Back in the day when the DfE had a rainbow in its logo, Peter Blatchford and his colleagues were commissioned to start their longitudinal study into the impact of teaching assistants on pupil outcomes.[1] The Deployment and Impact of Support Staff Project (DISS Project) lasted 5 years, from 2003-2008 and researched in primary, secondary and special schools. It remains the most comprehensive piece of research into teaching assistants[2] in the world to date.

The key finding took everybody by surprise. While TA’s had a positive effect on teachers’ work load, job satisfaction and levels of stress, their deployment made pupil outcomes worse. Researchers were stunned to find out that even after adjusting for SEN, fsm etc. there was a very strong negative correlation between the amount of TA support a pupil received and the progress they made. The more support, the less progress. Clearly, things had to change.

The researchers were clear that the reasons for the poor outcomes originated in managerial decisions about their deployment and preparedness – rather than in some deficiency innate in teaching assistants.   TA’s were overwhelmingly deployed to work either with children who had some kind of special educational need or with children with lower prior attainment. Very often, they acted as a kind of replacement teacher rather than an additional, enhanced provision on top of excellent teaching from the class teacher. What is more, very few TA’s had any meaningful liaison time with the class teacher and generally learnt what was being covered that day along with the children by listening to the teacher. It is not surprising then, that under these circumstances and lacking sustained professional development, TA’s tended to focus upon task completion rather than the actual learning.

Following the report, the Education Endowment Foundation released guidance for schools to alert them to the pitfalls of poor TA deployment and make seven recommendations to maximize the impact of teaching assistants. Three of these (V, V1 and V11) are to do with interventions – the one area where TA’s could be shown to be making a difference. Only last week, the EEF released further evidence that TA’s can and do make a real positive difference when deployed and trained properly to deliver high quality, well researched interventions.

Back in Bethnal Green we already knew that our interventions had positive outcomes for the very reasons given. For example we used several maths interventions from the Every Child counts stable, such as 1stClass@Number[3].

As a staff we had been looking at growth mindset research and that had already challenged us to move away from grouping by ‘ability’ to letting children choose their own level of challenge from a range of differentiated tasks.   There was no point in us promoting a growth mindset message if we then went on to undermined it by corralling children in tacitly fixed-ability groups. We were amazed on what a difference it made and how a sizable proportion of children made stunning progress once liberated to work at higher levels of challenge.

However, we still had teaching assistants sitting next to the statemented children and their lower achieving acolytes. Did the very presence of a teaching assistant give out the sign ‘abandon hope, all ye who sit on this table’ – whatever other positive messages we were promulgating? And were our TA’s focused on task completion? Did our statemented childrem – like their DISS counterparts – mainly interact with TA’s rather than other children? And what could we do to change?

What was surprising was that our teaching assistants agreed with the research. ‘If they sit with us, they become lazy;’ said one TA. ‘They rely on us to do the thinking for them’. What transformed the situation was discovering the follow up to the DISS report – the Effective Deployment of Teaching Assistants Project (EDTA). This action research project looked at ways of enabling schools to use TA’s effectively. As a direct result, the researchers published Maximising the Impact of Teaching Assistants,   a book so good and so accessible we bought a copy for every TA and class teacher and which I highly recommend. It is highly readable and set out to be used directly for professional development purposes. There is also a MITA website to accompany the book – that’s definitely worth a look too.

We did, of course, reflect on deployment and preparation of TA’s, but the key thing we have been concentrating on has been how TA’s can ensure that their presence promotes resilience, self reliance and autonomy in the pupils they support – the entire ‘growth mindset’ table d’hôte as it were.

One of the first things we did was conduct a survey among a subset of our pupils. We surveyed all the children in ks2 with statements, a selection of lower prior attainment children, and a selection of higher prior attainment children. When asked, who helps you learn in school – the ‘higher’ children almost all named their teacher whereas the ‘lower’ and SEN children were more likely to name a TA and not mention the class teacher at all. But the most illuminating finding was when asked to answer ‘true’ or false’ to the question ‘I learn better when a TA is there to help me’, a third responded no (including the SEN and ‘lower children), and explained that it often interrupted their thought processes and was an unwanted distraction. Another third said that TA presence was sometimes helpful and the final third were positive.

We shared these findings, along with highlights from both the DISS and the EDTA research, with TA’s and teachers during an INSET day this January. Prior to the training, we also videoed 3 out of 10 TA’s as a benchmark tool to help us evaluate the impact of our initiative. This video was only seen by the TA themselves and the 3 teachers leading the MITA project group in school. The best thing about the MITA project is what we now call ‘the MITA triangle’. This is a simple visual reminder of a hierarchy of when to intervene – or not intervene – when supporting students. I say ‘not intervene’ advisedly, because a key message from the project is to intervene as little as possible when supporting students. Assume they can do it, observe carefully, and only intervene after sustained careful observation shows you that the child needs some support.

Our teaching assistants found this really challenging. The idea that it was ok to just sit next to a pupil and effectively do nothing but watch for a few minutes was difficult to take on board. Fortunately, I had videoed myself ‘being a TA’ and supporting a year 6 child with his maths. The sight of the headteacher sitting alongside a pupil and just watching him work for a good few minutes was liberating for them. I was, of course, watching very carefully, so when I did intervene, it was because I could see a misconception getting in the way of learning. We talked a lot about the difference between this and spoon feeding a pupil. ‘I definitely don’t spoonfeed’ said one TA. Imagine her surprise when she saw her video…

The triangle outlines 5 possible ways of intervening. At the top of the triangle the pupil is autonomous and self scaffolds their own learning. Nothing to see here – move along.

The next rung down is prompting (or I prefer to call it nudging) a pupil. This might be as gentle as a meaningful look or strategically focused cough. It might be as simple as saying ‘so, what have you got to do now’ and that will be enough to get the ball rolling. If that isn’t sufficient we might then suggest a very generic strategy, leaving as much of the leg work as possible to the pupil. ‘So, is there anything on the board that might remind you of what you are meant to do now?’

The next step down the triangle is to give clues to the pupil. These will be more specific to the learning at hand than prompts. So we might say ‘ would it help you remember if you looked at the success criteria/spelling bank/100 square/periodic table/working wall? Remember this is happening after some quality input from the class teacher. This is not the child encountering information for the first time. Or being re-presented with it after it has become obvious that the first time didn’t work. This is not an intervention situation where the TA is acting as a teacher and imparting information. This is when the pupil is working with information or a process they have just had explained to them during whole class teaching.

If that doesn’t still work then the TA has to do some explicit modeling. This means the original teaching has failed in some way – it might mean it was too hard for the pupil in the first place. Modeling is basically re-teaching the concept or information. Modeling is fine – but the idea is that the pupil gets it first time along with the rest of the class and doesn’t get to have their own private mini lesson with a TA. That way, dependency lies. I am sure we have all come across pupils who realise that they don’t have to listen the first time around because they will be rewarded with special 1:1 time if they switch off when the teacher is talking.

At the very bottom of the triangle is just telling the pupil what to do. Not because telling pupils things is bad per se – but because we tell them instead of expecting them to work for the answer. Obviously this means that the answer has to be something they should be able to work out for themselves. It isn’t something they don’t, in theory, already know. This isn’t about championing discovery learning – this is about expecting a child to use strategies to find information we have already shared. If, for example, we have shown pupils how they can find the atomic number of an element by looking at the periodic table – it is not ok for them to shrug and look hopeless when asked to find the atomic number of sodium when there is a copy of the periodic table readily available to them.

If they’ve got counters in front of them, and worked examples, and – if all else fails – someone to re model the concept to them again, they should be able to re-create a 3×2 array. What shouldn’t happen is for them to be told,’ put three counters in the first line and another three counters in the second line – there – that’s three times two, draw that in your book.’   In this case, the TA might as well have done the work in the book herself and cut out the middle man!

Following the INSET day, we shared the videos with the original 3 ‘benchmark’ TA’s, to see what they thought about how they performed, now that they had received the training. Each TA watched their video alongside one of the 3 teachers leading the project. The reaction was amazing. The very TA who had been adamant she didn’t spoonfeed was running around the staff room afterwards exclaiming ,’oh my God – I do everything – I turn the pages, I give him a pen, I jump straight in before he’s even had a chance to think, I interrupt him when he is talking, I practically tell him the answer – I’m spoon feeding all the time.’ The child she was supporting is in year 5.

Another TA was similarly horrified. ‘I don’t give him any space, I hold the book like his a baby, I turn the pages – my hand are in the way.’ Indeed, the child she supports being a feisty character – there are veritable tussles for who gets to turn the page and hold the book. The child she was supporting was in year 4.

Throughout the video was an endless litany of ‘Good boy! Good boy!’ It was suggested to her that pointing out what successful strategy the child had just used was likely to be much more helpful. In discussion, each TA came up with 3 or 4 things they were going to try to do differently and the teacher summarized the discussion and these targets on one side of A4. After this, we videoed the other 7 TA’s for the first time and went through the same process. Unsurprisingly, because these TA’s had already had the training, they were already using the techniques in their practice. The discussions with the teacher were really useful. Everyone found it really hard to back off and wait at the start of the session. Everyone wanted reassurance they were doing it right. People were a bit confused if they were clueing or modeling at various points. Looking at the footage, it becomes obvious that we all flit up and down the triangle at different points as the child encounters and then overcomes difficulties. It didn’t really matter if we couldn’t quite decide if a particular intervention was a prompt of a clue – what mattered was that we were all thinking hard about how to let the child – or children in the case of a group – do the thinking.

After a couple of weeks, we re-videoed the original three. The results were astonishing. I actually cried watching the footage of one child. Here was a boy who used to do anything to avoid reading, and now here he was, blending those sounds like a pro, going back again and again if it didn’t make sense, showing amazing levels of self motivation and ploughing through the book like an Olympian. You know I never thought we’d ever teach this child to read and he is miles and miles behind – a 1C probably in old money. But he now has the determination and drive to go for it. It’s two and a half years until he hits secondary school and we are going to move heaven and earth to get him as far as we possibly can so he transfers a fluent (ish) reader. I had better make it clear that his failure to read previously was not down to the kind of support he had received previously – there’s a long and complicated story and I’ve named my school so I can’t say more – nor was the new style of support the only factor in his resurrection (no other word will really cover the extent of the transformation).   However, now that his TA was supporting him in a way that took the stabilisers off – he was flying. No more holding the book for him or being the one pointing at the words on the page, no more vacuous praise, lots more praise for using specific strategies of for showing perseverance, lots and lots more waiting and observing. Really insightful intervention just at the right time, building on strategies. It was like was like watching a masterclass in supporting reading. I watched the footage alone on the weekend before I was due to meet with the TA to look at it with her. It was so brilliant I texted her there and then to thank her for making such a difference to this child’s life.

But is wasn’t just her. Both the other TA’s have really changed how they do things. And all three children they work with have taken off. The gap between them and the rest of the class is closing rapidly. By chance, both classes having been doing fractions – notoriously challenging and made even more so by the hike in expectations from the new curriculum. Both children – and they both have statements for learning difficulties – can how add and subtract fractions with ease, and in the case of the older child, including when they have different denominators.   Before they come and take the money away, I had better add that their difficulties, particularly with processing language, remain as severe as ever. But as for maths – extraordinary!

And the main thing the TA’s are doing differently? Waiting. Taking time before jumping in. in fact, not jumping in…gently putting a toe in the water if really necessary.   Reminding pupils of strategies they could use, giving students space to struggle and become a little frustrated without rushing in to soothe and calm.

With the other TA’s we have also seen a sea change –waiting, gentle prompting – expecting independence as the new normal. There’s been a real buzz about this project with TA’s chatting about it a lot on the staff room. When your TA’s are reflecting about ways to improve their teaching and students learning on their tea breaks, you know you are on to a winner.

Particularly when TA’s work with groups, I can see how strong the temptation to go for task completion is and how we need to reinforce that sometimes you just need to model the input again – and then try the other strategies. One group of young children were meant to use objects such as toy dinosaurs to tell a ‘real story’ based on an equation such 3+2 and then draw a pictorial representation of this. The word ‘story’ was a distractor to some in the group who started on elaborate tales – as if it were a literacy lesson – without reference to the maths. Despite the TA reminding the children about how the class teacher had told a real story based on the maths story, the link between the two – obviously the whole point of the exercise, had not been grasped by a couple of children. Instead of modeling this again – or – even better – sharing the work of other children who had been successful – the TA spent most of her energy trying to get the children to draw first 3, then 2. When actually the drawing was an afterthought – a recording of the thinking and not the thinking itself. But the MITA model allowed us to have a great conversation around this and for her to self identify her own desire to please the teacher by getting the task done – whether or not this actually helps the child learn. So a powerful learning experience for her.

The project’s not over yet. The idea is that a second round of videoing is shared in TA triads – sort of lesson study style – but less threatening. That was meant to be this week – but we have fallen behind. The main problem was that videoing in class made it really difficult to capture the sound properly – even when the rest of the class were really trying to work as quietly as possible to facilitate the video, the sounds of chairs scraping and background talking just made the soundtrack too hard to analyse. So we re-videoed with TA’s taking their group or child out just for this lesson. We could invest in a directional mike, I suppose – but for our purposes videoing lessons in a different room worked fine. We had to make sure that the prompts available in class were still present, of course.

The project has also helped class teachers reflect on how they promote independence for all children.  In one notoriously dependent class, the teacher has shared the triangle with all pupils – suggesting that some of them rather like life at the bottom of the triangle. She now gives them feedback about how well they are able to self-scaffold – the expectation being that children who actively listen during whole class time should usually be able to get down to solid independent work without needing further reassurance. Her TA and herself will circulate, reinforcing where things are going well and spotting where children are making mistakes and then prompting, clueing and possibly re-modeling as necessary. But the children need to risk doing something wrong first.

At a recent open day day review – when teachers from 7 different schools came to observe classes and look in books – all present noted the high levels of resilience and autonomy. I am sure we have further to go with this and it will be interesting to see how this impacts on end of year data in the summer. I wholly heartedly recommend the MITA project to you – it’s been instructive, it’s been transformative and it’s been a blast.

[1] The rainbow doesn’t have anything to do with the rest of this post – I just included it because it looks hopeful and I was feeling wistful.

[2] I’m going to use the term TA – teaching assistant for all classed based support staff. I know a variety of other terms are used and some school differentiate between learning support assistants for SEN pupils and TA’s for more general support – but for brevity and clarity I am using the one term – in the same way the researchers themselves do.

[3] Full disclosure: husband works for ECC. Also, to be quite clear, ECC’s teacher led intervention has been subject to published independent study which found to made a very positive impact but its suite of TA led interventions do not yet have published independently verified research –although EFF trails are taking place. Their own databank is substantial and shows positive impact, as this report from Learning Wales shows.