Escape, attention, tangibles, sensory. The core concept behind every functional behaviour assessment, explained properly.
This article is written for behaviour support practitioners, support coordinators, and anyone who wants to understand why behaviours of concern happen. It’s part of our series on the NDIS Commission’s new practice guide on behaviour support assessment.
At the end of this article, you will know:
- What “function” actually means and why every behaviour has one
- The four functions of behaviour, with everyday examples of each
- Why “power and control” and a person’s diagnosis are never functions, according to the new practice guide
- How one behaviour can serve multiple functions, and why that changes the strategy
Think about the last time you sent a text you didn’t need to send, made a coffee you didn’t need to drink, or left a conversation early. None of that was random. Every one of those behaviours did something for you. It got you connection, comfort, or a way out of a room you didn’t want to be in.
Behaviour that challenges works exactly the same way.
Every behaviour serves a purpose for the person. The purpose is called the function, and finding it is the entire point of a functional behaviour assessment.
The NDIS Commission’s new practice guide puts function at the centre of the assessment process. Functions of behaviour, in the guide’s words, are typically framed in terms of what the person is trying to get (seek out) or reject (avoid, escape) by using the behaviour. That’s the simplest way to hold the whole concept: get or get away from.
From there, most practitioners work with four broad functions. Let’s go through them.
1. Escape or avoidance.
The behaviour gets the person away from something. A demand, a task, a place, a sound, a social situation, a feeling of pressure.
Here’s what it can look like. A participant is asked to start a non-preferred activity. They push the materials off the table and yell. Staff pause the activity and give them space. The behaviour worked. It ended the demand.
The practice guide uses a similar example in its model of a good hypothesis statement: when asked to complete a non-preferred task, the person engages in physical aggression, staff typically leave the immediate area, and the hypothesised function is escape.
Escape-maintained behaviour is common, and it’s often accidentally strengthened by the people around it. Not because anyone is doing anything wrong. Ending a demand when someone becomes distressed is a natural, caring response. It just also teaches the person that the behaviour is the fastest way out.
If a behaviour reliably ends a demand or removes the person from a situation, escape should be on your list of hypotheses.
2. Attention, or more accurately, social interaction.
The behaviour brings people to the person. Conversation, eye contact, comfort, even a firm telling-off. For someone who is lonely or under-engaged, any interaction can be worth having.
The guide’s second example covers this one. A person engages in self-harm behaviour when left alone for longer than 20 minutes or when staff are supporting others. Staff respond by redirecting them to a preferred activity and engaging them in conversation. The hypothesised function is to obtain social interaction and access to preferred activities.
A quick note on language here. “Attention-seeking” has become a dismissive label, and it deserves better. Humans are wired for connection. If a person’s most reliable way of getting interaction is a behaviour of concern, the problem is not the person. The problem is an environment where connection is scarce.
Behaviour maintained by social interaction is a signal about unmet connection needs, not a character flaw.
3. Tangibles: access to items and activities.
The behaviour gets the person a thing. Food, a device, a preferred activity, a favourite place.
You’ve seen this one in every supermarket in the country. A child asks for a chocolate bar. The parent says no. The child escalates. The parent, exhausted and watched by strangers, hands over the chocolate. The behaviour was more effective than asking.
In disability support settings, the same pattern shows up around locked kitchens, screen time, outings and preferred staff. If a behaviour reliably results in the person being given an item or activity, especially after an initial “no”, access to tangibles is a likely function.
The strategy implications matter here. The answer is rarely “hold the line harder”. It’s usually about predictability, fair access, and teaching the person a reliable way to request what they want that actually gets honoured.
4. Sensory, sometimes called automatic reinforcement.
The behaviour itself feels good, or it relieves something that feels bad. It doesn’t rely on anyone else responding at all.
Rocking, humming, skin picking, hand biting during pain, seeking deep pressure. These behaviours can occur when the person is completely alone, which is often the clue that the function is sensory rather than social.
Two important cautions on this one.
First, sensory behaviour can mask a medical issue. A person who suddenly starts hitting the side of their head may have an untreated ear infection or dental pain. The practice guide is firm on this point: assessment should confirm whether medical factors have been ruled out as causing or contributing to behaviour, including when the person last had a medical, medication and dental review.
Second, not every sensory behaviour needs intervention. The guide is explicit that features of a person’s disability or differences in processing style should never be the target behaviours for change. Stimming that harms no one is not a behaviour of concern. It only becomes a focus of support when it causes harm or seriously limits the person’s life.
Before treating any behaviour as sensory, rule out pain and medical causes first.
What is not a function.
This might be the most practical section of the practice guide, so we’re quoting it close to word for word.
A person’s diagnosis or disability, or “power” and “control”, should not be described as a function of behaviour.
You’ve probably read reports that say things like “the function of the behaviour is to gain control of their environment” or “the behaviour is an expression of feeling overwhelmed due to their diagnosis”. The guide lists both of these as examples of poor practice. They sound clinical, but they explain nothing, and they lead nowhere. You cannot design a strategy around “control”. You can design a strategy around escape from unpredictable demands, which is often what “control” is standing in for.
The same applies to internal states. Anxiety, anger and frustration are real, and they belong in the assessment as setting events and contributing factors. But they are not functions, because they don’t describe what the behaviour achieves for the person.
If a report names a diagnosis, an emotion, or “control” as the function, the functional assessment isn’t finished yet.
One behaviour, many functions. Many behaviours, one function.
Here’s where it gets interesting, and where real practice separates itself from textbook practice.
The practice guide notes that multiple behaviours may serve the same function, and one behaviour may serve multiple functions.
A person might hit out to escape demands in the morning when they’re tired, and hit out to get staff interaction in the evening when they’re bored. Same behaviour, two functions, two different strategies needed. Treat it as one thing and half your plan will fail, and nobody will know why.
Equally, a person might yell, throw objects and drop to the floor, all serving the single function of escaping crowded environments. Three behaviours, one function, one core strategy: change the environments and teach a way to leave safely.
This is why data matters so much. Patterns across time, settings and people are what reveal these distinctions. Gut feel alone won’t get you there.
Final words.
The four functions give everyone around a person a shared language for asking the right question. Not “how do we stop this behaviour?” but “what is this behaviour doing for this person, and how do we meet that need a better way?”
For practitioners, the new practice guide sets a clear standard: hypothesis statements should name the trigger, the behaviour, the maintaining consequence and the function, in observable terms, backed by data. For support coordinators and families, you now know what a real function looks like, and what a lazy one looks like.
If a plan you’re reading names “control” as a function, ask about it. The question alone raises the standard.
We’ll keep working through the practice guide in the coming weeks. If anything here raises questions about a participant you support, reach out any time. We’re happy to help.
From the Insight PBS team to yours 🙂
Resources
Read the practice guide: NDIS Quality and Safeguards Commission, Behaviour Support Assessment, including Functional Behaviour Assessment: Practice Guide (2026)
Read more of our blog articles: https://www.insightpbs.com.au/news/
Refer to Insight Positive Behaviour Support: insightpbs.com.au/referral
Contact us: insightpbs.com.au/contact-us