beadle & Tirella]
1. Start only when the learner is in HRE and has assented (approaches the table or teaching area; it should be possible to leave).
2. Give a brief, clear instruction.
3. Add a prompt if needed (C3), fading over trials.
4. Learner responds.
5. Correct: immediate reinforcement (ideally the learner's chosen item or activity). Incorrect: immediate correction, e.g. guiding or modelling the correct response, then a fresh trial.
6. Short intertrial interval (about 1 to 5 seconds).
7. Intersperse mastered items, vary materials, and return to learner-led HRE frequently (at least 30 seconds between teaching opportunities for learners with severe behaviour histories).
Example language (illustrative): "Touch red." (learner touches red) "Yes, red!" (hands over the toy for a short play).
Watch for: decreased responding, more errors, needing more reinforcement to return, turning away: treat as assent withdrawal and change the session, not as behaviour to work through. [source: Beadle & Tirella 2023]

C6. Naturalistic and incidental teaching

Purpose / when: teaching within play and routines where the learner's own interest supplies motivation; aids generalization. [source: ASAT DTI page recommends incidental teaching alongside DTI; PRT described as naturalistic in Wikipedia ABA; Hart & Risley incidental teaching, uncertain]
Steps: [synthesis from standard practice; uncertain]
1. Arrange the environment so preferred items are visible but need help to obtain (e.g. a clear box).
2. Wait for the learner to initiate.
3. Prompt a slightly more elaborate response than the learner's current level.
4. Deliver the item or activity immediately as the natural consequence.
5. Follow the learner's lead to the next opportunity.
Example language (illustrative): learner reaches for the bubble jar: adult waits, then models "open" and opens it when the learner tries.
Watch for: sabotage that frustrates rather than motivates.

C7. Differential reinforcement

Purpose / when: strengthening an alternative behaviour while the problem behaviour produces less or no reinforcement. [source: Wikipedia ABA]
Steps: identify the function (A5, A6); choose an alternative that achieves it (DRA) or an incompatible behaviour (DRI), or reinforce periods without the behaviour (DRO); reinforce the alternative more richly and quickly than the problem behaviour; thin gradually (B4). [source: Wikipedia ABA for concept; variant labels from standard texts, Cooper, Heron & Heward 2020, uncertain]
Example language (illustrative): "Thanks for showing me the break card. Break time."
Watch for: DRO alone teaches no skill; prefer DRA. [synthesis]

C8. Antecedent and motivating-operation strategies

Purpose / when: making problem behaviour unnecessary and skills more likely by changing the setting before behaviour occurs. [source: Wikipedia ABA on motivating operations]
Steps:
1. Identify establishing operations from the interview (A5).
2. Provide identified reinforcers noncontingently during crises and early treatment. [source: Hanley 2018 notebook]
3. Offer choices of people, materials, locations and order; adapt the environment for sensory needs. [source: Beadle & Tirella 2023]
4. Use visual schedules, warnings before transitions and priming. [uncertain: common practice, not in the primary sources]
Example language (illustrative): "Puzzle first or drawing first?"
Watch for: antecedent changes that remove all learning opportunities; pair with B1 to B3. [synthesis]

C9. Token systems

Purpose / when: bridging delays to reinforcement for learners who understand exchange. [source: Wikipedia ABA mentions token economies]
Steps: [uncertain: standard practice, Cooper, Heron & Heward 2020] teach the exchange first (one token, immediate exchange); specify the behaviours that earn tokens; build the number required gradually; let the learner choose backup reinforcers; thin over time.
Example language (illustrative): "You got your star. Five stars, then trampoline."
Watch for: poorly managed thinning is a common problem; Hanley has cautioned against rigid rule adherence with token systems. [source: Wikipedia ABA; Behavioral Observations session 219 description]

D1. Caregiver and staff training (instruction, modelling, rehearsal, feedback)

Purpose / when: whenever caregivers, teachers or technicians will implement procedures; required before generalization (B7). [source: Hanley 2018 notebook (implementer sequence); Taylor et al. 2019 (role-play, video review, in-vivo practice)]
Steps:
1. Explain the procedure and its rationale without jargon.
2. Model it live or on video, showing good and poor examples.
3. Have the trainee rehearse in role play, then in vivo.
4. Give specific feedback; repeat until the trainee meets an integrity criterion.
5. Check integrity periodically after training. [source: Taylor et al. 2019 Table 5 activities; the four-part 'behavioural skills training' label is standard (e.g. Parsons, Rollyson & Reid 2012), uncertain]
Example language (illustrative): "Watch me do the 'not now' step. Then you try, and I'll tell you one thing that went well and one thing to change."
Watch for: caregivers agreeing in session but not implementing; ask about obstacles (D2).

D2. Compassionate collaboration with caregivers

Purpose / when: every contact with families; Taylor et al. argue outcomes may improve when the relationship is attended to. [source: Taylor, LeBlanc & Nosik 2019]
Steps: [source: Taylor et al. 2019]
1. Positive interaction: greet warmly, comment positively on the child and parent, ask how the parent is doing.
2. Empathy: attentive posture, open questions, pauses, paraphrase, name the parent's feelings, respond to non-verbal cues.
3. Compassion: make room for feelings, validate without judgement, frame concerns as a team effort, suggest actions that ease distress.
4. Collaboration: seek the parent's ideas, explain rationale, ask whether the plan is acceptable and what obstacles exist, compromise, acknowledge and apologise for mistakes, adapt goals to the family's culture.
5. Avoid: interrupting, jumping to solutions, jargon, authoritarian stance, defensiveness, appearing distracted.
Example language (illustrative): "It sounds like mornings leave you exhausted before the day's even started. What part of the morning would you most want to change first?"
Watch for: Taylor et al. note the training outline in their Table 5 has not been empirically evaluated. [source: Taylor et al. 2019]

D3. Treatment integrity monitoring

Purpose / when: throughout, especially when results stall or a new implementer joins. [source: Hanley 2018 notebook; BACB 2.18]
Steps: list the critical steps of each procedure; observe sessions and score each step as done correctly or not; retrain (D1) on steps below criterion. [synthesis for the scoring method]
Watch for: low integrity looks like treatment failure; check integrity before changing the plan. [synthesis]

E. Historical procedures (described, not taught)

Lovaas-model EIBI and discrete trial training (UCLA Young Autism Project, from 1970). Intensive (up to about 40 hours a week), mostly one-to-one, table-based discrete trials delivered at home over years by trained therapists and parents, working through curricula from compliance and imitation to language and pre-academic skills. Early versions prioritised compliance and reduction of autistic behaviours such as stimming. [source: UO Autism History Project; Wikipedia ABA; compliance emphasis per Wikipedia critiques]
Aversives. Lovaas's early UCLA work used slaps, scolding, withholding of food, and electric shock (including a hand-held device) as a stated last resort for severe self-injury and also to suppress other behaviours; Lovaas (1987) reported that contingent aversives formed part of the intensive programme [uncertain: 1987 paper]. Electric skin shock continues at the Judge Rotenberg Center in the US; the UN Special Rapporteur on torture has called this practice torture; ABAI issued a 2022 policy statement condemning contingent electric skin shock, though a 2023 ABAI task force report described such methods as not inherently unethical in limited circumstances. [source: UO Autism History Project; Wikipedia ABA] No steps are given here.
1974 Rekers & Lovaas study. Behavioural treatment aimed at suppressing gender-nonconforming behaviour in a young child; JABA issued an expression of concern in 2020 and ABAI condemned conversion therapy in 2021. [source: Wikipedia ABA]

5. Handling problems

The controversy (neutral summary). Many autistic adults and advocacy organisations, including the Autistic Self Advocacy Network and the European Council of Autistic People, oppose ABA as an autism intervention. They report that it suppressed autistic traits, encouraged masking (linked to poor mental health), and was experienced as distressing or traumatic; a 2020 study of autistic adults who received ABA as children described many accounts of this kind. Critics also point to conflicts of interest in the research literature and to an ethics code they see as lenient on aversives and reliant on parental rather than client consent. Defenders, such as Leaf and colleagues, argue that the evidence does not show that all ABA is abusive, and some practitioners have reformed their methods in response, including the assent-based and trauma-informed approaches this guide follows. [source: Wikipedia ABA]

Stuck learner / no progress on skills.

Check whether HRE is present before teaching; if not, go back to A3. [source: UNL HRE handout]
Check treatment integrity (D3) and medical factors (A2). [source: BACB 2.12, 2.18]
Re-run preference assessment; break steps down; change materials, people or location. [source: Beadle & Tirella 2023]
Review whether the goal is meaningful to the learner; if not, change the goal. [synthesis]

Assent withdrawal (what older manuals called "resistance" or non-compliance). Treat it as feedback that the teaching conditions are aversive. Remove the instruction, re-establish motivation (pairing, preference assessment), retry later; if assent is still not given, change the instruction and return to the original goal later; honour the learner's refusal however it is communicated; graph withdrawal with skill data. [source: Beadle & Tirella 2023 citing Breaux & Smith; Fabrizio 2005]

Overwhelm, escalation or severe problem behaviour in session.

Empathetic statement; stop demands; give what the learner likely needs without requiring a request; adjust the environment (light, sound, space). [source: UNL HRE handout]
During IISCA: deliver reinforcers noncontingently; revise the test so it targets precursors; lengthen and vary reinforcement intervals. [source: Hanley 2018 notebook]
Design open-door settings that avoid physical management. [source: ACT Community Hanley page] Where local law or a service's crisis plan permits emergency restraint to prevent serious injury, it is an emergency measure, not treatment; document, debrief and revise the plan. [synthesis]
Severe self-injury: arrange urgent medical assessment of injuries and possible pain sources; involve the team and, where required, a human rights or behaviour review committee. [source: BACB 2.12, 2.15; synthesis for the sequence]

Resurgence after thinning or a life change. Return to the last stable step of the SBT sequence and thin more slowly; check whether reinforcers still work. [source: resurgence named on ACT Community Hanley page; response is synthesis]

Caregiver non-adherence. Ask what gets in the way; simplify the procedure; rehearse in role play and in vivo with feedback; reduce the number of procedures; check acceptability and compromise. [source: Taylor et al. 2019; synthesis for simplification]

Ruptures with caregivers. Name the problem, apologise for your part, listen without defending, renegotiate goals together. [source: Taylor et al. 2019 module on assessing and repairing relationships]

Ruptures with the learner (e.g. after an episode involving restriction, or after pushing past withdrawal). Return to non-demand HRE time until approach behaviour returns; reduce demands below previous levels; involve a familiar person. [synthesis drawing on UNL HRE handout and Rajaraman et al. 2022]

Pressure from funders or schools for more hours or compliance targets. Base hours and goals on the learner's assent and goals; document the rationale; cite BACB 2.01 and 2.14 (client rights first, positive reinforcement, stakeholder preference). [synthesis]

Requests to reduce harmless autistic behaviours (e.g. stimming, lack of eye contact). Explore with the family what harm is feared; target only behaviour that causes injury or real barriers, with the learner's agreement where possible; offer alternatives that meet the same need. [synthesis, prompted by autistic critiques in Wikipedia ABA]

Adolescent or adult learners, or suspected prior trauma from services. Involve the person directly in all goal-setting and decisions; give maximal control over session structure; consider referral for trauma-focused psychological support adapted for autistic clients. [synthesis; Rajaraman et al. 2022 for trauma-informed framing]

Co-occurring suicidal thoughts or self-harm in verbal adolescents and adults. Use standard suicide risk assessment and safety planning adapted for autistic communication, including lethal-means counselling on safe storage and removal; do not treat suicidal statements as behaviour to extinguish. [synthesis; not addressed in the ABA sources]

Sources

Sources:

Books and papers cited:

Baer, D. M., Wolf, M. M., & Risley, T. R. (1968). Some current dimensions of applied behavior analysis. JABA, 1, 91-97.
Lovaas, O. I. (1987). Behavioral treatment and normal educational and intellectual functioning in young autistic children. J Consult Clin Psychol, 55, 3-9. [uncertain]
Hanley, G. P. (2018). Practical Functional Assessment and Treatment Notebook (rev.).
Hanley, G. P. (2020). A perspective on today's ABA. [cited via UNL handout; not drawn on]
Gover, H. C., Staubitz, J. E., & Juárez, A. P. (2022). Revisiting reinforcement: A focus on happy, relaxed, and engaged students. TEACHING Exceptional Children, 55(1), 72-74. [uncertain]
Rajaraman, A., Austin, J., Gover, H., Cammilleri, A., Donnelly, D., & Hanley, G. (2022). Toward trauma-informed applications of behavior analysis. JABA, 55(1), 40-61.
Taylor, B. A., LeBlanc, L. A., & Nosik, M. R. (2019). Compassionate care in behavior analytic treatment. Behavior Analysis in Practice, 12(3), 654-666.
Breaux, C. A., & Smith, K. (2023). Assent in applied behaviour analysis and positive behaviour support. [uncertain]
Morris, C., Detrick, J. J., & Peterson, S. M. (2021). Participant assent in behavior analytic research. JABA, 54, 1300-1316. [uncertain]
Fabrizio, M. A. (2005). Voting with their feet: The role of assent in behavior analytic intervention. [uncertain]
Bannerman, D. J., et al. (1990). Balancing the right to habilitation with the right to personal liberties. JABA, 23(1), 79-89. [uncertain]
Smith, T. (2001). Discrete trial training in the treatment of autism. Focus on Autism and Other Developmental Disabilities, 16(2), 86-92. [uncertain]
Carr, E. G., & Durand, V. M. (1985). Reducing behavior problems through functional communication training. JABA, 18, 111-126. [uncertain]
Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). Applied Behavior Analysis (3rd ed.). [uncertain]
Behavior Analyst Certification Board (2020). Ethics Code for Behavior Analysts.

Not found: the elements of Hanley's Universal Protocol; session counts for SBT; definitions of each CAB class. [uncertain]

← Previous · II-021Anxiety management training Next · II-023 →Positive behaviour support