Positive behaviour support

1. Introduction

Version followed: for individual (tertiary) support, the UK consensus definition of Gore et al. (2013) together with NICE NG11 (2015) and the Willis and LaVigna multielement model for plan structure and reactive strategies; for schools, the OSEP Center on PBIS Implementation Blueprint Part 1 (version 18 October 2015), the SWPBIS Tiered Fidelity Inventory v2.1 (2017), Check-In/Check-Out as in Hawken, Crone, Bundock and Horner (2020, 3rd ed.) and Prevent-Teach-Reinforce (Dunlap et al., 2nd ed.). Horner et al. (1990), the founding paper, is not drawn on [uncertain].

Positive behaviour support (PBS) is not a single therapy. It is a values-led, multicomponent framework for understanding why a person's behaviour challenges (typically a person with learning or intellectual disability or autism, but also any child or adult in a school), and for building a personalised, long-term system of support around them. Positive behavioural interventions and supports (PBIS, school-wide PBIS) is the same logic scaled to a whole school as a three-tier prevention system. [source: Gore et al. 2013; PBIS Blueprint 2015]

Core idea of change: behaviour that challenges develops because it serves an important function for the person (getting attention or contact, escaping demands, obtaining items or activities, sensory consequences). Change comes from altering the person's environment and the behaviour of the people around them so that the behaviour is less needed and less effective, while teaching and reinforcing other ways to meet the same need and improving quality of life. Reduced challenging behaviour is treated as a secondary gain; quality of life is the primary aim. Much of the change effort is directed at carers, staff, teachers and systems, not only at the focal person. [source: Gore et al. 2013; Carr et al. 2002]

For a psychotherapist, the usual role is assessor, formulator, plan writer and coach to the people who deliver support day to day, rather than weekly one-to-one therapy. [synthesis]

2. Framework

PBS has two operational forms. Use the one that fits your setting; they share the same assessment-to-plan logic. [source: Gore et al. 2013; PBIS Blueprint 2015]

2a. Individual PBS (services, families, tertiary school support)

Phase 1: Referral, engagement and initial assessment. Goal: describe the behaviour (severity, frequency, duration, impact) from the person (where possible) and from carers; screen physical health, pain, medication side effects, communication, mental health, sensory profile, trauma, recent life changes and the physical and care environment; assess risk. Move on when health causes have been checked and a baseline exists. [source: NICE NG11 1.5.4-1.5.7]
Phase 2: Functional assessment and formulation. Goal: identify triggers, setting events and maintaining consequences across daily routines; form hypotheses supported by direct observation; record baseline frequency and intensity and quality of life; write a formulation of what led to the behaviour and what it achieves. Scale depth to the case: brief tools for recent-onset behaviour; interviews, observation and record review for complex cases. Move on when the team shares one hypothesis that explains when, where and why the behaviour happens. [source: NICE NG11 1.5.6, 1.5.9-1.5.11; Gore et al. 2013]
Phase 3: Plan development. Goal: a written, multicomponent behaviour support plan, devised with stakeholders and internally consistent with the assessment: operational definitions; proactive strategies (quality of life, removing triggers, functionally equivalent alternatives, coping skills); early calming strategies; a smaller set of reactive strategies; named responsibilities, training needs and data to be collected. Move on when the people who will carry it out agree it is doable with their resources. [source: Gore et al. 2013; NICE NG11 1.6.1]
Phase 4: Implementation and coaching. Goal: carers and staff carry out the plan accurately; you train, model, observe and give feedback. Move on when implementation is being checked and is reasonably accurate. [source: Gore et al. 2013; synthesis for the coaching detail]
Phase 5: Monitoring, review and maintenance. Goal: track challenging behaviour, quality-of-life indicators and plan implementation, and revise. NICE suggests review fortnightly for the first two months, then monthly, and sooner if behaviour worsens, restrictive interventions increase or quality of life falls. Gore et al. frame PBS as long term, with outcomes expected over years. [source: NICE NG11 1.6.1; Gore et al. 2013; Carr et al. 2002]

Format and dose: no fixed session count. Work happens through home, residential, day-service or classroom visits, observation periods, team meetings and carer coaching; a typical complex case involves several weeks of assessment followed by months of supported implementation [synthesis; no session count in the primary sources]. Where parent training is used for children under 12, NICE describes manualised groups of 10-15 parents, about 8-12 sessions of 90 minutes. [source: NICE NG11 1.7.2]

2b. School-wide PBIS (three tiers)

Tier 1 (universal; all students, staff and settings). Leadership team with active administrator; a small set of positively stated school-wide expectations, defined and taught; classroom routines consistent with them; a continuum for encouraging expected behaviour and a continuum for discouraging problem behaviour; family partnership; data-based monitoring. Commonly cited as meeting the needs of roughly 80-85% of students. [source: PBIS Blueprint 2015; Wikipedia for the percentage, uncertain]
Tier 2 (targeted; small groups). Everything in Tier 1, plus an intervention team with a coordinator and behavioural expertise, formal screening, more instruction in self-regulation and social skills, more adult supervision, more reinforcement, more antecedent changes such as precorrection, more precision in not rewarding problem behaviour, and academic support. CICO is the most widely used Tier 2 intervention. [source: PBIS Blueprint 2015; Hawken et al. 2020 publisher page]
Tier 3 (intensive; individual). Everything in Tiers 1 and 2, plus a multidisciplinary team, comprehensive function-based assessment (FBA), an individual plan covering prevention, teaching, reinforcement, controlled reduction of the natural rewards for problem behaviour and safety, and wraparound and culturally responsive person-centred planning with the family. [source: PBIS Blueprint 2015]
Implementation stages for a school: exploration and readiness agreement, installation, initial implementation, full implementation, sustainability and scaling; plans run 3-5 years. [source: PBIS Blueprint 2015]
Criteria for moving between tiers: the Blueprint gives no numeric cut-offs and stresses that a continuum is an array of practices, not student placements. Schools use their own screening rules and data (office discipline referrals, screening tools, teacher request); Tier 2 non-response triggers a fuller FBA. Fidelity: a TFI score of 70% per tier is the generally accepted marker of adequate implementation. [source: PBIS Blueprint 2015; TFI v2.1] Local referral thresholds (for example a set number of office referrals) are common practice but vary by school [uncertain].

3. Guiding principles

[source for the first ten: Gore et al. 2013, which states the components are used together, not as a menu; remaining items: Carr et al. 2002 and NICE NG11]

Quality of life first. Practical meaning: every plan includes concrete changes to activities, relationships, choice and inclusion, and you measure them alongside behaviour.
Constructional, not aversive. Build skills and opportunities; do not plan punishment or restrictive practices as treatment.
Stakeholder participation. The person, family and staff help define the problem, design the plan and judge whether outcomes matter; their own quality of life also counts.
Behaviour serves a function. Ask what the behaviour achieves for the person before choosing any strategy.
ABA as the primary method. Assessment and intervention rest on antecedent-behaviour-consequence analysis and reinforcement principles.
Complementary approaches as secondary. Add other evidence-based methods (health treatment, communication supports, CBT for anger, sensory assessment) where the formulation calls for them.
Data-driven decisions at every stage. Collect baseline and ongoing data and change the plan when data, not impressions, say so.
Function-based intervention. Every proactive strategy should link to a hypothesised trigger or function.
Multicomponent: proactive plus reactive. Most of the plan is prevention and teaching; a smaller part covers safe responses.
Long-term implementation support. Plan for coaching, monitoring and review over years, not a one-off report.
Lifespan perspective. Expect gradual change and plan for transitions. [source: Carr et al. 2002]
Ecological validity. Design for the real setting and the real support people, not ideal conditions. [source: Carr et al. 2002]
Social validity. A strategy must be practical, acceptable and a good fit for those who use it; if staff will not do it, redesign it. [source: Carr et al. 2002]
Systems change. Target problem contexts (staffing, routines, organisational culture), not only the individual. [source: Carr et al. 2002]
Prevention emphasis. Do most work when the behaviour is not happening. [source: Carr et al. 2002]
Dignity. Avoid strategies community members would judge dehumanising or degrading. [source: Carr et al. 2002 citing Horner et al. 1990]
Least restrictive response. Reactive and restrictive measures are graded, last resort, legal, proportionate and paired with a reduction plan. [source: NICE NG11 1.9.1-1.9.5]
Rule out health causes. Pain, illness, sleep and medication effects are checked before or alongside behavioural work. [source: NICE NG11 1.5.5, 1.10.1]

4. Intervention inventory

Organised as: A. assessment and formulation; B. the plan; C. proactive strategies; D. early-warning and reactive strategies; E. implementation, monitoring and family support; F. school-wide tiers. Step-by-step detail is drawn from the cited sources where they give it; where they name a component without procedure, the steps are standard behaviour-analytic practice marked [synthesis]. All example language is illustrative.

A. Assessment and formulation

A1. Comprehensive initial assessment

Purpose / when: At referral, and again whenever behaviour changes suddenly. Rules out health, pain, communication and environmental causes before behavioural formulation. [source: NICE NG11 1.5.4-1.5.5]
Steps: (1) Get a behaviour description (what it looks like, how often, how long, how intense, impact) from the person and at least two people who know them. (2) Check physical health: pain (dental, gastrointestinal, menstrual, ear, constipation), sleep, epilepsy, medication and side effects; request a GP or annual health check if not recent. (3) Assess communication: how the person says yes, no, stop, help, more. (4) Screen mental health, developmental history including autism, trauma or abuse, recent losses or moves. (5) Note the sensory profile and the physical environment (heat, light, noise, smell, crowding) and the care environment (staffing, routines, predictability). (6) Record past interventions and how the person responded. (7) Consider a rating scale for baseline and a screening tool for function. [source: NICE NG11 1.5.4-1.5.5; examples of pain sites are synthesis]
Example language (illustrative): To a carer: "Before we look at what the behaviour gets him, I want to be sure nothing is hurting. When did he last see a dentist?"
Watch for: Sudden onset or change in pattern, new self-injury to one body area, behaviour worse at night or after meals: prioritise medical review. Working when a health cause has been excluded or treated and you have a usable baseline.

A2. Risk assessment

Purpose / when: From the start and at every review; risks shape whether reactive measures are needed before proactive work can take effect. [source: NICE NG11 1.5.7, 1.9.1]
Steps: (1) Review suicidal ideation, self-harm and self-injury, harm to others, self-neglect, risk of family or placement breakdown, exploitation, abuse or neglect, and rapid escalation. (2) Follow your service's standard suicide and self-harm risk assessment; if a person has access to dangerous items, agree safe storage or removal with carers. (3) Write the agreed risk management into the plan, including who does what in an emergency. [source: NICE NG11 1.5.7; step 2 synthesis]
Example language (illustrative): "What is the worst thing that has happened during an episode, and what stopped it getting worse?"
Watch for: Carer exhaustion and placement breakdown are risks too. Working when everyone knows the agreed safety steps.

A3. Indirect functional assessment (interviews, questionnaires, records)

Purpose / when: First step of functional assessment; enough on its own for mild or recent-onset behaviour. Generates hypotheses for observation to test. [source: NICE NG11 1.5.11; Carr et al. 2002; CQC brief guide]
Steps: (1) Interview carers and staff (and the person, where able) using a structured functional assessment interview. (2) For each behaviour, ask about setting events (sleep, illness, hunger, staffing changes), immediate triggers (demands, transitions, being alone, denied items, noise), what happens right after (attention, removal of task, item given, nothing), and what the person gets or avoids. (3) Ask what the person likes, dislikes and does well, and how they communicate. (4) Use function rating questionnaires as screening aids, without treating their scores as proof. (5) Review incident records, daily notes and any existing charts for patterns by time, place, person and activity. [source: NICE NG11 1.5.9-1.5.11; CQC brief guide] Named tools commonly used (O'Neill et al. Functional Assessment Interview; Motivation Assessment Scale; QABF) are not drawn on [uncertain]; describe, do not reproduce, their items.
Example language (illustrative): "Think of the last three times it happened. What was going on just before? What did people do straight after? Did it get her out of anything, or get her something?"
Watch for: Informants disagreeing, or answers that blame the person ("he does it for no reason"); probe for specific episodes. Working when you can state a testable hypothesis.

A4. Direct observation and recording

Purpose / when: To test hypotheses and establish a baseline; needed for severe, complex or unresponsive behaviour. [source: NICE NG11 1.5.9-1.5.11; Gore et al. 2013]
Steps: (1) Agree an operational definition of each behaviour (observable, countable). (2) Set up continuous incident recording for carers: date, time, location, people present, mood, what happened before, what the person did, how long, who responded and how, what happened after. (3) Use a time-of-day grid to spot patterns across the week [synthesis: commonly called a scatterplot]. (4) Observe yourself across high- and low-rate times and activities. (5) Graph frequency and intensity weekly. [source: CQC brief guide; NICE NG11; steps 3 and 5 synthesis]
Example language (illustrative): To staff: "Write what you saw, not what you think she felt: 'pushed table, shouted no' rather than 'kicked off'."
Watch for: Recording falling off after a week; simplify the form. Working when data show a pattern (for example, most episodes during personal care or after 3 p.m.).

A5. Experimental functional analysis (specialist option)

Purpose / when: When indirect and observational data cannot separate functions and the behaviour is severe. Carr et al. note traditional analogue functional analysis is often impractical in community settings, so PBS usually relies on indirect and observational methods. [source: Carr et al. 2002]
Steps: Only with a behaviour analyst trained in the method, written risk safeguards and consent: brief, controlled conditions that each test one possible function, compared with a control condition. [synthesis; procedure not in the sources]
Watch for: Risk of injury during test conditions; stop criteria must be agreed in advance.

A6. Formulation and hypothesis statement

Purpose / when: Ends assessment; the single shared explanation from which every strategy follows. [source: NICE NG11 1.5.6; Gore et al. 2013]
Steps: (1) For each behaviour, write: in [setting events], when [trigger], the person does [behaviour], which results in [consequence], so the function is [get/avoid what]. (2) Add why this behaviour, rather than another, serves the function (missing skill, it works faster, nothing else gets a response). (3) Add broader factors: health, communication, history, quality-of-life gaps. (4) Share it with stakeholders in plain language and revise until they recognise it. (5) Re-evaluate after further assessment. [source: Gore et al. 2013 for the two questions; NICE NG11 1.5.12; template wording synthesis]
Example language (illustrative): "When Sam has slept badly and is asked to get dressed quickly, he bites his hand; staff then stop the task. We think the biting gets him out of rushed demands he cannot yet say 'wait' to."
Watch for: Multiple functions for one behaviour; different functions in different settings. Working when carers can predict episodes from the formulation.

A7. Quality-of-life and person-centred assessment

Purpose / when: Alongside functional assessment; sets the primary goals of the plan. [source: Gore et al. 2013; NICE NG11 1.5.9; PBIS Blueprint 2015 (person-centred planning at Tier 3)]
Steps: (1) Map a typical week: activities, places, relationships, choices, time spent waiting or with nothing to do. (2) Identify what the person enjoys, values and would like more of (by asking, observing and testing preferences). (3) Ask family about hopes for the person. (4) Pick a few measurable quality-of-life targets (community outings, contact with friends, choices offered daily). [source: Gore et al. 2013; Carr et al. 2002; steps synthesis]
Example language (illustrative): "What does a good day look like for Aisha? How many of those has she had this month?"
Watch for: Goals that are only about less behaviour. Working when the plan names concrete life changes.

B. The behaviour support plan

B1. Writing the multicomponent plan

Purpose / when: After formulation; it is the operational document carers use daily. [source: Gore et al. 2013; NICE NG11 1.6.1]
Steps: (1) Title the behaviours with operational definitions. (2) Summarise the formulation in two or three sentences. (3) Proactive section: quality-of-life changes; changes to remove or soften triggers and setting events; replacement skills that meet the same function; coping and tolerance skills; reinforcement arrangements. (4) Early-warning section: the person's signs of rising distress and what to do at each sign. (5) Reactive section: the graded, least restrictive responses to the behaviour itself, and safety arrangements. (6) Post-incident: recovery, debrief and recording. (7) Implementation: who does what, training needs, coordinator, resources that can realistically be sustained. (8) Data to collect and review dates. Keep it short enough to be read; add a one-page summary. [source: Gore et al. 2013; NICE NG11 1.6.1; Willis and LaVigna 2004; one-page summary is synthesis]
Example language (illustrative): To the team: "Most of this plan is about what we do when things are going well. The red section is short on purpose."
Watch for: Plans that list reactive strategies only, contradict the formulation, or require resources the service does not have. Working when staff can tell you the three most important proactive strategies without looking.

C. Proactive strategies

C1. Ecological and environmental changes (antecedent and setting-event modification)

Purpose / when: First-line for all functions; reduces how often triggers occur. [source: Willis and LaVigna 2004 (ecological manipulations); Gore et al. 2013; NICE NG11 1.6.1]
Steps: (1) From the formulation, list triggers and setting events. (2) For each, choose a change: alter the setting, the people involved, how interactions happen, how tasks are presented, noise and crowding, pace and predictability. (3) Address setting events directly (treat pain, protect sleep, offer food before demanding tasks). (4) Use visual schedules, warnings before transitions, and choices within tasks [synthesis: common examples]. (5) Fade changes later only if needed and only gradually.
Example language (illustrative): "Give a two-minute warning and show the picture of the next activity before you ask him to stop the tablet."
Watch for: Changes that simply remove the person from all demands or restrict their life; the goal is a better fit, not an empty day. Working when episode rate falls quickly in the targeted situations.

C2. Structured, personalised daily activity and enrichment

Purpose / when: When the person spends long periods unoccupied or the day is unpredictable; also a quality-of-life goal in itself. [source: NICE NG11 1.6.1, 1.7.8; Carr et al. 2002]
Steps: (1) Build a daily plan of meaningful, preferred and valued activities with predictable structure. (2) Support participation rather than doing things for the person [synthesis]. (3) Review its effect on behaviour. [source: NICE NG11 1.7.8]
Example language (illustrative): "Let's plan Tuesday so there is something she chooses every couple of hours."
Watch for: Activity plans that exist on paper only; check what actually happened. Working when engagement time rises.

C3. Teaching functionally equivalent replacement skills

Purpose / when: For every function identified; gives the person a better way to get the same outcome. [source: Gore et al. 2013; NICE NG11 1.6.1; Willis and LaVigna 2004]
Steps: (1) Name the function (for example, escape from task). (2) Choose a replacement that is easier and faster than the behaviour (a break card, sign, word, device button, gesture). (3) Teach it at calm times, prompting and then fading prompts. (4) Make sure the replacement works every time at first, more reliably than the behaviour does. (5) Once established, gradually build tolerance (for example, 'break after two more'). [source: Gore et al. 2013 for the principle; procedure is synthesis from standard functional communication training, which the sources describe only as functionally equivalent alternatives]
Example language (illustrative): To a carer: "Every time she hands you the 'stop' card, stop straight away for now. We will stretch it later."
Watch for: The replacement being ignored by some carers (the old behaviour then returns). Working when the replacement is used spontaneously and the behaviour falls.

C4. Teaching general skills and coping or tolerance skills (positive programming)

Purpose / when: Longer-term; builds communication, domestic, vocational, leisure, social, community and coping skills so challenging behaviour is less needed. [source: Willis and LaVigna 2004; Gore et al. 2013; NICE NG11 1.6.1]
Steps: (1) Pick skills linked to the formulation (waiting, accepting no, asking for help, relaxation, emotional regulation). (2) Teach in small steps at calm times with practice and reinforcement. (3) Practise in the real setting. (4) For adults with anger difficulties, consider adapted CBT of about 15-20 hours. [source: NICE NG11 1.7.6]
Example language (illustrative): "We'll practise waiting with a timer when he is calm, with things he doesn't mind waiting for first."
Watch for: Teaching only during crises. Working when the skill appears in harder situations.

C5. Focused support strategies: differential reinforcement and stimulus control

Purpose / when: For relatively rapid reduction while longer-term skills are built. [source: Willis and LaVigna 2004]
Steps: Choose from: (1) reinforcing the absence of the behaviour over a set interval (DRO); (2) reinforcing lower rates (DRL); (3) reinforcing a specific alternative behaviour (DRA); (4) reinforcing following instructions; (5) stimulus control and antecedent control (bringing behaviour under the control of cues). For any of these: set the interval or criterion from baseline so the person succeeds most of the time, deliver preferred reinforcers promptly, and stretch the criterion gradually. [source: Willis and LaVigna 2004 for the list; setting criteria from baseline is synthesis]
Example language (illustrative): "Start at ten minutes, because baseline shows she usually manages that; each ten minutes without hitting earns a token."
Watch for: Intervals too long (no reinforcement earned), reinforcers losing value. Working when reinforcers are earned most of the time and rate falls.

C6. Changing carer responses so problem behaviour no longer pays off

Purpose / when: When consequences currently maintain the behaviour; the PBIS Blueprint calls it controlled reduction of the natural rewards for problem behaviour. Always paired with C3, because the person needs another route to the same outcome. [source: PBIS Blueprint 2015; NICE NG11 1.7.5 (changes to staff and carer responses)]
Steps: (1) Identify what carers do after the behaviour that delivers the function. (2) Agree a different response that minimises that payoff while keeping everyone safe. (3) At the same time, ensure the replacement skill reliably produces the outcome. (4) Warn carers that the behaviour may briefly increase. [synthesis for steps 2-4]
Example language (illustrative): "If hitting gets the task removed, we'll keep a very small part of the task going, then stop when he uses the break card."
Watch for: Escalation beyond safe limits; never withhold response to genuine pain, distress or risk. Working when episodes shorten after a short temporary rise.

D. Early-warning and reactive strategies

Reactive strategies aim to end an episode as quickly and safely as possible, without escalating it and without injury or damage; they are not expected to reduce future frequency, which is the job of proactive strategies. Their outcome measure is episodic severity (for example, average episode duration or proportion of episodes with injury). [source: Willis and LaVigna 2004] Use the least restrictive option first and involve the person and carers in planning them. [source: NICE NG11 1.9.2] The reactive list below is from Willis and LaVigna's table, which was partly garbled in the copy opened [uncertain on exact list and order]; steps are [synthesis] unless cited.

D1. Early-warning signs and calming strategies (secondary prevention)

Purpose / when: When the person shows early signs of distress, before the behaviour itself. [source: NICE NG11 1.6.1]
Steps: (1) With carers, list the person's early signs (pacing, repeated questions, flushed face, certain vocal sounds). (2) For each, agree calming actions linked to function: reduce demands, offer the replacement skill, offer a preferred activity or quiet space, lower voice and proximity. (3) Write them as a short sequence in the plan. [steps synthesis]
Example language (illustrative): "I can see you're getting wound up. Shall we take a break? Here's your card."
Watch for: Calming offers that always involve high-value rewards may begin to teach escalation; keep offers plain. Working when more episodes stop at the early stage.

D2. Ignoring (planned non-reaction to low-risk behaviour)

Purpose / when: Low-risk behaviour likely maintained by attention, when the person is safe. [source: Willis and LaVigna 2004]
Steps: Keep the person safe from a distance; avoid eye contact or comment on the behaviour; respond warmly as soon as an appropriate behaviour occurs. [synthesis]
Example language (illustrative): None during; afterwards: "Thanks for asking me nicely, let's go."
Watch for: Unsafe to use for injurious behaviour or when the function is pain or distress.

D3. Redirection

Purpose / when: Early or mild episodes; moves attention to an incompatible or preferred activity. [source: Willis and LaVigna 2004]
Steps: Calmly prompt a different activity, ideally one using the replacement skill, without discussing the behaviour. [synthesis]
Example language (illustrative): "Can you help me carry these to the kitchen?"
Watch for: Redirection to highly preferred items right after the behaviour can reinforce it when function is tangible.

D4. Active listening

Purpose / when: Verbal people who are upset or angry. [source: Willis and LaVigna 2004]
Steps: Reflect feelings and content, do not argue or correct, give time. [synthesis]
Example language (illustrative): "You're really angry that the trip was cancelled."
Watch for: Shift to problem solving once arousal drops.

D5. Problem solving

Purpose / when: When the episode is about a concrete need the person can express. [source: Willis and LaVigna 2004]
Steps: Identify the problem, offer or generate options, agree one, act on it. [synthesis]
Example language (illustrative): "The bus is late. We could wait inside or walk to the next stop. Which?"
Watch for: Overly wordy problem solving with people at high arousal.

D6. Contingent instruction

Purpose / when: Short, clear directions that interrupt the behaviour where the person reliably follows simple instructions. [source: Willis and LaVigna 2004]
Steps: Give a brief instruction for an easy, incompatible action; reinforce compliance. [synthesis]
Example language (illustrative): "Hands on your lap, please."
Watch for: Avoid with escape-maintained behaviour if instructions are themselves the trigger.

D7. Contingent relaxation

Purpose / when: For people who have been taught a relaxation routine at calm times. [source: Willis and LaVigna 2004]
Steps: Prompt the practised routine (breathing, counting, a calming object). [synthesis]
Watch for: Unpractised techniques rarely work in crisis; teach first (C4).

D8. Stimulus change

Purpose / when: To interrupt an escalating episode by an abrupt, unexpected and non-threatening change in the situation. [source: Willis and LaVigna 2004]
Steps: Introduce a sudden neutral change (a new person enters, an unrelated request, a change of room or activity), then move to calming. [synthesis]
Watch for: Loses effect with repetition; it is a one-off interruption, not a plan.

D9. Strategic giving way to a demand (sometimes called strategic capitulation)

Purpose / when: In an escalating, high-risk episode, giving the person what they are seeking (stopping the demand, providing the item) may be the safest way to end it. [uncertain: term not confirmed in the opened table; the principle follows from the stated reactive goal of rapid, safe resolution in Willis and LaVigna 2004]
Steps: Agree in advance in the plan when this is allowed; record it so the proactive plan can address why the episode reached this point. [synthesis]
Watch for: Carers fear it rewards behaviour; explain that reactive strategies aim for safety and proactive strategies change frequency.

D10. Geographical and physical containment; restrictive interventions

Purpose / when: Only for serious risk of harm, as a last resort. [source: Willis and LaVigna 2004; NICE NG11 1.9.1-1.9.6]
Steps: (1) Before any planned restrictive intervention, carry out a risk assessment covering physical health, psychological history including abuse, biomechanical risks and sensory sensitivities. (2) Use the least restrictive graded approach first (moving others away, guiding to a safe space) before any physical intervention. (3) Restrictions must follow the relevant legal framework, be in the person's best interests, necessary and proportionate; staff use only accredited physical intervention techniques they have been trained in [the training point is synthesis]. (4) Pair any restrictive intervention with a reduction programme. (5) Document every use and debrief with those involved. (6) Review frequency, safety and continued need regularly. [source: NICE NG11 1.9.3-1.9.6]
Example language (illustrative): "We're going to move everyone else into the garden so you have the room to yourself."
Watch for: Restrictive practice creep (more frequent use, longer durations); rising restrictive use is a trigger for urgent plan review. [source: NICE NG11 1.6.1]

D11. Post-incident recovery and debrief

Purpose / when: After every significant episode. [source: NICE NG11 1.9.3, 1.9.6]
Steps: (1) Give the person time and low demands to recover; check for injury. (2) Record the incident. (3) Debrief the staff or carers involved (what happened, what helped, how they are). (4) Where possible, debrief the person in an accessible way. (5) Feed learning into the plan. [steps 1, 4 synthesis]
Example language (illustrative): To staff: "What was the first sign you noticed, and what would you do differently at that point?"
Watch for: Blame. Working when debriefs produce plan changes.

E. Implementation, monitoring and family support

E1. Training and coaching carers and staff

Purpose / when: From plan sign-off; PBS assumes the people around the person must change their behaviour too, supported by the same reinforcement principles. [source: Gore et al. 2013]
Steps: (1) Explain the formulation in plain language. (2) Demonstrate key strategies. (3) Have carers rehearse with feedback. (4) Observe in the real setting and give specific praise and corrective feedback. (5) Repeat at reviews and for new staff. [source: Gore et al. 2013 for principle; steps synthesis, following the instruct-model-rehearse-feedback sequence of behavioural skills training]
Example language (illustrative): "That was spot on: you gave the warning and showed the picture. Next time try waiting three seconds before repeating."
Watch for: Training given once with no follow-up. Working when observed implementation is accurate.

E2. Fidelity checks and monitoring

Purpose / when: Throughout; data on implementation as well as on behaviour and quality of life. [source: Gore et al. 2013; NICE NG11 1.6.1] In the multielement model this is done by a Periodic Service Review system [uncertain].
Steps: (1) Turn the plan's key strategies into a short checklist. (2) Observe or sample records against it. (3) Graph behaviour frequency, episodic severity, restrictive interventions and quality-of-life indicators together. (4) Review fortnightly for two months, then monthly, and sooner if things worsen. (5) Decide to continue, adjust, fade or reformulate. [source: NICE NG11 1.6.1; Willis and LaVigna 2004; steps 1-2 synthesis]
Watch for: Concluding a plan failed when it was never implemented; check fidelity first.

E3. Parent training programmes and family support

Purpose / when: Parents and carers of children under 12 at risk of developing behaviour that challenges; and for all families, carer support. [source: NICE NG11 1.3.1-1.3.4, 1.7.1-1.7.2]
Steps: Offer an accessible manualised group (10-15 parents, about 8-12 sessions of 90 minutes, with childcare or out-of-hours timing) focused on communication and social skills; link families to carer assessment, short breaks and advocacy. [source: NICE NG11]
Watch for: Families too stretched to attend; offer individual sessions [synthesis].

E4. Sensory assessment before sensory interventions

Purpose / when: When sensory factors appear in the formulation. Do not use sensory interventions such as multisensory rooms before a sensory profile has been assessed. [source: NICE NG11 1.7.7]
Steps: Refer for sensory assessment (usually occupational therapy) and build findings into environmental changes. [synthesis]

E5. Coordinating health and medication review

Purpose / when: NICE places medication as an add-on only after psychological approaches fail within an agreed time, health treatment does not help, or risk is very severe, always combined with other interventions and specialist-led. Your role is to supply behavioural data for those reviews and to support reduction of medication used for behaviour. [source: NICE NG11 1.8.1-1.8.6; your role is synthesis] No dosing is given here.

F. School-wide PBIS

F1. Tier 1: forming the leadership team and using data

Purpose / when: First step of school adoption. [source: PBIS Blueprint 2015; TFI v2.1]
Steps: (1) Form a team with an active administrator and representative staff, with regular meeting routines. (2) Secure a commitment from staff (the Blueprint gives over 80% agreement as an example readiness threshold). (3) Set up a discipline data system recording what, where, when, who and why. (4) Review data at least monthly using the Blueprint's six questions: what problem, which practice, how prepared are staff, how well implemented, how well are students responding, what to adjust. (5) Measure fidelity annually or more often with the TFI. [source: PBIS Blueprint 2015; TFI v2.1]
Example language (illustrative): "Our data show most referrals happen in the corridor between 12 and 1. What can we change there?"
Watch for: Data collected but not used. Working when the team changes practice from data.

F2. Tier 1: defining and teaching school-wide expectations

Purpose / when: The core universal practice. [source: PBIS Blueprint 2015; Horner et al. 2004 as cited on Wikipedia]
Steps: (1) Choose three to five positively stated expectations (for example, be safe, be respectful, be responsible). (2) Define what each looks like in each setting (classroom, corridors, lunch, buses) in a matrix. (3) Teach them explicitly like academic content, with examples, non-examples and practice, at the start of the year and after breaks. (4) Align classroom rules and routines with them. [source: PBIS Blueprint 2015; Wikipedia; step 3 detail synthesis]
Example language (illustrative): "In the lunch queue, being respectful means hands to yourself and an indoor voice. Show me what that looks like."
Watch for: Expectations displayed but never taught. Working when students can state them and referrals fall.

F3. Tier 1: acknowledgement system (encouraging expected behaviour)

Purpose / when: A continuum of procedures for encouraging expected behaviour. [source: PBIS Blueprint 2015]
Steps: (1) Train all staff in specific, frequent praise linked to the expectations. (2) Add a school-wide acknowledgement (tokens, notes home, privileges) that any adult can give. (3) Aim for many more positive than corrective interactions [synthesis; a specific ratio is not given in the sources].
Example language (illustrative): "Thanks for holding the door, that's being respectful."
Watch for: Rewards concentrated on already well-behaved students; check distribution.

F4. Tier 1: continuum for discouraging problem behaviour

Purpose / when: Consistent, instructional responses to rule-breaking. [source: PBIS Blueprint 2015]
Steps: (1) Define minor (staff-managed) versus major (office-managed) behaviours. (2) Agree a graded response sequence that reteaches the expectation. (3) Apply it consistently and record it. [source: PBIS Blueprint 2015; TFI v2.1 (definitions of problem behaviour, discipline policies)]
Example language (illustrative): "That's running. Show me walking to the door. Thanks."
Watch for: Disproportionate referrals for particular groups; review by subgroup [synthesis, reflecting the Blueprint's culturally responsive emphasis].

F5. Tier 1: precorrection and active supervision

Purpose / when: In predictable problem times and places. [source: Wikipedia on primary tier; PBIS Blueprint 2015 names precorrection as an antecedent strategy]
Steps: Before a known difficult transition, remind students of the expectation; in non-classroom areas, staff move, scan and interact positively. [synthesis]
Example language (illustrative): "We're going to the hall. Remember, walking and quiet voices."

F6. Tier 1: family partnership

Purpose / when: Throughout. [source: PBIS Blueprint 2015; TFI v2.1]
Steps: Share the expectations with families, invite input on them, and send positive notes home. [synthesis]

F7. Tier 2: screening and request-for-assistance

Purpose / when: To find students who need more than Tier 1. [source: PBIS Blueprint 2015; TFI v2.1]
Steps: Use agreed decision rules on referral data, screening instruments and teacher or family requests; match the intervention to the student's likely function. [source: TFI v2.1; specific thresholds uncertain]

F8. Tier 2: Check-In/Check-Out (CICO)

Purpose / when: For students with chronic, low-level, non-dangerous problem behaviour not responding to Tier 1; works best when behaviour is maintained by adult attention, less well for escape. [source: Vanderbilt CICO tip sheet; Hawken et al. 2020]
Steps: (1) The team names a consistent check-in adult and a place and time. (2) Morning: the student checks in, receives a daily progress report listing the school-wide expectations, and sets the day's goal. (3) Each lesson: the teacher rates each expectation on a simple scale and gives brief positive or corrective feedback. (4) Afternoon: the student checks out, points are totalled, and a reward is given if the goal is met. (5) Home: a parent reviews and signs the report, returned next day. (6) The team reviews data weekly and decides to continue, modify, fade or stop; fading often moves to self-monitoring. [source: Vanderbilt tip sheet; fading detail uncertain]
Example language (illustrative): "Morning! Yesterday you hit 80%. What's your target today?"
Watch for: Students whose behaviour is escape-maintained may need function-based adaptations or an FBA. Prevent reward hoarding and waning interest with student-chosen rewards and booster training. [source: Vanderbilt tip sheet; Hawken et al. 2020]

F9. Tier 2: small-group social skills, self-regulation and self-management

Purpose / when: Students needing more instruction in social or self-regulation skills. [source: PBIS Blueprint 2015; Wikipedia]
Steps: Group teaching with modelling, role play, practice and feedback, plus self-monitoring of target behaviours, linked to school-wide expectations. [synthesis]

F10. Tier 2: increased supervision, reinforcement and academic support

Purpose / when: As add-ons for students at risk. [source: PBIS Blueprint 2015]
Steps: More adult supervision, more reinforcement opportunities, more precorrection, fewer rewards for problem behaviour, and academic help where work difficulty triggers behaviour. [source: PBIS Blueprint 2015]

F11. Tier 3: individual FBA-based plan with wraparound and person-centred planning

Purpose / when: Students with persistent or serious behaviour who do not respond to Tier 2. [source: PBIS Blueprint 2015; TFI v2.1]
Steps: Follow A1-E2 in a school context: comprehensive FBA, hypothesis statement, plan with prevention, teaching, reinforcement, controlled reduction of rewards for problem behaviour and safety, quality-of-life indicators, and continued access to Tiers 1 and 2. For complex needs across home and community, use wraparound and culturally responsive person-centred planning (the TFI also names RENEW for older students). [source: PBIS Blueprint 2015; TFI v2.1]

F12. Tier 3 option: Prevent-Teach-Reinforce (PTR)

Purpose / when: A manualised, team-based school version of individual PBS. [source: Dunlap et al. PTR 2nd ed. excerpt]
Steps: (1) Teaming and goal setting. (2) Simple data collection. (3) PTR functional assessment organised around prevention, teaching and reinforcement. (4) Choose at least one strategy each from prevent, teach and reinforce, linked to the hypothesis. (5) Progress monitoring and data-based decisions. Strategy menus are in the manual, not reproduced. [source: Dunlap et al. excerpt; step 4 requirement uncertain]
Watch for: PTR may be insufficient when behaviour has medical roots or the home is severely disrupted. [source: Dunlap et al. excerpt]

5. Handling problems

The primary sources address implementation failure, escalation and restrictive practice; most of the rest below is [synthesis] and marked as such.

No change after several weeks (stuck). First check fidelity: was the plan actually carried out? If not, simplify it and coach (E1-E2). If yes, re-run functional assessment: the hypothesis may be wrong, the behaviour may serve more than one function, or a setting event (pain, sleep, mental health) was missed. NICE asks for re-evaluation of the formulation after further assessment and for multidisciplinary specialist input when behaviour is severe, complex or unresponsive. [source: NICE NG11 1.5.8, 1.5.12; Gore et al. 2013 on implementation support; ordering synthesis]
Behaviour worsens when the plan starts. A temporary rise is common when a behaviour stops paying off (C6). Check that the replacement skill reliably gets the same outcome, review safety, and bring the next review forward; NICE says review sooner if behaviour worsens or restrictive interventions increase. [source: NICE NG11 1.6.1; extinction-burst explanation synthesis]
Carers or staff not implementing (non-adherence). Treat it as a behaviour to understand: is the plan too long, unclear, unrealistic for staffing, or at odds with carers' beliefs ("he's just being naughty")? Reduce to a few high-value strategies, model them, praise accurate use, and ask what would make it doable. Gore et al. expect the people around the person to need reinforcement and support to change. [source: Gore et al. 2013 for principle; Carr et al. 2002 on social validity; tactics synthesis]
Disagreement or rupture with the family or team. Return to shared goals (quality of life), share the data, and revise the formulation together; check you have not implied that parents caused the behaviour. If a family rejects a strategy, find one that fits their values and routines. [synthesis, grounded in stakeholder participation and social validity principles: Gore et al. 2013; Carr et al. 2002]
The person objects to the plan or to being 'managed'. Involve the person in the plan in an accessible way, ask what they want more and less of, and where they lack capacity for specific decisions, follow the relevant mental capacity or best-interests framework. [synthesis; NICE NG11 1.9.2 and 1.9.5 on involvement and legal framework]
Overwhelm or crisis risk at home or in placement. Prioritise safety and carer support: short breaks, more staff at high-risk times, advocacy; reactive strategies may come first where risk is serious, with proactive work introduced once stable. [source: NICE NG11 1.3, 1.9.1]
Restrictive practice creeping up. Treat rising restraint, seclusion or as-needed medication as a red flag for plan failure: urgent review, check implementation and staffing, ensure every restrictive intervention has a reduction programme and debrief. [source: NICE NG11 1.6.1, 1.8.4, 1.9.4-1.9.6]
Targeting harmless behaviours such as stimming. A known risk of PBS is labelling self-stimulatory or repetitive behaviour as a target when it is harmless and may help the person regulate (a concern raised by many autistic people). Before targeting any behaviour, ask: does it cause real harm or significantly limit the person's life? If not, do not target it; instead adjust the environment or others' tolerance. Where it is harmful, teach a safer alternative that meets the same sensory or regulatory need rather than suppressing it. [synthesis; consistent with Gore et al. 2013 values and NICE NG11 1.7.7 on sensory assessment]
Behaviour driven by pain, illness or mental health. Behavioural strategies alone will not hold. Arrange medical review and treat coexisting conditions; NICE advises considering this first. [source: NICE NG11 1.8.1, 1.10.1]
Trauma history. Avoid physical interventions and strategies that resemble past abuse; build predictability and choice; consider trauma-informed therapy for the person where suitable. [source: NICE NG11 1.9.3 on psychological history; remainder synthesis]
Staff turnover and plan drift. Build induction on the plan into staffing procedures; keep a one-page summary; schedule fidelity checks after any change of staff. [synthesis; PBIS Blueprint 2015 calls for procedures for selecting, training and coaching new personnel]
School: CICO not working. Check fidelity and whether the student's behaviour is escape-maintained (CICO suits attention-maintained behaviour better); adapt to function, layer extra supports or move to a Tier 3 FBA. [source: Vanderbilt tip sheet; Hawken et al. 2020 chapter list]
School: low staff buy-in or inconsistent Tier 1. Return to readiness (staff commitment), use data to show the problem, start with one high-incident setting, and use the TFI to target weak features. [source: PBIS Blueprint 2015; TFI v2.1; tactic of starting with one setting is synthesis]
School: disproportionate discipline for particular groups. Disaggregate referral data by group, review definitions and responses for bias, and involve families and community in setting expectations. [synthesis; Blueprint emphasises culturally responsive practice]
Internalising problems (anxiety, withdrawal) missed by referral-driven systems. Use screening rather than relying only on referrals; adapted CICO for internalising problems is described in Hawken et al. (2020) but is not drawn on [uncertain].
Your own scope as a psychotherapist. PBS draws primarily on ABA; for complex cases involving experimental functional analysis or physical intervention, work with a qualified behaviour analyst or PBS specialist. In the UK, CQC expects services to have a psychologist or behaviourally trained specialist supervising staff. [source: Gore et al. 2013; CQC brief guide; referral advice synthesis]

Sources

Sources: