The Importance of Success in Assessments

I was privileged to be invited to share some thoughts at Teachers Talk Radio Connect 2026 in Manchester yesterday. When I considered what to present on, I chose to talk about assessment, reflecting on some thoughts from my time out of the classroom at AQA. I called my session 'The Importance of Success in Assessments'.

In May 2025, Impact Ed released 'Mind the Engagement Gap: A National Study of Pupil Engagement in England's Schools'. I'm sure that you've seen the graph as part of whole school CPD this year, but if you've forgotten or you haven't, here is that graph:

I think it's easy to draw a conclusion from the graph - students like being at primary school, and hate secondary school. But there are significant questions around why this might be the case. I don't have answers, but I do have some thoughts.

I think that we massively underestimate the transition between a primary setting and a secondary setting, and don't do nearly enough to support students. A former colleague's son started in Year 7 in September, and in late September/early October, I asked how he was getting on. 'He's lost', he said, 'doesn't know his teachers' names, who he sits next to in lessons, what his timetable is. The change is huge.'

This young man is high-attaining, but his dad was worried about him settling in to his new environment, and now I'm worried for my own child's transition in two years, too.

I think it's understandable. Going from a primary school of 300 students to a secondary school of 1000 must be overwhelming. Day-to-day, they spent the last six years of their education in the same room with the same teacher, developing a strong bond. I've experienced the other side of this when asking Year 7 students what 19 - 11 + 8 is. 16, obviously, but the response when you explain that the answer is actually not 0 is quite similar to 'No. It's not. My primary teacher said so, and I love them!' It's quite the bond that carries over.

Engagement levels are high between Years 3 and 6. They remain relatively stable with slight dips across Key Stage 2, but in Year 7 there is a significant drop. The aforementioned reasons of a lack of relationship with trusted adults, not knowing their classmates, having to juggle equipment expectations with a timetable, and a different approach to conduct and behaviour. There's so much to remember without even considering new content within the curriculum, and there are so many LARGE people around - I speak from experience as a vertically challenged man, some of those Year 11s are huge.

This issue hit home for me in November. I tried to be prepared by informing one of my Year 7 classes of the date of their next assessment - 'Write this in your planners' - but I wasn't prepared for one response...

'OH! WE'RE DOING ANOTHER TEST?! BECAUSE THAT'S ALL WE DO IN HIGH SCHOOL!'

I went home, spoke to my 8-year-old, who was two-to-three months into Year 4, and asked when her last test at school was. Year 3. My year 7 classes had done five tests - a Sparx baseline, CATs testing, GL assessments and two unit tests - in maths alone, none of which this young person had been successful in. My Year 10 class had done three unit assessments already, too. My own year 4 daughter? Not aware that she'd done an assessment in this academic year.

Let's focus on the phrase above - 'none of which this young person had been successful in' - and move on to Year 9, as I don't teach Year 8. I teach a Year 9 class of low-attaining learners - set 8 of 8. They're incredibly successful in lessons, but I face the same battle every day. They reach the top of the stairs, turn right, and put their guard up. 'Eurgh. Maths!' goes through their heads, and it takes 5-10 minutes to get them out of their slump and to start showing them what they can do again. But then the assessment comes, and we're back to square one. Is the dip in Years 8 and 9 down to repeated failure?

Year 10, when pressure ramps up as GCSEs loom in Year 11, doesn't appear to have a larger drop off, but could their engagement actually get much lower?! And as Year 11 looms, engagement increases a little, but not significantly. Is this down to the light at the end of the tunnel? No more school uniform soon? No more state-mandated PE? The idea of a fresh start? Maybe a bit of maturity and responsibility? Again, I'm not armed with the answers. Only some thoughts, and those are mostly around assessment.

So, what's the point of assessment? What are we aiming to do? 

In Year 11, we have terminal assessments and mocks, which are used to give accurate predictions for those high-stakes assessments. In Year 10, we have the same mocks, used to give predictions to leadership teams and colleges. But in Years 7 to 9, what is the purpose? From my experience, any assessments in Years 7 to 9 are mainly about ranking students to identify whether we should move anyone between groups. It's felt aimless for a while, and I think that the withdrawal of KS3 SATs was detrimental to disadvantaged students who perhaps require external validation of their efforts, struggling for intrinsic motivation.

Ultimately, the point of assessment should always be success. I know this is sometimes a challenge, but it should always be the aim. We need to kick-start the success-motivation cycle, manufacturing success for students, motivating them to seek out more successes through hard work, which feeds into more motivation down the line.

Here's where I shared some actionable suggestions:

I teach a Year 11 set 5 of 8. In their Year 10 mocks, they achieved grades of 1s and 2s, with a maximum mark of around 30%. Giving them past papers from September is a waste of a useful resource, and a weekly reminder that they can't do over half of the questions they're being asked to do. That's not a recipe for success. Delay giving Year 11 past papers until they need familiarity with full papers. They can get exam practice from topic-based practice quizzes, which can be found on many websites, or created yourself using Exampro, Exam Wizard or Exam Builder.

For their Year 10 mocks, remove content that they haven't yet encountered. There is no wonder my Year 11s struggled on full papers - they're nowhere near completing the content from the Foundation GCSE specification. The purpose of Year 10 mocks is mostly around familiarity with the processes of exams - revising for each exam, entering in silence, completing a GCSE paper in the sports hall in complete silence, becoming more accustomed to exam conditions, and waiting patiently for all papers to be collected before leaving in silence. Do they need to do each paper for 90 minutes?

Those Year 10s know what they know, and whether they sit a full series of GCSE papers or something amended, they won't know any more at the end of those exams. The average Grade 2 learner scored 24 marks out of 100 on Paper 2 of OCR Foundation GCSE Maths exam in Summer 2025. If we remove the hardest 50 marks (according to performance data), that score comes down to 20, and this means that those learners only scored 4 marks in the last hour (approximately) of that exam. 

When this is a terminal assessment, this is unavoidable, but suppose that we gave Year 10 learners the experience of sitting an exam, for half the time, with half the questions. The same learners scored 40% (20 out of 50) instead of 24% (24 out of 100), and I think that this can only be a good thing for motivation and feeling successful. It's also possible to use the assessment data provided by the exam boards to provide accurate grades on those half papers, so our data wouldn't suffer either.

In Years 7 to 9, we can do the same thing. Engineer assessments so that students are more successful. Stop building assessments that consist of 25 marks of the 'main content' from a unit and 25 marks which would be considered 'challenging content', and accepting 50% as the expected standard. Instead, build assessments which are made up of 20 marks of 'pre-requisite knowledge' (that we expect students to score all of the marks on), 20 marks of 'main content' and 10 marks of 'challenge content'. Higher attainers are still being challenged, but score close to 100% instead of 70%, and the expected standard is now closer to 75%-80% rather than 50%, feeding into motivation and later successes.

We tend to manufacture success for students in lessons - carefully selecting tasks that they'll be successful with - but not so much with assessments. This is a noble thing to do - give students success every day, and the assessment is just a one-off. But students don't see it that way - their perceived importance of an assessment is greater than that of the importance of each lesson. Just consider the behaviour of your worst class when you give them an assessment - a significant improvement on their day-to-day, working in silence. This lesson matters because they want to do well in assessments. 

We should be carefully selecting their assessments to allow them to feel success too, because (and I think this is a quote from a Craig Barton podcast) "it's really hard to have a growth mindset when you keep doing shit in tests".

Ultimately, we should be engineering opportunities for students to feel successful in assessments. When students feel more successful, they have a greater sense of enjoyment and belonging and are more likely to be engaged in their education.

The End of The Lesson

I started this little series of blogs by considering transitions and the effect of these on teacher tiredness, beginning with the start of lessons, so it feels like I should end it by looking at the end of lessons.

Just before the end of a lesson, I'm alerted to the time by the silent alarm on my watch. I have this set to go off 3 minutes before the end of each lesson. 3 minutes might feel a little short, but it adds some urgency, and 5 minutes felt too long.

The first thing I do is quickly identify where we are in our learning episode and make a quick plan for how we continue in our next lesson.

Have learners engaged and been successful in a period of independent practise on the content of our lesson?

Have they engaged in mixed practice?

Can I move on to the 'next' maths using their current level of understanding as a starting point?

These questions, and more, decide where our journey takes us next.

But then there's a practical element to this. We have a transition between two classes to get underway.

We can quickly finish up where we are up to. We can provide answers, share thoughts, or discuss a problem. Then we start the transition.

Lately, I've been considering the best time for retrieval practice, and I think it might be the end of the lesson. Retrieval practice is so important, and if we put it at the beginning of lessons, latecomers miss out. I don't know about you, but in my experience, those kids are the ones who need it the most. Perhaps others have come from break or lunch to your lesson and aren't yet fully focused, but are more likely to be focused at the end of a lesson, having just done 45 minutes of maths.

So now I think I should add retrieval practice to the end of lessons. 

Finish up our activity, and I can go into practical mode. Learners close their books, and I can start collecting them following the route that I intend to give them out next lesson, being effective with time. I can start checking that the classroom and equipment haven't been disrespected and that they're ready for the next class to come into and use.

Whilst I do this, I can put 2-3 questions on the board that we have studied yesterday, last week, last month and three months ago, and I can select them based on what's appropriate at the end of a lesson for the time I plan to have left, as well as taking into account equipment requirements. Sometimes we can do this on whiteboards, others on sticky notes for collection and checking, and others on printed sheets where necessary, but this means I can be efficient with time, having learners do while I get organised between classes, and also ensure that attention is directed at specific topics in the most effective way possible to develop longer-term learning.

This may take longer than 3 minutes, and my alarms will need some adjusting, but this is a simple fix. Just typing this makes me think 5 minutes might be more appropriate, with more activities to complete than I previously did.

Then, after having collected the books for my current class and put them away, collected equipment that we've used and dismissed the class in an orderly manner, I can stand by the door collecting whiteboard pens in a box, so that they can be taken out again by another class of 30 learners. I can get to their books that were collected in following the same route that they're now going to be given out, before completing the register, all while moving between the character who teaches year 7 bottom set and year 10 top set one after the other! Feels like a chaotic 10 minutes - no wonder teachers are tired!

The Lesson

I feel like 'the lesson', or 'the teaching bit' is something that gets lots of attention, and definitely should do, too! The thing is, we can't address the differing needs of each individual mathematical idea in one blog post. Instead, this will serve as a place for some general thoughts around the main body of a lesson, and the things we need in place in order for learners to be successful - both in the lesson, and over time.

The first thing necessary for learning is that learners have a solid understanding of the prerequisites for the idea they're about to encounter. If this doesn't exist, there's no point in teaching the new mathematical idea.

Don't know their square numbers? Don't look at square roots. 

Don't know what factors are? Forget about highest common factor.

Can't plot quadratic graphs? Don't bother with solving quadratic equations.

How can we know that learners hold this knowledge? Well, check for it.

We can check in the lesson by using mini whiteboards to question the whole class, asking questions about the ideas that we expect learners to know already. But what if it isn't there? That's a big pivot to do in the moment, having planned a lesson on the Pythagorean Theorem and finding out that learners struggle with squaring numbers or finding square roots.

We could always check ahead of time? At the end of the lesson before? That way we can prepare for where learners are, not where they're expected to be on a scheme of learning that was designed 10 years ago.

Better still, curricula should adapt to reflect the current strengths of our learners and allow us to teach the 'next' maths based on what they already know well. This data should be available to teachers, so they can be confident about what their class have been successful with, directing them to the 'next' maths to support long-term learning.

Once we're confident that learners are ready to approach the new maths, we can move on, safe in the knowledge that they stand every chance of success in maths over time. This is better than scaffolding to support learners in the moment, as performance is a poor proxy for learning, and long-lasting learning is more likely to develop if learners are making connections between new learning and their existing knowledge.

Performance isn't a bad thing in itself. Learning cannot occur if learners aren't successful with the maths in today's lesson - if they can't do it it now, there's no chance of it sticking around a week, a month or a year later. Performance today means that learners stand a chance of being able to recall it next week, next month or next year.

This is where example-problem pairs come in - show learners what and how to do, and have them mimic the steps with a minimally different question. Once everyone has shown that they can do what's expected, we can give them some independent practice without concern that they'll encounter significant difficulties.

Considering learners have just shown that they can complete the steps necessary to be successful, this is then an opportunity to begin to develop automaticity with the method. Ideally completed independently, in silence, learners are working on questions similar to those they've just been successful with. This means that learners develop confidence in their abilities, feeling more and more successful.

A golf podcast I listened to had a guest on who said something that resonated with me a lot. Speaking about a golf coaching app, he said "If you suck at something, you get a lot of blocked practise".

I don't think that there should ever be a chance of learners struggling with a concept in a lesson. The content should be selected such that it is just beyond their current level of understanding. If this is selected well, learners don't need too much blocked practise to solidify their understanding. However, if a learner's prior knowledge isn't close to that required to bridge effectively to new knowledge, it follows that they'll need a lot of blocked practise to learn the new content. If learners can't confidently identify the factors of a number, they might need 20 questions on finding the highest common factor to develop confidence. However, learners who can confidently find the factors of a given number see finding the highest common factor as a small step, and might only need 5 questions.

Suppose that the content has been selected so that the learners' chances of learning the new content are as great as possible. Example-problem pairs followed by a short period of independent practise should be sufficient to begin to develop automaticity, and spaced practice will make this permanent. The time that we save without giving learners lots of blocked practise can be given across to mixed practise - an opportunity to develop method selection.

We could do this within the same domain, revisiting prerequisites and highlighting links between related concepts, or across domains, mimicking an exam paper. Either way, learners now have the opportunity to engage with retrieval practice - revisiting material that they have been successful with previously, increasing the chances that learning will occur over time.

In a 50-minute/one-hour lesson, time might be scarce, so this might need to be spread out over a couple of lessons, but 100-minute lessons offer the opportunity to develop method selection with a significant portion of each lesson handed over to mixed practise. Rosenshine stated that 80% of each lesson should be spent working with known ideas, and I see mixed practise as a way to engage with content in this way, as well as by engaging with known content before bridging to new knowledge, revealing the unknown maths as an extension of that which is already known.

After writing down the LO, checking prerequisite knowledge, bridging to new information, checking for understanding using example-problem pairs, and engaging in a period of independent study followed by mixed practise, we're likely to be out of time in my upcoming 50-minute lessons. A learning episode is likely to take more than one timetabled lesson in many cases, and that's absolutely fine. Not all maths fits nicely in 50-minute, one-hour or 100-minute blocks.

Just After The Start of The Lesson

My last blog could apply to many subjects, but I think this one will only apply to maths lessons. That said, I could be wrong. Mrs Taylor and Little Miss Taylor do tell me that often!

This could be considered the beginning of lessons, but taking into account my last blog, this is the point when learners are working on a task using a pen they've collected from me, on a whiteboard that's kept on desks. Books have been handed out, and the register has been completed.

So... What happens next?

And what have learners been working on?

There's a balancing act to be had when answering the latter question, it seems. 

To manage potential cover implications and develop consistency between lessons, taking something from a pre-written resource feels like a good way to go, but this supports management of a team and not learning. 

Supposing that learners encounter a question that they haven't yet progressed to, it's a question that's inaccessible to them, but they might be able to guess/perform next time it comes up with some quick instruction. Unfortunately, without giving content the time it needs, we run the risk of embedding misconceptions with the incomplete explanations that come with rushed answers to starters, and learners begin lessons without success.

In my experience the best use of starters is for spaced retrieval, revisiting topics that learners have had success with recently and will need to continue having success with over time for learning to happen. The expectations around the beginning of the lesson in my new environment differ slightly from this, and I suspect that this might be delayed and delivered as an exit ticket.

In keeping with the policy of my new department, learners will complete two questions aimed at improving fluency with core concepts. They'll do this on mini whiteboards whilst I greet, hand out books and complete a register.

At this point, we'll begin the lesson. I think it's important to signify to learners that we're moving on from what we were doing (those starter questions) to a new topic, and specifically, one that needs their full attention.

As an attentional cue, I'll share the lesson's focus. You might call this a learning objective, but it'll just be a title that we'll write in our books along with the date, to signify that we're moving to a different phase of the lesson.

This will also be written on the whiteboard, which I'm planning to write up each morning to avoid rushed (and therefore illegible) handwriting on the board. This is less likely to be an issue with the Learning Objective, but is likely to be an issue with sharing Success Criteria, which is something I'm less familiar with.

I'll be handwriting these on laminated A3 sheets in the mornings for a couple of reasons. The main one is that I will forget in the heat of the lesson, but also because I'll write them more neatly ahead of time, and I can revisit my day ahead of teaching each lesson to get my thoughts in order.

To recap... Learners are welcomed at the door, take a whiteboard pen and settle into the questions on the board to develop fluency with core concepts. When all the kids are in, I'm going to hand the books out and do the register. Then go through the answers to the questions and direct learners to wipe their boards off and open their books. This acts as an attentional cue to learners that we're shifting to new learning, and we will write the LO and date in their books whilst I display a pre-written learning objective and success criteria.

Next: The Lesson

The Start of The Lesson

"There's nothing quite like teacher tired", I recall a colleague saying several years ago. I imagine there's a long line of people ready to take umbridge with this, and whilst I don't have the wide-ranging experience to support this statement, I can say that teaching is very tiring.

Over the last 17 years, I've worked in challenging schools across Leeds, and a couple of years ago left a school where we had 100-minute lessons. I was surprised to find that I was more tired at the end of a 5-lesson day than I was when we had a 3-lesson day. The same amount of breaks, the same amount of time spent teaching, but 'double lessons' bringing together two 50-minute lessons - so, 3 lots of 100 minutes rather than 5 lots of 60 minutes or 6 lots of 50 minutes.

I gave this some thought, and I eventually decided that this was due to a greater number of transitions and the increased number of beginnings and ends of lessons. I don't find the teaching part of teaching tiring. Once the kids are in, settled and getting on, standing at the front directing attention and circulating in periods of independent study doesn't take much energy... But getting them out and getting them in... Eesh!

I am also very aware that on Monday, I start at a school where the timetable is six lots of 50 minutes - the largest number of transitions, with more beginnings and ends than other timetable plans!

To help me with preparing for these transitions, I have a silent alarm on my watch that alerts me when it's 3 minutes before the end of the timetabled lesson. Time to wrap up the activity that we're working on, or for me to stop talking. Then we're putting away equipment, collecting things together like protractors, books and the like, and making sure that there's no mess left around the classroom. An orderly dismissal, and then we're waiting for our next class. Somehow, without fail, they arrive ten seconds after the bell went, and we're at it again. Making sure that all the kids are coming in with the right attitude on display, having their equipment out, and starting on the activity on the board. Books need to be given out, there are likely to be issues with equipment, and the register needs to be done with detentions needing to be added for those who arrived late.

This 10-minute-ish period that typically happens twice with hour-long lessons and thrice with 50-minute lessons doesn't happen with 100-minute lessons.

Lesson 1 (1/2) can be set up before the school day starts properly. Books are out, and that's a job that doesn't need juggling. Equipment is ready, too. At the end of this lesson, we have a break, and the time to dismiss the class without the pressure of a new class waiting at the door. Lesson 2 (3/4) can be set up over break, and we run into lunch, and lesson 3 (5/6) can be set up over lunch with that class dismissed at the end of the day. There are other differences between 50-60m lessons and 100m lessons, but that's not for this blog. I've been considering my routines for the start of lessons as I embark upon another September in a new environment.

I think that the best way is to consider what I want to achieve and how I'm going to achieve it. So...

#1 Learners to enter the room in a calm and focused manner. Stand at the door to welcome each learner, directing them to the task on the board. This will be completed on mini whiteboards. MWBs are to be left on desks by the previous class, and I'm thinking that upon entering, learners take a dry-wipe pen from a box in my hands. This manages the flow of learners into the room and minimises the chances of any outside distractions continuing into the classroom.

#2 Books handed out in an efficient way. This will not involve learners - they're getting on with the task on the board, being impactful with their time. Books will be stored in boxes at the back of the room and given out/collected in the same way each time. I've done this well previously, and handing out books takes at most 20 seconds rather than the 3 minutes of a kid talking to their mates as they go, asking 'Where does _____ sit?' every third book. This is very easy to achieve, making sure that you have a route around the room that you follow, which puts the books in the same order each time.

#3 Register completed without disturbing the peace. I see doing the register in a traditional sense as a huge waste of time, and my go-to is to compare empty seats with the seating plan and make sure that the number of full seats and names on the register match. That said, I also believe that in the first few weeks it's a good way to learn names, and if norms go the wrong way it's a good way to settle a class. For the most part, I'm not interested in calling out names and waiting for a response when they could be thinking, and I could have it over and done with in 9.58 seconds.

In conclusion, I think that's the start of lessons. Stand by the door with a box of whiteboard pens, greeting learners and handing out (working) MWB pens. When everyone is in and settled, move to handing out books which have been collected following a specific route to support faster handing out. Complete the register with knowledge of who is and isn't in, considering books and gaps in the seating plan.

What are learners doing now? What will happen next? That's for the next blog.

Curriculum: A longer read.

The word curriculum is apparently a near-constant part of my life lately.

I tweeted on November 10, 2023, this:

Curriculum... a thread 🧵

Since 2017 I've thought a lot about curriculum. In my sixteen years in the classroom, I've followed schemes of learning based on textbooks, from exam boards and 'for our kids'.

— Dave Taylor (@taylorda01) November 10, 2023

It led to a bit of attention, and ultimately to a guest appearance with Matt Findlay and Femi Adineran on the Beyond Good podcast. Since recording, Jonathan Hall started an event called a Maths Fuddle through Mathsbot, and his first talking point referenced curriculum too.

Tonight we're talking all things curriculum and schemes. Let's kick of with a binary poll...

What’s more important, a good teacher or a good curriculum?#MathsFuddle1

— Jonathan Hall (@StudyMaths) January 12, 2024

This is one the things, and there aren't many of them, where Jonny and I disagree. He's team 'Teacher', and I'm team 'curriculum'.

I want to put my thoughts down in full, so here goes...

Google's define: function returns this for 'define: curriculum'... the subjects comprising a course of study in a school or college. Straight off the bat, I want to clarify what I mean by curriculum, by saying that I mean the content that is intended to be taught to students. I use the term curriculum within the maths department interchangeably with scheme of learning or scheme of work.

In my 16 years in the classroom, I have followed various schemes with various cohorts of children. I've followed textbooks with KS3, and exam-board provided route maps with KS4. I've followed textbooks with KS4 too, as well as bespoke curriculum at KS3 (spoiler alert: it wasn't bespoke, it was self-made). I've followed off-the-shelf curriculums, and I've looked in to others that I haven't followed, and one thing was true every time... It wasn't the curriculum that was the problem, it was the implementation.

The implementation came down to many factors, but all led back to the thought that 'This is costing us money/has been shared by an exam board/is used by many, so it must be what we should be doing'. My main thought here is 'Well, just because everyone else is doing it wrong, doesn't mean we should'.

Take the exam-board-provided example (and I'll consider things outside of my lived experience, but close to). AQA provide route maps for their qualifications, and they're pretty good. They provide a decent journey through the specification (as pretty much all curricula do), but they don't provide the journey that my cohort needs. 

Their one-year resit route map begins with equivalent fractions, decimals and percentages, and this is often question 1/2/3 on a Foundation paper, but if I gave that to a group of students who achieved a grade 3, this is likely to be unnecessary for them to be taught. For those who achieved a grade U or 1,  this is likely to be the right thing. But, what will happen (and this is probably well-meaning, and it makes sense) is that leaders will see this journey, see that it has an AQA badge on it, and say 'The exam board are saying this is what we should do, so we'll do that'. I don't think that they're saying that you should, but they are saying that you could.

The same goes for CIMT/Complete Maths/Lumen/Sparx/White Rose Education, those Collins textbooks sitting on the shelf in the workroom, and those 1990s Maths Frameworking textbooks that have seen more dust than classroom time in the last 15 years. The main issue is that when an age is attached (as a guide, more than anything), children of that age must be expected to work with that content.

But what about the Year 7 child who hasn't grasped the column method for subtraction, the Year 8 child who hasn't mastered the idea of division, the Year 9 who hasn't learnt how to measure an angle, or the Year 10 child who hasn't learnt his times tables?! (These are all children I've worked with in the last week).

Should they be asked to look at Graphical Forms of Sequences, or Sharing Amounts in a Given Ratio, or Trigonometry, or Simultaneous Equations? (Spoiler: absolutely not). Experienced teachers know that they shouldn't, and effective teachers alter their instruction to the needs of their students. But here's what I believe... 

Simultaneous Equations as a topic is likely to be appropriate for about a fifth of the students in Year 10 (based on my experience), so the other four-fifths need something different. The experienced teachers know this, but those without that experience don't. The experienced teachers then teach something else - something that they're expected to decide, and the inexperienced teachers teach simultaneous equations with disastrous outcomes, because that's what the curriculum tells them to do. 

The curriculum should tell teachers what the students in their class need to cover next, not what the top 20% need to cover, and the rest of you should do something else. The journey through maths is well-defined, and differentiating down means that this isn't the journey that students see. They see something that's a bit like it, but easier. Because they're journey isn't well-defined, they find it harder to make progress, because their journey isn't coherent.

The gold standard of a curriculum is one that adapts to the needs of the students. I have experience here, and spent around 400 hours in a past role collating a curriculum that took students on a journey from the Year 3 National Curriculum, to AQA's Level 2 Further Maths qualification. We found that student behaviour improved, enjoyment levels increased, and attainment improved, taking us from 61% 4+ in 2018 to 75.7% 4+ in 2022. 7+ increased from 8.0% to 18.1%. Teaching became a joy and privilege.

So, how did I go about putting this in motion?

I think my first experience of realising that we needed something better was with a Year 10 bottom set, scoring less than 10% on tests, and then laughing about it. This had become so natural to them, that they expected to fail, and then it became a race to the bottom. My solution was to have them do the Entry Level Certificate in lessons, score 80% on each assessment, and development a culture of success, and wanting to do better.

A few months later I was approached by a colleague who'd taught Pythagoras to students. This was the second topic in the textbook. He commented that they were fine with the procedure, but hadn't a clue how to find square roots of square numbers. This was topic 5 in the textbook.

All students were doing the same tests, from bottom sets to top sets, with the top sets scoring 70% plus, and the bottom scoring 20% and below. This was accepted as the norm, and those students would just score that and be happy with it. This is where I targeted my first action - by having multiple assessments that staff could choose from to give their classes, that were comparable between groups, but gave students more success.

I worked, and continue to work, in challenging circumstances. High levels of disadvantage, high proportions of EAL, low levels of aspiration, and prior attainment below the national average. I'd experienced motivation in KS3 fall away since KS3 SATs were abolished, and the same at KS4 when modular qualifications were culled. Whether this is the outcome of external tests that means that students are taught the right maths when 'revising', or the students' motivation to perform on external assessments, I'm not sure, but either way, I wanted to harness this, and took inspiration from both.

I decided that I'd split the year in to 5 modules. The content of each module wasn't overly important, but the expectation that students would improve on their modular assessment from the year before was. Each of the 5 modules was split in to 9 units, where the content was hierarchical, and content is expected to be mastered before moving on.

The five modules were called Module A, B, C, D and E, with units going from 1 to 9. Module A consisted of Place Value, Symmetry and Transformations and Averages. B was Calculations, Angles and Sequences. Module C was Algebraic Manipulation, Charts, Fractions and Shapes. D consisted of Decimals, Equations, Properties of Numbers and Units. Module E consisted of Graphs, Perimeter, Area & Volume, Percentages, Probability and Ratio. I felt this gave time for skills to be taught before using them in later modules, but it's also important to state that content didn't necessarily begin in Unit 1 (Averages began in Unit 4, and Algebraic Manipulation was introduced in Unit 5). The whole journey is here (be warned of large file size).

A recent paragraph stated that 'content is expected to be mastered before moving on'. How do you achieve that? Well, this came down to assessment. Assessments were designed so that each Modular assessment had 9 sections (one for each unit). Each section consisted of 20 marks, so the Module A assessment was 180 marks long, but students only sat 60 marks from this assessment, decided by their teacher. If students had succeeded with content up to Unit 4, they sat the assessment consisting of sections 2, 3 and 4. If students achieved 15 marks in a section, they were able to achieve in the next unit, and have their assessment recorded as this.

A quick recap: A student would work on content from, say, Module C. They had been taught content from Units 3, 4 and 5, for example, so sat a 60-mark assessment on content from sections 3, 4 and 5. They might score 18 marks on the content from Unit 3, 12 marks on the content from Unit 4 and 9 marks on the content from Unit 5. This would mean that they had mastered the content from Unit 3, but not quite from Unit 4, and this is where their instruction would begin the following Year. In terms of recording this assessment grade, this would be recorded as 4.5 on the departmental tracker, indicating that they've had some success with Unit 4, and should work from this point in the next academic year when they return to Module C. Comparatively, a score of 18/16/9 would be recorded as '5.2' where .2 represents scoring 5 marks, .5 is 10 marks, and .8 is 15, which means they 'graduate' up to the next section of their assessment.

From a departmental leadership perspective, the tracker then provided an up-to-date record where students are on the journey in each Module (on each of the sections of 'the dartboard' as we referred to the image shared earlier), as well as a ranking based on the average of their 5 modules (updated with their most recent assessment. This means that sets can be assigned using the average level of proficiency across the 5 units, and tinkered with at the top and the bottom of each set to ensure that the spread of attainment in each group is as small as possible. Assessment data, as well as their set from each year remains visible so that trends can be spotted, but progress is calculated between each module from the current academic year and the previous academic year to celebrate progress made by students (rewarding the lowest attainers too, as they make good progress, rather than always those who go on to score the highest grades). The list of students making the most progress between modules was printed off and displayed proudly, so that those students who might not normally find their name on these lists were there, could tell their parents they were on the wall at school, and their peers can celebrate them too.

From a class teacher perspective, when you moved to another Module on the scheme, you looked at the tracker and saw where their attainment was last year. Acknowledging forgetting, this is where your instruction started, to activate prior knowledge, and then bridge to new content over time, confident that students were going to grasp the content that you were teaching. By the end of the Module, you 'ordered' assessments based on the levels of success that students were experiencing in lessons, and departmental leadership were able to check that staff were being aspirational when ordering their assessments, based on the spread of prior attainment visible from the tracker, and the numbers of each assessment ordered on the order form. 

From a student perspective, they were never out of their depth because they were building on success from the last time with content from that module, so they were motivated to succeed, their behaviour improved and so did their participation in lessons. They were accountable for making progress between modules, and celebrated when they did.

Heading back to the class teacher perspective, as in inexperienced teacher you were able to develop your pedagogy with motivated students, participating in lessons, knowing that if students weren't grasping the idea, that it was down to you, rather than the content. Teachers were also accountable for ensuring progress occurred.

I'll acknowledge some of the challenges associated with this operation, and there were many, but I won't remember all of them. But, where do you start with Year 7 students? Taking into account their prior attainment, start them on unit 1, 2, 3 or 4. It wasn't ideal, because their prior attainment wasn't recorded by us, and was sometimes quite different to what we experienced. But there was an acceptance that even if the level of complexity was a little off in Year 7, by the time we got back around to Module A we had assessment data from Year 7, and getting the level of maths right was much easier.
What about tiering at GCSE? They still did Year 10 mocks, Year 11 mocks, and from that we had grades and expected progress between each assessment point and the end of Year 11. We left tiering decisions until as late as possible, but in my eyes, if you're not going to achieve a grade 6, then you should be doing Foundation.
What about students who made no progress repeatedly? Well, we have their progress from previous years, and they're identified as needing more attention. They are the first students you check in with in periods of independent work, the first mini whiteboards that you check, seated close to the teacher for additional attention.
What if a class hasn't covered the content for the assessment? The class teacher chooses their assessment. A lack of progress is picked up at departmental leadership level if class-wide, and medium term planning can be looked at with the teacher in question.

Whilst I've been writing this, this has been in my notifications:

I think a strong curriculum can have a wider positive impact than a strong teacher but I think the impact of a strong teacher is greater on those that they teach.

A strong teacher will find a way of making the most bonkers curricula work

— Jshm (@jshmtn) January 13, 2024

This is why I think having a good curriculum model in place is more important than having the best teachers. The curriculum's impact is felt beyond the four walls of an individual teacher's classroom, and from a departmental leadership perspective, this is what we should be aiming for.

This blog may be incomplete in places - sorry if that's the case. Feel free to pull at loose ends and ask for clarification. I'm sure that you'll be able to pick holes in my beliefs, and that's fine by me, because I'm happy to be told that I'm wrong. I just happen to believe that I got most of 'how a maths curriculum can work' right, backed up by a significant improvement in outcomes, and would love to see more places adopt a similar strategy.

Assessment Feedback Lessons

With a little time having passed between the end of term and now, and with my mind turning its attention back to work mode, I've had a little time to collect my thoughts about leaving and come back to assessment feedback lessons.

Not every assessment feedback lesson can be ideal, due to time constraints, school events, school holidays, lack of technology, and other issues, but this is where my thinking is at in terms of how I'd like to do assessment feedback moving forward.

1. Assessment is completed.

2. Assessment is marked and scores are entered into a QLA document.

3. 'WAGOLL' provided to pupils.

4. Key points are fed back.

5. Identified areas are revisited.

Each of these steps is briefly outlined below.

1. Pupils sit the assessment. There's an argument that sitting assessments is a poor use of curriculum time, and whilst I see that all classroom time would be best spent teaching pupils content, I also want my assessment data to be as reliable as possible. For me, this is that it's sat in (as close as possible to) exam conditions, in front of me, so I know that I can trust it.

2. Marking is time-consuming, but when it comes to assessments, I think this is an appropriate use of time. I want to see the misconceptions that pupils still hold, I want to be able to feedback on common incorrect answers and the remedies for these, and I want to see the effort that pupils are putting in terms of resilience (answering every question) and presentation (taking pride in their work).

QLAs also seem to split the profession, but I like them. I get a feeling for how well pupils have done on a specific question when I mark, but after QLA'ing, I get cold, hard data that I can use to direct my intentions, and I love cold, hard data.

I don't see much value in the QLA document only providing information to the teacher, so I like my QLA to provide a printout for each pupil to give them a visual representation of their scores (in terms of Red, Amber and Green). This printout will also include three questions that are directed at the pupil, identified as being the three most important questions that I'd like them to be getting correct. Some pupils may have more that I want them to be getting correct, but these are the priorities.

I've been working on a new QLA spreadsheet recently, to 'supercharge' the information that I get. I might blog about that at another time.

3. If you're unaware, 'WAGOLL' stands for 'What a Good One Looks Like'. This could be in the form of a handwritten completed assessment, or it could be in the form of me completing the assessment on video. It could even be pictures of their own work, collated, to make a 'Best Of...' compilation. This allows pupils to re-engage with the questions and see what their answers should have been. The improvement in tech over recent years would probably see this uploaded to Google Classroom for pupils to access.

4. Whilst completing the QLA and preparing individualised printouts, I identified three key questions for pupils to complete, but in doing so, I've identified all of the key areas of deficiency within the understanding of my class. 

In my assessment feedback lesson, I would be sharing common wrong answers, why they're wrong, and what pupils should have done, and what's important here is that I'll follow this up with an opportunity for them to correct their mistakes in the form of example-problem pairs or backward faded activities. 'Going through the paper' with pupils annotating with the correct answers isn't enough - copying doesn't engage pupils in thinking, so it's important that we follow this up with pupils having something to 'do' with the feedback.

Following the example-problem pairs and backward faded activities pupils will be given their scores and printouts (these have been held back to increase curiosity and engagement), and the time to complete their targeted questions mimicking the exam conditions, allowing pupils to be successful with something that they previously weren't. I'll then mark these questions to check whether pupils have responded to the feedback they have been given.

5. The most important part of these lessons is that they should be focused on learning, rather than performance. We should all be well aware that simply because pupils are getting questions correct, doesn't mean that they've learnt it. Learning happens over time, rather than in the moment. With this in mind, it's important to provide opportunities for pupils to revisit this understanding at spaced intervals, to maximise the chance of pupils learning this stuff rather than just performing with it in the moment. I'd be adding these to my medium-term plans, to incorporate spaced retrieval opportunities a day later, a week later, a month later and three months later.

My aim is to provide pupils with many opportunities with which to learn from their mistakes. The 'WAGOLL' gives pupils who are keen and enthusiastic the opportunity to engage with the whole assessment. The individualised printout forces those who are less keen and less enthusiastic to engage with key material. The spacing of the ideas in terms of retrieval is the key aspect here, as learning occurs over time and performance is a poor proxy for learning. I want pupils to perform with the ideas many times, in multiple representations, which will imply that they've learnt the idea that they were previously struggling with.

This blog felt a little quick and short. If you have any queries, or if you have any suggestions to enhance this procedure in any way, please let me know how!

The Plan

So, my thoughts have settled, my plans are somewhat in place, and I'm hamstrung by events beyond my control.

Following their mock exams, I've seen my Year 10s once (immediately after their last exam, which was a maths paper) and they had a 'mock results day' on Friday (July 7). I had a timetabled lesson on that day and could not strike due to my union choices, but I was staffing a rewards trip for Year 7 and Year 8. My next lesson with them is on Tuesday (July 11) which is cut short by staff training, but I am also on an induction day at the school I am joining in September. I then have a lesson with them on Friday (July 14) and Tuesday (July 18) before we break up for the summer holidays.

The lack of time and a proper opportunity to follow up going into Year 11 has impacted what I intended to do, but here's the plan of action...

They sat two OCR papers for their mocks - a calculator and a non-calculator. They were given a revision list to aid their revision, a revision course on TUTOR, a practice paper geared towards their mock papers, and the opportunity to join us in the dining hall on the morning of their exams to settle their nerves. Paper 1 was the better-answered of the two papers, but both highlighted the need to revisit some key topics over the next year.

Due to the rewards trip and my induction day, I have had 150 minutes of cover to set, and I've been leaning on Backward Faded Maths, and the associated techniques, to resource these lessons.

The questions that I've chosen from paper 1 are questions on 'Reverse Compound Interest' (11% of the marks achieved), 'Angles in Polygons' (42%), 'Proportional Reasoning' (54%), 'Calculating With Bounds' (40%), 'Algebraic Proof' (36%) and 'Equation of a Circle' (39%).
The rest of the paper was better answered, but there are a few areas that I'll be reminded pupils about in a lesson (probably on Friday July 14), as well as a note that pupils need to revisit 'Circle Theorems' (26%) in more depth.

From paper 2, I'll be leaving backward faded activities on 'Writing Ratios in the form 1:n' (36%), 'Using Relative Frequency' (51%), 'Probability' (52%), 'Simplifying Surds' (18%), 'Multiplying Recurring Decimals' (51%).
These aren't the only needs, but they're the ones which are most suited to backward fading, and in my lesson on Tuesday, July 18 we're likely to spend some time looking at 'Calculating Percentage Increase' (51%), Indices and Powers (60%) and Rearranging Formulae (56%). I've also got more notes, with pupils needing to revisit 'Using Kinematics Formulae' (32%), 'Constructions and Loci' (37%), 'Enlargements from a Point' and 'Describing Transformations' (54% and 23%), 'Plotting and Using Quadratic Graphs' (28%) and 'Forming and Solving Equations' (14% and 10%).

The 100-minute cover lesson, with a non-specialist cover supervisor, that I set for Friday started with a 'Do Now' with solutions left, followed by six activities from Backward Faded Maths with 'follow-up' extension questions. The 50-minute lesson has been assigned with the same structure.

When I get back to seeing the pupils, I'll give them a 'Do Now' to be completed in exam conditions and then we'll mark them (comparing the mark schemes from their mocks to the 'shadow' questions), seeing the improvements made and adding these to their scores. We can then compare these to grade boundaries, to see if anyone tips over to the next grade. I've also, whilst marking, made scribbles of 'W' and 'M' within papers, where writing (more of) their Working would have had an impact, or where they've made a silly Mistake that they shouldn't be making. We'll add these to the scores we've just calculated and compare them to the grade boundaries again. The aim of this is to give them a confidence boost to increase motivation over the summer, as I hope that their first experience of sitting a 'formal' exam has been a learning process.

I'll finish up the lesson with a Q&A about their papers, where I'll answer any queries they might have about what they did in their exam and what they should have done, before sharing what this might look like as GCSE grades if they follow 'average models of progression from their Year 10 mock to GCSEs from our recent outcome data.

The lesson for their second paper will follow the same structure. This isn't what I'd liked to have done, but with time limited and a lack of an ability to continue this into next year, I hope that I've given it the best I can!

I might follow this up with an ideal structure for an assessment feedback routine, because I know that this is far from ideal!

What Am I Trying To Achieve?

The last five blogs have led to this one, and I think the main question that anyone should be asking when they do anything is 'What am I trying to achieve?'. A task without a goal, a purpose, feels a little pointless.

What am I trying to achieve when I'm feeding back assessments to my classes? Well, that's a big one.

Ultimately, I want them to be proud of what they've achieved and to be spurred on to greater things, or I want them to take away an impetus to improve in some way. Dylan Wiliam says that 'feedback should improve the learner', so I don't simply want an exam script that looks more like a rainbow and has a larger number on it than it did when I gave it back to them. I want pupils to have a moment of realisation. It could be as small as a mistake that they made is corrected (not just once, but for the rest of time) or it could be the understanding that if they begin to apply themselves more, and give more time to practicing maths, they can be better at it.

My Year 10s - the class that I'm going to trial this with - sat the first paper of their mock exams, and as it stands I have marked the papers, but haven't yet analysed them for individual areas for improvement, or whole-group areas for improvement. My plan is to sort this tomorrow, identifying areas of need for individuals/small groups, and identifying areas that the whole class need corrective instruction with.

In terms of what to do with the individuals, this tweet gave me some inspiration:

Get them to identify if it is a minor mistake or something they have to work on. Watch a video on DrFrost Maths or the model answer for the older ones which we make ourselves. Then they need to practice the skills on DrFrost Maths as our scheme of work/courses are linked on it

— Mrs Hudson (@MrsHuds43995523) January 8, 2023

We have included Mathswatch clip numbers on pupils' printouts from the QLA spreadsheet but it's always been self-directed. It's felt like a tick-box exercise, and it hasn't done much to improve the learner, their responsibility, or them as a mathematician. I'm tempted to record some example-problem pairs for key topics and share these with pupils, or provide backward-faded worked examples with further problems to be completed independently.
If I record videos I can post these on Google Classroom and use a computer room, whereas the backward-faded activities can be used in the classroom. I also have a couple of days that need cover setting coming up, and these two could be completed with cover teachers.

Pick 4 or 5 questions where the majority made silly mistakes. Go through them and then do shadow questions. They then compare the marks to see how many extra they would have got. It's superficial but it boosts their confidence for how they are going to move forward..

— Laura Jacobs (@Miss_L_Jacobs) January 9, 2023

I think I can set the videos/BFM tasks where pupils have made daft errors or have taken a wrong turn somewhere, and set the problem (in the example-problem pairs) as the shadow question. Or maybe I can write the problem part and they'll complete the shadow questions at a later date. This should allow me to discern performance from learning and they can re-do them in conditions akin to an exam hall for 20 minutes or so.

I have a couple of weeks until we can share the results with the pupils, and I'll formulate a plan between now and then. The next blog is likely to be the one which outlines my plans.

Take Responsibility

In a previous blog post, I wrote about three key learning points. In this blog, we'll look at the third of those...

Key learning point 3: Don't leave revision to pupils, and plan in activities to revisit key information.

I've worked in schools in challenging circumstances in Inner City Leeds for 15 years, and I love it. I love the relationships, I love the challenge and I love teaching. I am, however, a realist, and I wouldn't expect many of my pupils to be able to direct their own ongoing revision so that they were improving. They could definitely revisit ideas, but they'll go for the ones that make them feel good about themselves, or they'll go for the ones that everyone found hard, and this approach to learning and revision isn't one that will maximise the impact.

I'd be willing to bet that less than 10% of pupils I've taught over the last 15 years have had an area set aside for keeping their school work organised, never mind planning out their 'ongoing revision and study'.

With that in mind, I space out their learning for them and provided opportunities for retrieval so that ideas stick and learning occurs. Rome wasn't built in a day, and in a similar way, learning happens over time rather than in time. This means repeated exposure to ideas that pupils have succeeded with is important, so that learning occurs. So important that, following an assessment, we can't just expect pupils to be self-directed to revisit the things that they don't understand and we need to recognise that these areas of need should be revisited regularly.

We should take responsibility and present them with spaced repetition opportunities to improve the likelihood that learning DOES take place following the identification of this idea as an area of need from the assessment and following their first re-exposure in lessons.

To help with this, I use this spreadsheet. Adding topics into the 'Completed Learning' column will map the ideas forward in time by a day, a week, a month and three months (ish...) and provide a reminder to revisit these ideas with your classes.

In the next blog, I'll outline my plan for feeding back Year 10 mocks, taking into account others' responses to the original tweet.

Feedback in Moderation

In a prior blog, I wrote about three key learning points. In this blog, we'll focus on the second of those... 

Key learning point 2: Don't go through the whole paper. Focus on key questions.

This learning was in response to going through all of a paper, and whilst there's probably some merit to going through the whole thing I think you'll get more out of giving feedback in moderation.
Exercise is good for the body, until you push the body too far and cause injury (as I type this, I can feel my knees creaking), and the same can be said for feedback. Too much feedback can overwhelm our learners, but can also dilute the important messages that we're sending. We can be more impactful by focusing our feedback on the things that we need pupils to take away.

In response to key learning point 1, I stated that I find three areas of need for each pupil and try to maximise my gains by getting as much as possible out of each question I choose. In an ideal world, I'll be able to only revisit three questions, but it's more likely that we'll look at ten or so due to gaps in pupils' learning.

Only looking at three key skills would be great, but looking at ten or so questions rather than the 20 from the test still prevents pupils from becoming overwhelmed during our 'assessment feedback lesson', and we know that our feedback will garner more attention from pupils as we narrow our focus.

In the lesson, pupils will see the correct solution to the question from the assessment and be presented with a question with both a similar surface structure and a similar deep structure to complete, to develop near transfer. Follow-up questions could go on to alter either of these structures, but the likelihood is that we want pupils to develop confidence with these ideas without alteration as they've recently struggled with this idea in the assessment. We'll repeat this for the ideas that we're asking pupils to re-do.

They'll then work on three of these ideas identified from their assessment, independently, in order to give further attention to ideas that they've recently shown a lack of understanding over (in the 'test') and had success with (in our example-problem pairs).

This is how this is done:

I always mark tests and then enter the scores in a QLA spreadsheet. A template is here, and I recommend downloading it as .xlsx rather than using it within Google Sheets.

On the first tab, enter pupil names, and enter their scores. You can see the percentage score for the class across the bottom, and this is great for identifying needs. It has to be said that this data must be used in context, and specific issues should have been picked up during marking.

The second tab has a list of scores, and only scores, and for graded exams normally looks up grades and assigns these rather than having to type them individually.

The third tab is labelled 'Next Steps', and this requires further use of the first sheet. These are the needs identified for each pupil. I do this by highlighting cells, highlighting the question number at the top and then typing the question number in the 'NS' boxes at the end. This assigns individual questions for pupils to work on, and highlighting keeps track of the questions I need to go through in full as well as reminding me to write a new question in this third tab. Writing questions on this tab will mean that individualised questions are automatically assigned to pupils on their printouts when I go on to the fourth tab.

The fourth tab brings this all together. The mark for each question for each pupil is RAG'd when their surname is entered, and the identified questions for them to work on are on the same sheet, ready to be printed and given to them for completion independently of each other, to check that we've fed back has landed for each learner.

It's then incredibly important that pupils are given further retrieval opportunities, so they'll be written into my medium-term plans for starters/settlers/Do-Nows/homework as retrieval practice. They couldn't do it to begin with - revisiting the idea once in a feedback lesson when emotions are running high isn't going to solve the problem. This brings me back to Josh's tweet from the last instalment:

Yeah absolutely. I think when identifying the questions to go through we should ask ourselves ‘do they fundamentally not understand this? Or are they close to getting there?’

The latter are for review lessons, the former are for your medium term plans.

— Jshm (@jshmtn) January 8, 2023

The way that I've done this over the years has varied, but I've settled on this spreadsheet.
I write my medium-term plans in the 'Intended Learning' column and update this based on how well the lesson has gone.
In 'Completed Learning' I type a few words to describe the content covered. This is then mapped forward by one day, seven days, 31 days and 90 days.
This then forms my list of topics that need to be dropped in for retrieval opportunities. The dates don't always line up, so I keep track by shading 'covered' topics and adding them to the 'Completed Learning' column where there isn't a lesson in the event that I want it to come up in the rotation again.

Focus on Feedback, not on Performance

In the last blog, I wrote about three key learning points.

In this blog, we'll look at the first of these...

Key learning point 1: Try to give feedback without returning pupils' scores/grades/papers. This removes the emotional response to a score/grade, and maintains pupils' curiosity in the questions that the teacher has identified to go through.

When this was suggested to me, it sounded weird.
When I first tried it, it felt weird.
But I tried it and I've observed a trainee do the same. 

I'm sold. Pupils are more open to corrections, they're more curious about the answers, and they're more engaged in the feedback.

Logistically, this is how I've gone about feeding back.

1. Pupils complete a 'test', and I mark it before the next lesson. I've always valued a speedy turnaround, and like to plan my 'assessment lessons' so that I have time to mark and prepare an 'assessment feedback lesson' for the next lesson on the timetable.

2. I identify three areas of need for each pupil and set them follow up work on these questions. I try to make sure that each area of need catches as many pupils as I can, to minimise my work load and maximise the impact for the minimum effort, and also that the area of need isn't too far from their current level of understanding.
I do this by completing a QLA document (template here) and identifying the greatest areas of need by looking for the topics I expect 100% on, and seeing where gaps exist. I'll highlight the question in yellow, type the question number in at the end as their 'Next Step' and create similar questions in another sheet which are then sent to each pupils' individual printout. 

3. I write example-problem pairs for the areas of need. The example that I use is always the test question, as that develops motivation - pupils recognise it, and they want to know the answer. The problems have a similar surface structure and similar deep structure, as my aim is to develop near transfer. Additional problems can vary in terms of the surface structure or their deep structure, but my main aim is to develop near transfer with this idea for those that didn't get this right in the 'test'.

4. We run through the example-problem pairs as a class with whole-class response systems. Pupils answer questions on mini-whiteboards and hold them up on my say-so, or we use multiple-choice questions to draw out misconceptions. At this stage pupils don't know what they scored on each question, so engagement is maximised. Pupils are responding on mini-whiteboards or with their fingers, so participation is maximised.

5. The pupils complete independent work on their three areas of need on their individual printouts. This way, their independent work is focused on their areas of need, not on the simple stuff that others got wrong, or on the stuff they found way too hard.

6. It's important to keep this stuff 'live' in the classroom, with opportunities for retrieval prepared in future lessons. With this in mind, the topics are added to my medium term plans as settlers/starters/Do Nows/homeworks with retrieval in mind.

I've mentioned this in a previous blog, and it's definitely important - emotions run high in 'assessment feedback lessons'.
Either there's a feeling of contentment, of relief, after successes that leave the learner feeling relaxed and like they don't need the feedback because they've 'smashed it'. Not giving them their 'test' back avoids this lack of focus and maintains their motivation.
The other side of the coin is that the pupil has 'failed' the 'test' and is now questioning their self worth, and the point of working hard if they 'keep doing crap in tests'. Not giving them their 'test' back until later delays this emotional response, maintains motivation and increases participation in the example-problem pair part of the lesson.

Karen Hancock had this to say on the issue:

I rarely put marks on the students' work. I mark directly into a spreadsheet and then I'll either do the paper live for them to mark or we create video solutions and they watch those.
We usually have a shadow paper so that they can demonstrate they can do the questions now.

— Karen (@karenshancock) January 8, 2023

This strikes me as being similar to what I've done, in that there's no focus on performance, but supercharged. Completing a QLA is something I do anyway, so I think the only difference is a lack of writing on scripts. I would, however, be wary of marking live, as there can be so many slight differences that can score marks on a five-mark question (for example) that I'd feel as though I'm losing control of the lesson. The video solutions is something I could go with - my current thoughts are to share worked solutions (on something like Google Classroom), but recording me writing these and talking about my thought process would be useful to pupils, I expect.
Using a shadow paper is an excellent idea, too, and something we do, but I think that their use could be more impactful too. Maybe they don't need to complete the entire shadow paper, but I can pick and choose those questions based on pupil need.

Nicky Gale replied with this tweet:

I work in a sixth form college. We do a video solution for all the questions and a shadow paper. In the lesson, students watch videos and correct any they didn’t get full marks for, then try the similar questions. We then have an individual conversation with them and set targets.

— Nicky Gale (@nicnaks79) January 9, 2023

This sounds excellent, and where I'd like to go with Karen's suggestion, but I don't think I could do this for 30 pupils in a lesson. I'd love to handover responsibility to the pupils, to watch the videos, to correctly answer the questions, and then work on the similar questions, but I'm not sure that I'd get what I want out of some of my pupils. 
I feel like I need to direct this, and this is what Helen Scott had to say:

Pick 2-3 "should" have been able to model, DIRT questions after this. Ticks the policy box.
We all know the real learning comes from the retrieval and learning over time when they see the question again over starters and then in a quiz.

— Helen Scott (@HelenScott88) January 9, 2023

I feel as though 2-3 questions wouldn't cover everything from the paper in terms of pupil need, but I'm fully alongside the message of the tweet. I think choosing 2-3 per pupil, which might become 8-10 over the class is achievable and acceptable, with pupils working on the shadow questions for those 2-3 questions in terms of the 'feedback cycle' expected within schools, with (as Helen said) the real learning coming over time as pupils are re-exposed to the idea (using questions with a similar structure in the short term, and altering the surface or deep structure as we go).

In terms of how we go about this, Josh tweeted:

Yeah absolutely. I think when identifying the questions to go through we should ask ourselves ‘do they fundamentally not understand this? Or are they close to getting there?’

The latter are for review lessons, the former are for your medium term plans.

— Jshm (@jshmtn) January 8, 2023

This seems like an important point to end the blog on. Just because a pupil has got a question wrong doesn't mean that they don't understand the idea. On top of this, just because a pupil got a question correct doesn't mean they do understand the idea. Sometimes a pupil guesses, sometimes they fumble through their working, taking leaps and getting to the right answer despite not understanding the journey, and sometimes pupils struggle to apply their understanding in a high-pressure situation. 

Be wary of assessment data.