Your School Has a Fat Tail
Dozens of ways to apply research so students make great progress. Just click here.
This post is about behaviour and inclusion and my favourite topics - data and research.
We’re going to use it to find your fat tail, and then we’re going to slim it down with lifestyle changes or Ozempic - you get to decide.
Find the Fat Tail*
People who like data are fascinated by graphs with a fat tail.
Imagine plotting the behaviour points of your students on a graph. Most students would have none, or hardly any. But there will be a fat tail. Students who have a huge number, 10, 20, 50 times higher than the average.
When you find these students, you need to be inclusive. This is a moral position, not just the new Ofsted buzzword.
But what does inclusive look like?
Firstly, these students must conform to your normal behaviour policy. But as a part of that, you also need to diagnose the student’s trigger points: (do not call it a need - this will drive you towards making excuses - you need to find solutions).
Is their trigger that they feel unsafe? Or that they are easily angered and don’t know how to deal with these emotions? Or that they have fallen so far behind that they are convinced that they will fail? Or do they have a problem processing information, and lessons are just too fast paced? Do they have few boundaries at home? You know what this list looks like - it can be endless.
Not so the solutions. These need to be finite. First, it is taxpayers’ money. Secondly, and more importantly, it is the other students’ money. They are subsidising any intervention. So, it is reasonable to cost your interventions. And then ask yourself these questions:
What is the maximum spend on interventions we can justify for one child?
What are the maximum number of interventions we can run (given the cost in hours of colleagues’ time?)
Once you start budgeting your interventions in this way, you are far more likely to measure their impact. Are they delivering value?
If they are, and the students is still not altering their behaviour, you have no choice but to permanently exclude. You should be brave about this. The average grades a student achieves in AP or PRU is numerically slightly worse than that student might achieve in your school.
But it is not a real difference. Gaining three grade 2s in an AP is not meaningfully worse than foregoing the one grade 3, two 2s and one 1 they might have gained in your school. Bankrupt is bankrupt, whether your bank account has £5 in it or £10.
So cost your interventions - run them, evaluate them, and then, when the money or the time runs out, PEX.
Find the Network Spillover*
This is where your data is going to pay real dividends. Map the networks of your students with the worst behaviour records.
You will see a statistical relationship. If a student does not have one of these worst offenders in their network, their chances of getting a behaviour point is very, very low. But it goes up dramatically, based on whether you have 1, 2, 3 or 4 such students in your network. This is network spillover.
When you map the network you can see who the super-spreaders of bad behaviour are.
A student with a SEND need of anything but SEMH is highly unlikely to be a superspreader. Superspreaders need a social network. Their influence grows exponentially.
They are a virus spreading through your school. I don’t mean to dehumanise them. I mean the very opposite. Like a virus is transmitted through human interaction, so the impact of that superspreader depends on them behaving in very human ways - interacting.
When you find the super-spreaders you have two options.
Lifestyle Changes
Remove them from their social networks. Run their interventions at lunch or after school, or both.
But invest in them, seek every positive and communicate it to parents and to the student.
Have a significant adult check in with them several times a day - spot early warning signs, praise where you can, anticipate and correct where you have to.
Ozempic
Remove them from their social networks. Run their interventions at lunch or after school, or both. Make the cost of misbehaviour as high as you can.
Go to the places they misbehave - classes, outside spaces - and catch them at it or resisting it. Prove that their behaviour is changing, either for the better or worse.
Tell parents that any meeting has to be between the hours of 9 and 4.
Show them your graduated response, with the costings of interventions next to it, so that parents can see the graduated journey students are on, and the cost to the school in resources of those interventions.
Print out the parent/school agreement and show them where they are not complying and where the school is.
Prove to the parent and student that they are unlikely to meet the behaviour standards that 95% of students have no problem meeting.
Let them draw their own conclusion about their next steps. It will probably stop you having to PEX.
*These terms were introduced to me by research into criminal behaviour and policing. The podcast they came from is a brilliant examination of the murder of George Floyd. Your host is Malcolm Gladwell, who never fails to help me see the world in a slightly different way.
How to Decide
How many superspreaders can you justify accommodating in your year group?
Look back at the last 3 years of GCSE results. Take the lowest performing 30% of year 11s in that cohort. Then map them for behaviour points and networks.
Now you will see how much your super-spreaders are costing other students in your school. You can simply compare the P8 of students in the network with those who are not in the network.
What minus figure can you morally accept? A third of a grade in each subject? Half a grade in each subject? A whole grade in each subject? Two grades?
Some superspreaders are only spreading a cold - that’s fine. But some will be spreading Covid in a nursing home. That’s not.
Find a number which is unacceptable to you, that fits the moral values of you and your team. Use the Ozempic solution for any superspreader who is at that threshold. For the rest, you have the lifestyle change option.