Extinction Bursts: A Teacher's Guide to the Behaviour Spike

Extinction bursts are a temporary worsening of a behaviour that occurs when the reinforcement that previously maintained it is suddenly removed. Skinner (1953) described this pattern as a normal feature of operant conditioning, not a sign that an intervention has failed. The student who calls out, suddenly ignored after months of being told off, will usually call out louder and more often before the behaviour fades.

This article explains what extinction bursts look like in the classroom, why they happen, how long they last, and what teachers can do to stay the course when the behaviour gets worse before it gets better.

‍

‍

‍

‍

‍

What an Extinction Burst Looks Like

An extinction burst is a short-term increase in the frequency, intensity, or duration of a behaviour after its reinforcer is withdrawn. A pupil who has been getting attention for desk-banging will, when ignored, bang harder, longer, or add new behaviours like shouting or kicking the chair. The pattern is paradoxical to anyone who hasn't met it before: do less, and the behaviour gets worse.

Skinner (1953) showed this pattern in operant conditioning experiments — animals trained to press a lever for food would press faster and harder when the food was withdrawn, before the pressing eventually stopped. Lerman, Iwata and Wallace (1999) replicated the pattern with classroom-aged children: 39% of behaviours under extinction conditions showed a measurable burst within the first three sessions.

Classroom example: Year 4 pupil Aisha calls out the answer to every question. Ms Patel decides not to acknowledge the call-outs and instead praises the next pupil who raises their hand. On day one, Aisha calls out twice as often. By day three, she stands on her chair to be noticed. By day five, the calling-out has dropped below baseline. Without the prior knowledge that this spike was coming, Ms Patel might have abandoned the strategy on day two.

‍

Why Extinction Bursts Happen

The behaviour got the student something they wanted — usually attention, a break from work, or social status. When that something stops arriving, the brain doesn't immediately conclude "this no longer works." Its first hypothesis is "I haven't tried hard enough." So it tries harder.

This is consistent with operant learning theory: behaviours are maintained by their consequences (Skinner, 1953). When the consequence vanishes, the behaviour increases its rate of emission to test whether the contingency has truly changed. Cooper, Heron and Heward (2007) describe this as a "behavioural rebound" and note it is one of the most robust findings in applied behaviour analysis.

There is a second mechanism at work: variable reinforcement. If the teacher previously responded to calling-out 80% of the time, the student has learned that responses are intermittent. Intermittent reinforcement schedules produce the most extinction-resistant behaviours (Ferster and Skinner, 1957). The student "knows" that sometimes you do respond, so they keep trying for longer.

Classroom example: Year 8 pupil Marcus has been disrupting maths for two terms. The deputy head once told off Marcus by name in a corridor incident, and Marcus saw two friends laugh. Since then, Marcus disrupts to recreate that laugh. When the maths teacher and deputy head agree to give Marcus zero verbal acknowledgement during lessons (sanctions are issued silently in writing), the disruption escalates for a week. Then it stops.

‍

How Long an Extinction Burst Lasts

Most classroom extinction bursts last 3 to 5 days when the new contingency is applied consistently across all adults the student encounters (Lerman et al., 1999; Cooper et al., 2007). Inconsistency extends the burst indefinitely — if the student is ignored by their teacher but reinforced by the cover teacher on Wednesday, the schedule of reinforcement has merely become more variable, which strengthens the behaviour rather than weakening it.

The burst's intensity peaks early and the rate of decline accelerates after day three. By day seven, a behaviour that previously occurred 15 times per lesson is typically occurring fewer than three times per lesson. By day fourteen, it is largely gone — though spontaneous recovery (see below) means you should expect occasional return episodes.

A useful rule of thumb: if the burst hasn't peaked by day three or shown any decline by day seven, check whether reinforcement is leaking in from somewhere. Common leaks are teaching assistants who haven't been briefed, parents who reward the behaviour at home with attention, or the student earning peer status from the disruption itself (which the teacher cannot control directly).

Classroom example: SENCO Mrs Hughes is supporting a Year 7 pupil with PDA whose work-refusal behaviour has been reinforced by being sent to the library when he refuses. The new plan is that the library option is removed; instead, the work is left on his desk and he is given a private space at the back of the room. The first week is brutal — desk-flipping, walking out, swearing. By the start of week two, Mrs Hughes is on the verge of reverting. By Friday of week two, the pupil completes his first piece of independent work in five months.

‍

Spontaneous Recovery: The Comeback After a Holiday

Spontaneous recovery is the temporary reappearance of an extinguished behaviour after a delay (Skinner, 1953). For schools, this typically means the behaviour comes back briefly after a weekend, a half-term, or a long absence. It is not a relapse and it does not mean the strategy has failed.

Pavlov (1927) first documented spontaneous recovery in classical conditioning, and Skinner (1953) showed it applies equally to operant behaviours. The mechanism is not fully understood, but the practical implication is clear: hold the new contingency in place, and the recovery extinguishes itself within one or two days. Bouton (2004) reviewed 50 years of research on this pattern and concluded that the original learning is never erased — extinction simply layers a new "context cue" over it. Returning to a familiar context (the classroom after a holiday) reactivates the original learning briefly.

Classroom example: Aisha (from earlier) has not called out for three weeks. After the May half-term, she returns and calls out twice in the first lesson. Ms Patel does nothing differently — she ignores the call-outs and praises hand-raising. By the second lesson back, Aisha has stopped again.

‍

‍

‍

‍

‍

When Extinction Is the Wrong Strategy

Extinction is appropriate for behaviours maintained by attention or escape from low-stakes work. It is not appropriate for behaviours maintained by automatic reinforcement (self-stimulating behaviours), behaviours that pose a safety risk during the burst, or behaviours where the student is in genuine emotional distress.

A pupil who self-harms or harms others cannot be "extinguished" — the burst phase would be dangerous. Iwata, Pace, Cowdery and Miltenberger (1994) showed that combining extinction with differential reinforcement of an alternative behaviour (DRA) is more effective and safer than extinction alone. In classroom terms: don't just stop responding to the unwanted behaviour, also actively notice and reward the behaviour you want.

For students with developmental trauma, demand-avoidance profiles, or anxiety-driven school refusal, the calling-out or work-refusal may be a regulation strategy rather than an attention-seeking strategy. Conners (1997) and more recently Greene (2014) argue that for these students, extinction can intensify dysregulation rather than resolve it. The functional behavioural assessment matters: identify what the behaviour is doing for the student before you decide whether to extinguish it.

‍

How to Run an Extinction Plan in a Classroom

Run the plan in five steps:

1. Identify the function: What is the student getting from the behaviour? Attention? Escape? Peer status? If you can't answer this, you cannot run extinction.
2. Brief everyone: Every adult the student encounters this week needs to know the plan. Cover teachers, lunchtime staff, the SENCO, parents, the head of year. Inconsistency wrecks extinction.
3. Pair extinction with active reinforcement: Decide what behaviour you want to see instead, and notice it loudly when it happens. The student needs a clear new path to the social need.
4. Plan for the burst: Tell yourself, the student's parents, and your colleagues to expect 3 to 5 days of worse behaviour. Pre-decide what you will and will not do during the burst.
5. Hold the line through spontaneous recovery: After half-terms and long absences, expect brief reappearances. Do not interpret these as the strategy failing.

Classroom example: Year 6 teacher Mr Patel runs a four-week extinction plan for a pupil whose lateness was being reinforced by a tour of the classroom and a brief one-to-one chat about why being on time matters. New plan: the pupil enters silently, the lesson continues, and Mr Patel speaks to the pupil at the end of the lesson about progress on the work, not the lateness. Burst peaks at day four (pupil is 12 minutes late and slams the door). By day ten, lateness is at zero.

‍

Common Mistakes That Extend the Burst

The most common mistake is partial reinforcement — the teacher ignores the behaviour during direct instruction but responds during transitions, which converts a continuous reinforcement schedule into a partial one and strengthens the behaviour (Ferster and Skinner, 1957).

The second mistake is missing the alternative behaviour. Extinction without differential reinforcement of an alternative leaves the student with no functional way to meet the social need that the original behaviour was meeting (Iwata et al., 1994). They will either invent a new problem behaviour or escalate the existing one.

The third mistake is mistaking spontaneous recovery for relapse. Teachers who have not been briefed about the recovery pattern see the behaviour return after a holiday and conclude the strategy has failed. They reintroduce the old reinforcement, and the behaviour comes back permanently, now on a much more intermittent and therefore more durable schedule.

The fourth mistake is running extinction on a behaviour the student cannot yet replace. A student with limited language who calls out because they cannot raise their hand quickly enough needs an explicit replacement behaviour (e.g. holding up a coloured card) before extinction is fair.

‍

Limitations and Critiques

The applied behaviour analysis literature on extinction is robust but not without critique. Three limitations are worth noting.

First, the underlying operant model treats classroom behaviour as a series of discrete events shaped by external contingencies. Critics from constructivist and trauma-informed perspectives (Greene, 2014; Bath, 2008) argue that this misses the relational and developmental context — particularly for students with attachment disruption or developmental trauma, where the behaviour may be a survival response rather than a learned habit.

Second, much of the foundational research on extinction comes from clinical settings with one-to-one staffing ratios. Classroom application requires extrapolation, and Lerman et al. (1999) acknowledged that variability of the burst pattern is higher in naturalistic settings. The 3-to-5-day estimate is an average, not a guarantee.

Third, the strategy assumes the teacher can maintain absolute consistency. In a busy primary classroom with three adults rotating through, this is rarely realistic. Hattie and Yates (2014) note that the gap between theoretical behavioural intervention and the lived reality of classroom implementation is one of the most frequently underestimated variables in educational research.

‍

‍

Classroom Strategies to Survive an Extinction Burst infographic for teachers" loading="lazy">

‍

‍

‍

Next Lesson

In your next lesson, identify one behaviour in your classroom that you suspect is being maintained by your own response to it. Write down what the student is getting (attention, escape, peer status). Decide what behaviour you want to see instead. Brief any other adult who works with this student before the day ends. Tell yourself, in writing, that the behaviour will get worse for three to five days before it gets better.

Then start.

Free Resource Pack

Extinction Bursts: Teacher's Guide

3 essential resources to understand and navigate behaviour spikes in the classroom.

Behaviour Management CPD Visual Strategy Card Quick Reference Extinction Bursts Teacher Support

Download Now ▼

Download your free bundle

Fill in your details below and we'll send the resource pack straight to your inbox.

First name

Last name

School email

School name

Your role Select your role Class Teacher SENCo Subject Lead Phase Lead Deputy Head Headteacher Teaching Assistant Other

Current school focus Select focus area Metacognition and self-regulation Literacy and reading Numeracy and maths Behaviour and wellbeing SEND and inclusion Assessment and feedback Curriculum development Staff CPD and training Other / not applicable

Quick survey (helps us create better resources)

How confident are you in identifying and responding to an 'extinction burst' in student behaviour?

Not at all confident

Slightly confident

Moderately confident

Very confident

Extremely confident

To what extent do your colleagues and school leadership understand and apply strategies for managing extinction bursts consistently?

Not at all

Slightly

Moderately

Very

Completely

How frequently do you proactively plan for potential 'extinction bursts' when implementing new behaviour interventions?

Never

Rarely

Sometimes

Often

Always

Download Free Resource Pack

✅

Your resource pack is ready

We've also sent a copy to your email. Check your inbox.

Download Resources

Free resource from structural-learning.com · Evidence-based teaching tools