Sometimes, the SEMH support strategies we're using with a student just stop working - for no apparent reason - and we're left with the dilemma, "What do we do next?"
In this episode, I share how this simple concept from statistics might explain both why a pupil's behaviour support strategies might have become ineffective - and what to do next.
Get our FREE SEN Behaviour Handbook: https://beaconschoolsupport.co.uk/SEN-handbook.php
Join our Inner Circle membership programme: https://beaconschoolsupport.co.uk/inner_circle.php
Join our FREE Classroom Management and Student Behaviour FB Group: https://www.facebook.com/groups/school.behaviour
Share this podcast with your friends:
Show notes / transcription
Simon Currigan 0:00
Hi there. My name is Simon Currigan. And welcome to school behaviour secrets. This is the third quickfire strategy episode I've got for you, where I give you one simple idea or strategy to think about or news with your new class all condensed into an easy to listen to five to 10 minute episode, we'll get back to our normal format in September. But if you've been enjoying these episodes, or you're hating them, and you're listening to these and thinking, has someone lost his mind, what is he doing? Let me know on social media look for us on Facebook as beacon school support and on Twitter. We're at Beacon support and tell me what you think. And by the way, if you're listening to us in another part of the world where you've got a different school year, then these tips are going to be perfect for you to use for the students that you're already teaching right now. So here's the theme of today's Quick Fire strategy episode it is regression to the mean. Now, don't panic. I'm not about to give you a maths lesson in today's episode. But the concept of regression to the mean, which is taken from statistics actually has a very important applications in our work supporting individual kids with social, emotional, mental health and behavioural needs. And understanding this concept could be the difference between supporting your kids with strategies that are effective and throwing away perfectly good strategies based on poor evidence. I've touched on this before very briefly kind of tangentially in other episodes of the podcast and definitely inside of behaviour tactics video inside our inner circle membership. But I wanted to focus on the idea of regression to the mean, sort of individually today, when we're thinking about supporting kids with behaviour issues, many of us, me included often slip into sort of a mistaken way of thinking that we want the process of supporting the child to be scientific. So we want to look at the behaviours a child's presenting in class, and then look for what might be causing those behaviour issues, things like trauma, an underlying need, like ADHD or something else. And then we go about putting in place strategies to support the child with the underlying need. Because we know that if we're seeing challenging behaviour in class, that behaviour is most likely a symptom and the real cause of the behaviour is an unmet need. So if we address that need, over time, we will see less and less of the inappropriate behaviour in class, right? So far, so good. That all sounds very scientific great. But what this approach does is it reduces children down to programmes or robots, like they're running on very simple logical programming.
If I provide support strategy x, let's say segmenting a child's work in time chunks with learning breaks in the classroom, the result will always be Y, which might be improved engagement with a task, I'm saying, you know, I'm providing strategies and in my head, I'm assuming the child will respond robotically to our support strategies. And often when we introduce a new strategy, often because of its novelty, we see improvements in behaviour quite quickly, which encourages us as the adult to think of the child's behaviour very simply in terms of logical inputs and outputs. And we see the strategy work in the classroom until it doesn't and it stops working. Because humans are not robots. They're complex, emotional and intellectual creatures whose behaviour on a day to day basis is affected by hundreds and 1000s of factors that are completely out of our control. And that means to a sizable extents are people's behaviour is going to look random that on some days, whether they engage with a strategy can appear more down to luck than anything else. And then as professionals, we start asking ourselves, maybe it's because we're using the strategy wrong or we need to change the strategy because something with the student has changed or we just get frustrated because external professionals or senior leaders are telling us Well, science tells us if you do strategy x, it should result in strategy y so maybe you not doing it well or consistently, or it's panic stations because the child was making progress with a behaviour for a few weeks. But now there's been some incidents and the behaviours fallen off a cliff. And we've had serious incidents three days in a row. So something must be done, well, maybe nothing needs to be done maybe to change your strategy that was working previously, but hasn't been effective for a few days on the bounce is premature. This is where regression to the mean comes in regression to the mean tells us about what to expect about how a random system behaves. And we've already talked about how human behaviour has a big random component. It's all those tiny factors that are outside of our control things we thinks we can see and things we can't things we can plan for. And those that we never see coming things like did the child get enough sleep last night? Did they have a fight with their sibling on the way to school? Are they getting the right range of nutrients? Are they lost their phone this morning? Did the child's parents separate last night is there someone Ill close in the family all of these are effects that can make the child's behaviour look and feel random because we don't see those kind of like underlying factors. And it can make it feel like whether they do well in school is down to chance. And here's what the math says about chance and randomness and statistics. It says that the child's behaviour will have a mean an average. And to get this we need to think about their behaviour on an average day over time, say a period of four or five weeks, which is a good long term time to look at what average behaviour for a child looks like in the classroom within that period. Because that behaviour has got a random component to it, you're likely to get two or three very good days of behaviour one after another because that's how random works regression to the mean though means over time, the child's behaviour will get pulled back down to the average. So if you get two or three days very positive behaviour, statistics tells us that you're then likely to get a run of more difficult days to balance those good days out.
It is like an average you add up a series of numbers and the high numbers, the good days balance out the low numbers, the bad days. And with enough numbers, you'll find a true average underneath the average is made up of a range you get good days and bad days and days in the middle. So our pupils behaviour doesn't stick doggedly to in inverted commas. Average behaviour, we're likely to get a majority of average days, some good days and some bad days. And when we get two or three days in a row that are bad, which we should actually expect. If we stop doing the things that have worked in the past have been effective on all the average days and all the good days, then we might be throwing out effective strategies for no good reason at all, just for the sake of panicking because something must be done with throwing out the baby with the bathwater, we're forgetting that kids are robots. Behaviour does have a random, unpredictable, unplanned element. And we should expect good days, average days and bad days. That doesn't mean the strategies and now pass their sell by date. Now if you have 99 difficult days in a row, I'm not saying you should stick to strategies that are clearly not working. What you're seeing there is a trend. But what I am saying is it can feel tempting to give up on good strategies because of a few days of random events. So we have to keep in mind the role of regression to the mean and take a long term view of the child's progress. And when we think about progress, we need to ask what are we measuring, we're measuring the average behaviour in the long term over the weeks, because we want to see the average improving and that will involve short bursts of success and short bursts of difficulties along the way. What we're never going to do is just get average and successful days and completely eliminate the difficult ones not in the short term anyway. And that's it for today's quick fire strategy episode. If you liked today's episode, don't forget to share it with three friends who would find it useful. Simply open up your podcast app, hit the share button and your app will send you a direct link to the episode through text messaging, email, whatever you use. And if you've enjoyed the podcast today, while you've got that podcast app open, remember to hit the subscribe button to make sure that you don't miss future episodes. And don't forget if you work with kids with challenging behaviour, and you're not sure why they're acting the way they are, we've got a download that can help. It's called the Sen. D handbook. And it can help you link behaviours that you're seeing in the classroom with possible causes like autism, ADHD, and trauma. The idea here isn't for us to try and make a diagnosis as teachers because we're not qualified to do that. But we've got an important role in linking behaviours with possible causes and getting the right professionals involved getting the right early intervention strategies involved as quickly as possible and this will help you do that. It's a free download. Go to our website beaconschoolsupport.co.uk. UK and click on the free resources tab near the top to get a copy of your handbook. I'll also put a link in the episode description if you found the content today useful or less give you some food for thought to have a brilliant week and I look forward to seeing you next time. Bye
(This automated transcript may not be 100% accurate.)